fukasawa e60969
#if 0
fukasawa e60969
CHANGES - changes for libpng
fukasawa e60969
fukasawa e60969
version 0.1 [March 29, 1995]
fukasawa e60969
  initial work-in-progress release
fukasawa e60969
fukasawa e60969
version 0.2 [April 1, 1995]
fukasawa e60969
  added reader into png.h
fukasawa e60969
  fixed small problems in stub file
fukasawa e60969
fukasawa e60969
version 0.3 [April 8, 1995]
fukasawa e60969
  added pull reader
fukasawa e60969
  split up pngwrite.c to several files
fukasawa e60969
  added pnglib.txt
fukasawa e60969
  added example.c
fukasawa e60969
  cleaned up writer, adding a few new transformations
fukasawa e60969
  fixed some bugs in writer
fukasawa e60969
  interfaced with zlib 0.5
fukasawa e60969
  added K&R support
fukasawa e60969
  added check for 64 KB blocks for 16 bit machines
fukasawa e60969
fukasawa e60969
version 0.4 [April 26, 1995]
fukasawa e60969
  cleaned up code and commented code
fukasawa e60969
  simplified time handling into png_time
fukasawa e60969
  created png_color_16 and png_color_8 to handle color needs
fukasawa e60969
  cleaned up color type defines
fukasawa e60969
  fixed various bugs
fukasawa e60969
  made various names more consistent
fukasawa e60969
  interfaced with zlib 0.71
fukasawa e60969
  cleaned up zTXt reader and writer (using zlib's Reset functions)
fukasawa e60969
  split transformations into pngrtran.c and pngwtran.c
fukasawa e60969
fukasawa e60969
version 0.5 [April 30, 1995]
fukasawa e60969
  interfaced with zlib 0.8
fukasawa e60969
  fixed many reading and writing bugs
fukasawa e60969
  saved using 3 spaces instead of tabs
fukasawa e60969
fukasawa e60969
version 0.6 [May 1, 1995]
fukasawa e60969
  first beta release
fukasawa e60969
  added png_large_malloc() and png_large_free()
fukasawa e60969
  added png_size_t
fukasawa e60969
  cleaned up some compiler warnings
fukasawa e60969
  added png_start_read_image()
fukasawa e60969
fukasawa e60969
version 0.7 [June 24, 1995]
fukasawa e60969
  cleaned up lots of bugs
fukasawa e60969
  finished dithering and other stuff
fukasawa e60969
  added test program
fukasawa e60969
  changed name from pnglib to libpng
fukasawa e60969
fukasawa e60969
version 0.71 [June 26, 1995]
fukasawa e60969
  changed pngtest.png for zlib 0.93
fukasawa e60969
  fixed error in libpng.txt and example.c
fukasawa e60969
fukasawa e60969
version 0.8 [August 20, 1995]
fukasawa e60969
  cleaned up some bugs
fukasawa e60969
  added png_set_filler()
fukasawa e60969
  split up pngstub.c into pngmem.c, pngio.c, and pngerror.c
fukasawa e60969
  added #define's to remove unwanted code
fukasawa e60969
  moved png_info_init() to png.c
fukasawa e60969
  added old_size into png_realloc()
fukasawa e60969
  added functions to manually set filtering and compression info
fukasawa e60969
  changed compression parameters based on image type
fukasawa e60969
  optimized filter selection code
fukasawa e60969
  added version info
fukasawa e60969
  changed external functions passing floats to doubles (k&r problems?)
fukasawa e60969
  put all the configurable stuff in pngconf.h
fukasawa e60969
  enabled png_set_shift to work with paletted images on read
fukasawa e60969
  added png_read_update_info() - updates info structure with transformations
fukasawa e60969
fukasawa e60969
Version 0.81 [August, 1995]
fukasawa e60969
  incorporated Tim Wegner's medium model code (thanks, Tim)
fukasawa e60969
fukasawa e60969
Version 0.82 [September, 1995]
fukasawa e60969
  [unspecified changes]
fukasawa e60969
fukasawa e60969
Version 0.85 [December, 1995]
fukasawa e60969
  added more medium model code (almost everything's a far)
fukasawa e60969
  added i/o, error, and memory callback functions
fukasawa e60969
  fixed some bugs (16-bit, 4-bit interlaced, etc.)
fukasawa e60969
  added first run progressive reader (barely tested)
fukasawa e60969
fukasawa e60969
Version 0.86 [January, 1996]
fukasawa e60969
  fixed bugs
fukasawa e60969
  improved documentation
fukasawa e60969
fukasawa e60969
Version 0.87 [January, 1996]
fukasawa e60969
  fixed medium model bugs
fukasawa e60969
  fixed other bugs introduced in 0.85 and 0.86
fukasawa e60969
  added some minor documentation
fukasawa e60969
fukasawa e60969
Version 0.88 [January, 1996]
fukasawa e60969
  fixed progressive bugs
fukasawa e60969
  replaced tabs with spaces
fukasawa e60969
  cleaned up documentation
fukasawa e60969
  added callbacks for read/write and warning/error functions
fukasawa e60969
fukasawa e60969
Version 0.89 [June 5, 1996]
fukasawa e60969
  Added new initialization API to make libpng work better with shared libs
fukasawa e60969
    we now have png_create_read_struct(), png_create_write_struct(),
fukasawa e60969
    png_create_info_struct(), png_destroy_read_struct(), and
fukasawa e60969
    png_destroy_write_struct() instead of the separate calls to
fukasawa e60969
    malloc and png_read_init(), png_info_init(), and png_write_init()
fukasawa e60969
  Changed warning/error callback functions to fix bug - this means you
fukasawa e60969
    should use the new initialization API if you were using the old
fukasawa e60969
    png_set_message_fn() calls, and that the old API no longer exists
fukasawa e60969
    so that people are aware that they need to change their code
fukasawa e60969
  Changed filter selection API to allow selection of multiple filters
fukasawa e60969
    since it didn't work in previous versions of libpng anyways
fukasawa e60969
  Optimized filter selection code
fukasawa e60969
  Fixed png_set_background() to allow using an arbitrary RGB color for
fukasawa e60969
    paletted images
fukasawa e60969
  Fixed gamma and background correction for paletted images, so
fukasawa e60969
    png_correct_palette is not needed unless you are correcting an
fukasawa e60969
    external palette (you will need to #define PNG_CORRECT_PALETTE_SUPPORTED
fukasawa e60969
    in pngconf.h) - if nobody uses this, it may disappear in the future.
fukasawa e60969
  Fixed bug with Borland 64K memory allocation (Alexander Lehmann)
fukasawa e60969
  Fixed bug in interlace handling (Smarasderagd, I think)
fukasawa e60969
  Added more error checking for writing and image to reduce invalid files
fukasawa e60969
  Separated read and write functions so that they won't both be linked
fukasawa e60969
    into a binary when only reading or writing functionality is used
fukasawa e60969
  New pngtest image also has interlacing and zTXt
fukasawa e60969
  Updated documentation to reflect new API
fukasawa e60969
fukasawa e60969
Version 0.89c [June 17, 1996]
fukasawa e60969
  Bug fixes.
fukasawa e60969
fukasawa e60969
Version 0.90 [January, 1997]
fukasawa e60969
  Made CRC errors/warnings on critical and ancillary chunks configurable
fukasawa e60969
  libpng will use the zlib CRC routines by (compile-time) default
fukasawa e60969
  Changed DOS small/medium model memory support - needs zlib 1.04 (Tim Wegner)
fukasawa e60969
  Added external C++ wrapper statements to png.h (Gilles Dauphin)
fukasawa e60969
  Allow PNG file to be read when some or all of file signature has already
fukasawa e60969
    been read from the beginning of the stream.  ****This affects the size
fukasawa e60969
    of info_struct and invalidates all programs that use a shared libpng****
fukasawa e60969
  Fixed png_filler() declarations
fukasawa e60969
  Fixed? background color conversions
fukasawa e60969
  Fixed order of error function pointers to match documentation
fukasawa e60969
  Current chunk name is now available in png_struct to reduce the number
fukasawa e60969
    of nearly identical error messages (will simplify multi-lingual
fukasawa e60969
    support when available)
fukasawa e60969
  Try to get ready for unknown-chunk callback functions:
fukasawa e60969
    - previously read critical chunks are flagged, so the chunk handling
fukasawa e60969
      routines can determine if the chunk is in the right place
fukasawa e60969
    - all chunk handling routines have the same prototypes, so we will
fukasawa e60969
      be able to handle all chunks via a callback mechanism
fukasawa e60969
  Try to fix Linux "setjmp" buffer size problems
fukasawa e60969
  Removed png_large_malloc, png_large_free, and png_realloc functions.
fukasawa e60969
fukasawa e60969
Version 0.95 [March, 1997]
fukasawa e60969
  Fixed bug in pngwutil.c allocating "up_row" twice and "avg_row" never
fukasawa e60969
  Fixed bug in PNG file signature compares when start != 0
fukasawa e60969
  Changed parameter type of png_set_filler(...filler...) from png_byte
fukasawa e60969
    to png_uint_32
fukasawa e60969
  Added test for MACOS to ensure that both math.h and fp.h are not #included
fukasawa e60969
  Added macros for libpng to be compiled as a Windows DLL (Andreas Kupries)
fukasawa e60969
  Added "packswap" transformation, which changes the endianness of
fukasawa e60969
    packed-pixel bytes (Kevin Bracey)
fukasawa e60969
  Added "strip_alpha" transformation, which removes the alpha channel of
fukasawa e60969
    input images without using it (not necessarily a good idea)
fukasawa e60969
  Added "swap_alpha" transformation, which puts the alpha channel in front
fukasawa e60969
    of the color bytes instead of after
fukasawa e60969
  Removed all implicit variable tests which assume NULL == 0 (I think)
fukasawa e60969
  Changed several variables to "png_size_t" to show 16/32-bit limitations
fukasawa e60969
  Added new pCAL chunk read/write support
fukasawa e60969
  Added experimental filter selection weighting (Greg Roelofs)
fukasawa e60969
  Removed old png_set_rgbx() and png_set_xrgb() functions that have been
fukasawa e60969
    obsolete for about 2 years now (use png_set_filler() instead)
fukasawa e60969
  Added macros to read 16- and 32-bit ints directly from buffer, to be
fukasawa e60969
    used only on those systems that support it (namely PowerPC and 680x0)
fukasawa e60969
    With some testing, this may become the default for MACOS/PPC systems.
fukasawa e60969
  Only calculate CRC on data if we are going to use it
fukasawa e60969
  Added macros for zTXt compression type PNG_zTXt_COMPRESSION_???
fukasawa e60969
  Added macros for simple libpng debugging output selectable at compile time
fukasawa e60969
  Removed PNG_READ_END_MODE in progressive reader (Smarasderagd)
fukasawa e60969
  More description of info_struct in libpng.txt and png.h
fukasawa e60969
  More instructions in example.c
fukasawa e60969
  More chunk types tested in pngtest.c
fukasawa e60969
  Renamed pngrcb.c to pngset.c, and all png_read_<chunk> functions to be
fukasawa e60969
    png_set_<chunk>.  We now have corresponding png_get_<chunk>
fukasawa e60969
    functions in pngget.c to get information in info_ptr.  This isolates
fukasawa e60969
    the application from the internal organization of png_info_struct
fukasawa e60969
    (good for shared library implementations).
fukasawa e60969
fukasawa e60969
Version 0.96 [May, 1997]
fukasawa e60969
  Fixed serious bug with < 8bpp images introduced in 0.95
fukasawa e60969
  Fixed 256-color transparency bug (Greg Roelofs)
fukasawa e60969
  Fixed up documentation (Greg Roelofs, Laszlo Nyul)
fukasawa e60969
  Fixed "error" in pngconf.h for Linux setjmp() behavior
fukasawa e60969
  Fixed DOS medium model support (Tim Wegner)
fukasawa e60969
  Fixed png_check_keyword() for case with error in static string text
fukasawa e60969
  Added read of CRC after IEND chunk for embedded PNGs (Laszlo Nyul)
fukasawa e60969
  Added typecasts to quiet compiler errors
fukasawa e60969
  Added more debugging info
fukasawa e60969
fukasawa e60969
Version 0.97 [January, 1998]
fukasawa e60969
  Removed PNG_USE_OWN_CRC capability
fukasawa e60969
  Relocated png_set_crc_action from pngrutil.c to pngrtran.c
fukasawa e60969
  Fixed typecasts of "new_key", etc. (Andreas Dilger)
fukasawa e60969
  Added RFC 1152 [sic] date support
fukasawa e60969
  Fixed bug in gamma handling of 4-bit grayscale
fukasawa e60969
  Added 2-bit grayscale gamma handling (Glenn R-P)
fukasawa e60969
  Added more typecasts. 65536L becomes (png_uint_32)65536L, etc. (Glenn R-P)
fukasawa e60969
  Minor corrections in libpng.txt
fukasawa e60969
  Added simple sRGB support (Glenn R-P)
fukasawa e60969
  Easier conditional compiling, e.g.,
fukasawa e60969
    define PNG_READ/WRITE_NOT_FULLY_SUPPORTED;
fukasawa e60969
    all configurable options can be selected from command-line instead
fukasawa e60969
    of having to edit pngconf.h (Glenn R-P)
fukasawa e60969
  Fixed memory leak in pngwrite.c (free info_ptr->text) (Glenn R-P)
fukasawa e60969
  Added more conditions for png_do_background, to avoid changing
fukasawa e60969
    black pixels to background when a background is supplied and
fukasawa e60969
    no pixels are transparent
fukasawa e60969
  Repaired PNG_NO_STDIO behavior
fukasawa e60969
  Tested NODIV support and made it default behavior (Greg Roelofs)
fukasawa e60969
  Added "-m" option and PNGTEST_DEBUG_MEMORY to pngtest (John Bowler)
fukasawa e60969
  Regularized version numbering scheme and bumped shared-library major
fukasawa e60969
    version number to 2 to avoid problems with libpng 0.89 apps
fukasawa e60969
    (Greg Roelofs)
fukasawa e60969
fukasawa e60969
Version 0.98 [January, 1998]
fukasawa e60969
  Cleaned up some typos in libpng.txt and in code documentation
fukasawa e60969
  Fixed memory leaks in pCAL chunk processing (Glenn R-P and John Bowler)
fukasawa e60969
  Cosmetic change "display_gamma" to "screen_gamma" in pngrtran.c
fukasawa e60969
  Changed recommendation about file_gamma for PC images to .51 from .45,
fukasawa e60969
    in example.c and libpng.txt, added comments to distinguish between
fukasawa e60969
    screen_gamma, viewing_gamma, and display_gamma.
fukasawa e60969
  Changed all references to RFC1152 to read RFC1123 and changed the
fukasawa e60969
    PNG_TIME_RFC1152_SUPPORTED macro to PNG_TIME_RFC1123_SUPPORTED
fukasawa e60969
  Added png_invert_alpha capability (Glenn R-P -- suggestion by Jon Vincent)
fukasawa e60969
  Changed srgb_intent from png_byte to int to avoid compiler bugs
fukasawa e60969
fukasawa e60969
Version 0.99 [January 30, 1998]
fukasawa e60969
  Free info_ptr->text instead of end_info_ptr->text in pngread.c (John Bowler)
fukasawa e60969
  Fixed a longstanding "packswap" bug in pngtrans.c
fukasawa e60969
  Fixed some inconsistencies in pngconf.h that prevented compiling with
fukasawa e60969
    PNG_READ_GAMMA_SUPPORTED and PNG_READ_hIST_SUPPORTED undefined
fukasawa e60969
  Fixed some typos and made other minor rearrangement of libpng.txt (Andreas)
fukasawa e60969
  Changed recommendation about file_gamma for PC images to .50 from .51 in
fukasawa e60969
    example.c and libpng.txt, and changed file_gamma for sRGB images to .45
fukasawa e60969
  Added a number of functions to access information from the png structure
fukasawa e60969
    png_get_image_height(), etc. (Glenn R-P, suggestion by Brad Pettit)
fukasawa e60969
  Added TARGET_MACOS similar to zlib-1.0.8
fukasawa e60969
  Define PNG_ALWAYS_EXTERN when __MWERKS__ && WIN32 are defined
fukasawa e60969
  Added type casting to all png_malloc() function calls
fukasawa e60969
fukasawa e60969
Version 0.99a [January 31, 1998]
fukasawa e60969
  Added type casts and parentheses to all returns that return a value.(Tim W.)
fukasawa e60969
fukasawa e60969
Version 0.99b [February 4, 1998]
fukasawa e60969
  Added type cast png_uint_32 on malloc function calls where needed.
fukasawa e60969
  Changed type of num_hist from png_uint_32 to int (same as num_palette).
fukasawa e60969
  Added checks for rowbytes overflow, in case png_size_t is less than 32 bits.
fukasawa e60969
  Renamed makefile.elf to makefile.lnx.
fukasawa e60969
fukasawa e60969
Version 0.99c [February 7, 1998]
fukasawa e60969
  More type casting.  Removed erroneous overflow test in pngmem.c.
fukasawa e60969
  Added png_buffered_memcpy() and png_buffered_memset(), apply them to rowbytes.
fukasawa e60969
  Added UNIX manual pages libpng.3 (incorporating libpng.txt) and  png.5.
fukasawa e60969
fukasawa e60969
Version 0.99d [February 11, 1998]
fukasawa e60969
  Renamed "far_to_near()" "png_far_to_near()"
fukasawa e60969
  Revised libpng.3
fukasawa e60969
  Version 99c "buffered" operations didn't work as intended.  Replaced them
fukasawa e60969
    with png_memcpy_check() and png_memset_check().
fukasawa e60969
  Added many "if (png_ptr == NULL) return" to quell compiler warnings about
fukasawa e60969
    unused png_ptr, mostly in pngget.c and pngset.c.
fukasawa e60969
  Check for overlength tRNS chunk present when indexed-color PLTE is read.
fukasawa e60969
  Cleaned up spelling errors in libpng.3/libpng.txt
fukasawa e60969
  Corrected a problem with png_get_tRNS() which returned undefined trans array
fukasawa e60969
fukasawa e60969
Version 0.99e [February 28, 1998]
fukasawa e60969
  Corrected png_get_tRNS() again.
fukasawa e60969
  Add parentheses for easier reading of pngget.c, fixed "||" should be "&&".
fukasawa e60969
  Touched up example.c to make more of it compileable, although the entire
fukasawa e60969
    file still can't be compiled (Willem van Schaik)
fukasawa e60969
  Fixed a bug in png_do_shift() (Bryan Tsai)
fukasawa e60969
  Added a space in png.h prototype for png_write_chunk_start()
fukasawa e60969
  Replaced pngtest.png with one created with zlib 1.1.1
fukasawa e60969
  Changed pngtest to report PASS even when file size is different (Jean-loup G.)
fukasawa e60969
  Corrected some logic errors in png_do_invert_alpha() (Chris Patterson)
fukasawa e60969
fukasawa e60969
Version 0.99f [March 5, 1998]
fukasawa e60969
  Corrected a bug in pngpread() introduced in version 99c (Kevin Bracey)
fukasawa e60969
  Moved makefiles into a "scripts" directory, and added INSTALL instruction file
fukasawa e60969
  Added makefile.os2 and pngos2.def (A. Zabolotny) and makefile.s2x (W. Sebok)
fukasawa e60969
  Added pointers to "note on libpng versions" in makefile.lnx and README
fukasawa e60969
  Added row callback feature when reading and writing nonprogressive rows
fukasawa e60969
    and added a test of this feature in pngtest.c
fukasawa e60969
  Added user transform callbacks, with test of the feature in pngtest.c
fukasawa e60969
fukasawa e60969
Version 0.99g [March 6, 1998, morning]
fukasawa e60969
  Minor changes to pngtest.c to suppress compiler warnings.
fukasawa e60969
  Removed "beta" language from documentation.
fukasawa e60969
fukasawa e60969
Version 0.99h [March 6, 1998, evening]
fukasawa e60969
  Minor changes to previous minor changes to pngtest.c
fukasawa e60969
  Changed PNG_READ_NOT_FULLY_SUPPORTED to PNG_READ_TRANSFORMS_NOT_SUPPORTED
fukasawa e60969
    and added PNG_PROGRESSIVE_READ_NOT_SUPPORTED macro
fukasawa e60969
  Added user transform capability
fukasawa e60969
fukasawa e60969
Version 1.00 [March 7, 1998]
fukasawa e60969
  Changed several typedefs in pngrutil.c
fukasawa e60969
  Added makefile.wat (Pawel Mrochen), updated makefile.tc3 (Willem van Schaik)
fukasawa e60969
  Replaced "while(1)" with "for(;;)"
fukasawa e60969
  Added PNGARG() to prototypes in pngtest.c and removed some prototypes
fukasawa e60969
  Updated some of the makefiles (Tom Lane)
fukasawa e60969
  Changed some typedefs (s_start, etc.) in pngrutil.c
fukasawa e60969
  Fixed dimensions of "short_months" array in pngwrite.c
fukasawa e60969
  Replaced ansi2knr.c with the one from jpeg-v6
fukasawa e60969
fukasawa e60969
Version 1.0.0 [March 8, 1998]
fukasawa e60969
  Changed name from 1.00 to 1.0.0 (Adam Costello)
fukasawa e60969
  Added smakefile.ppc (with SCOPTIONS.ppc) for Amiga PPC (Andreas Kleinert)
fukasawa e60969
fukasawa e60969
Version 1.0.0a [March 9, 1998]
fukasawa e60969
  Fixed three bugs in pngrtran.c to make gamma+background handling consistent
fukasawa e60969
    (Greg Roelofs)
fukasawa e60969
  Changed format of the PNG_LIBPNG_VER integer to xyyzz instead of xyz
fukasawa e60969
    for major, minor, and bugfix releases.  This is 10001. (Adam Costello,
fukasawa e60969
    Tom Lane)
fukasawa e60969
  Make months range from 1-12 in png_convert_to_rfc1123
fukasawa e60969
fukasawa e60969
Version 1.0.0b [March 13, 1998]
fukasawa e60969
  Quieted compiler complaints about two empty "for" loops in pngrutil.c
fukasawa e60969
  Minor changes to makefile.s2x
fukasawa e60969
  Removed #ifdef/#endif around a png_free() in pngread.c
fukasawa e60969
fukasawa e60969
Version 1.0.1 [March 14, 1998]
fukasawa e60969
  Changed makefile.s2x to reduce security risk of using a relative pathname
fukasawa e60969
  Fixed some typos in the documentation (Greg).
fukasawa e60969
  Fixed a problem with value of "channels" returned by png_read_update_info()
fukasawa e60969
fukasawa e60969
Version 1.0.1a [April 21, 1998]
fukasawa e60969
  Optimized Paeth calculations by replacing abs() function calls with intrinsics
fukasawa e60969
  plus other loop optimizations. Improves avg decoding speed by about 20%.
fukasawa e60969
  Commented out i386istic "align" compiler flags in makefile.lnx.
fukasawa e60969
  Reduced the default warning level in some makefiles, to make them consistent.
fukasawa e60969
  Removed references to IJG and JPEG in the ansi2knr.c copyright statement.
fukasawa e60969
  Fixed a bug in png_do_strip_filler with XXRRGGBB => RRGGBB transformation.
fukasawa e60969
  Added grayscale and 16-bit capability to png_do_read_filler().
fukasawa e60969
  Fixed a bug in pngset.c, introduced in version 0.99c, that sets rowbytes
fukasawa e60969
    too large when writing an image with bit_depth < 8 (Bob Dellaca).
fukasawa e60969
  Corrected some bugs in the experimental weighted filtering heuristics.
fukasawa e60969
  Moved a misplaced pngrutil code block that truncates tRNS if it has more
fukasawa e60969
    than num_palette entries -- test was done before num_palette was defined.
fukasawa e60969
  Fixed a png_convert_to_rfc1123() bug that converts day 31 to 0 (Steve Eddins).
fukasawa e60969
  Changed compiler flags in makefile.wat for better optimization
fukasawa e60969
    (Pawel Mrochen).
fukasawa e60969
fukasawa e60969
Version 1.0.1b [May 2, 1998]
fukasawa e60969
  Relocated png_do_gray_to_rgb() within png_do_read_transformations() (Greg).
fukasawa e60969
  Relocated the png_composite macros from pngrtran.c to png.h (Greg).
fukasawa e60969
  Added makefile.sco (contributed by Mike Hopkirk).
fukasawa e60969
  Fixed two bugs (missing definitions of "istop") introduced in libpng-1.0.1a.
fukasawa e60969
  Fixed a bug in pngrtran.c that would set channels=5 under some circumstances.
fukasawa e60969
  More work on the Paeth-filtering, achieving imperceptible speedup
fukasawa e60969
    (A Kleinert).
fukasawa e60969
  More work on loop optimization which may help when compiled with C++
fukasawa e60969
    compilers.
fukasawa e60969
  Added warnings when people try to use transforms they've defined out.
fukasawa e60969
  Collapsed 4 "i" and "c" loops into single "i" loops in pngrtran and pngwtran.
fukasawa e60969
  Revised paragraph about png_set_expand() in libpng.txt and libpng.3 (Greg)
fukasawa e60969
fukasawa e60969
Version 1.0.1c [May 11, 1998]
fukasawa e60969
  Fixed a bug in pngrtran.c (introduced in libpng-1.0.1a) where the masks for
fukasawa e60969
    filler bytes should have been 0xff instead of 0xf.
fukasawa e60969
  Added max_pixel_depth=32 in pngrutil.c when using FILLER with palette images.
fukasawa e60969
  Moved PNG_WRITE_WEIGHTED_FILTER_SUPPORTED and PNG_WRITE_FLUSH_SUPPORTED
fukasawa e60969
    out of the PNG_WRITE_TRANSFORMS_NOT_SUPPORTED block of pngconf.h
fukasawa e60969
  Added "PNG_NO_WRITE_TRANSFORMS" etc., as alternatives for *_NOT_SUPPORTED,
fukasawa e60969
    for consistency, in pngconf.h
fukasawa e60969
  Added individual "ifndef PNG_NO_[CAPABILITY]" in pngconf.h to make it easier
fukasawa e60969
    to remove unwanted capabilities via the compile line
fukasawa e60969
  Made some corrections to grammar (which, it's) in documentation (Greg).
fukasawa e60969
  Corrected example.c, use of row_pointers in png_write_image().
fukasawa e60969
fukasawa e60969
Version 1.0.1d [May 24, 1998]
fukasawa e60969
  Corrected several statements that used side effects illegally in pngrutil.c
fukasawa e60969
    and pngtrans.c, that were introduced in version 1.0.1b
fukasawa e60969
  Revised png_read_rows() to avoid repeated if-testing for NULL (A Kleinert)
fukasawa e60969
  More corrections to example.c, use of row_pointers in png_write_image()
fukasawa e60969
    and png_read_rows().
fukasawa e60969
  Added pngdll.mak and pngdef.pas to scripts directory, contributed by
fukasawa e60969
    Bob Dellaca, to make a png32bd.dll with Borland C++ 4.5
fukasawa e60969
  Fixed error in example.c with png_set_text: num_text is 3, not 2 (Guido V.)
fukasawa e60969
  Changed several loops from count-down to count-up, for consistency.
fukasawa e60969
fukasawa e60969
Version 1.0.1e [June 6, 1998]
fukasawa e60969
  Revised libpng.txt and libpng.3 description of png_set_read|write_fn(), and
fukasawa e60969
    added warnings when people try to set png_read_fn and png_write_fn in
fukasawa e60969
    the same structure.
fukasawa e60969
  Added a test such that png_do_gamma will be done when num_trans==0
fukasawa e60969
    for truecolor images that have defined a background.  This corrects an
fukasawa e60969
    error that was introduced in libpng-0.90 that can cause gamma processing
fukasawa e60969
    to be skipped.
fukasawa e60969
  Added tests in png.h to include "trans" and "trans_values" in structures
fukasawa e60969
    when PNG_READ_BACKGROUND_SUPPORTED or PNG_READ_EXPAND_SUPPORTED is defined.
fukasawa e60969
  Add png_free(png_ptr->time_buffer) in png_destroy_read_struct()
fukasawa e60969
  Moved png_convert_to_rfc_1123() from pngwrite.c to png.c
fukasawa e60969
  Added capability for user-provided malloc_fn() and free_fn() functions,
fukasawa e60969
    and revised pngtest.c to demonstrate their use, replacing the
fukasawa e60969
    PNGTEST_DEBUG_MEM feature.
fukasawa e60969
  Added makefile.w32, for Microsoft C++ 4.0 and later (Tim Wegner).
fukasawa e60969
fukasawa e60969
Version 1.0.2 [June 14, 1998]
fukasawa e60969
  Fixed two bugs in makefile.bor .
fukasawa e60969
fukasawa e60969
Version 1.0.2a [December 30, 1998]
fukasawa e60969
  Replaced and extended code that was removed from png_set_filler() in 1.0.1a.
fukasawa e60969
  Fixed a bug in png_do_filler() that made it fail to write filler bytes in
fukasawa e60969
    the left-most pixel of each row (Kevin Bracey).
fukasawa e60969
  Changed "static pngcharp tIME_string" to "static char tIME_string[30]"
fukasawa e60969
    in pngtest.c (Duncan Simpson).
fukasawa e60969
  Fixed a bug in pngtest.c that caused pngtest to try to write a tIME chunk
fukasawa e60969
    even when no tIME chunk was present in the source file.
fukasawa e60969
  Fixed a problem in pngrutil.c: gray_to_rgb didn't always work with 16-bit.
fukasawa e60969
  Fixed a problem in png_read_push_finish_row(), which would not skip some
fukasawa e60969
    passes that it should skip, for images that are less than 3 pixels high.
fukasawa e60969
  Interchanged the order of calls to png_do_swap() and png_do_shift()
fukasawa e60969
    in pngwtran.c (John Cromer).
fukasawa e60969
  Added #ifdef PNG_DEBUG/#endif surrounding use of PNG_DEBUG in png.h .
fukasawa e60969
  Changed "bad adaptive filter type" from error to warning in pngrutil.c .
fukasawa e60969
  Fixed a documentation error about default filtering with 8-bit indexed-color.
fukasawa e60969
  Separated the PNG_NO_STDIO macro into PNG_NO_STDIO and PNG_NO_CONSOLE_IO
fukasawa e60969
    (L. Peter Deutsch).
fukasawa e60969
  Added png_set_rgb_to_gray() and png_get_rgb_to_gray_status() functions.
fukasawa e60969
  Added png_get_copyright() and png_get_header_version() functions.
fukasawa e60969
  Revised comments on png_set_progressive_read_fn() in libpng.txt and example.c
fukasawa e60969
  Added information about debugging in libpng.txt and libpng.3 .
fukasawa e60969
  Changed "ln -sf" to "ln -s -f" in makefile.s2x, makefile.lnx, and
fukasawa e60969
    makefile.sco.
fukasawa e60969
  Removed lines after Dynamic Dependencies" in makefile.aco .
fukasawa e60969
  Revised makefile.dec to make a shared library (Jeremie Petit).
fukasawa e60969
  Removed trailing blanks from all files.
fukasawa e60969
fukasawa e60969
Version 1.0.2a [January 6, 1999]
fukasawa e60969
  Removed misplaced #endif and #ifdef PNG_NO_EXTERN near the end of png.h
fukasawa e60969
  Added "if" tests to silence complaints about unused png_ptr in png.h and png.c
fukasawa e60969
  Changed "check_if_png" function in example.c to return true (nonzero) if PNG.
fukasawa e60969
  Changed libpng.txt to demonstrate png_sig_cmp() instead of png_check_sig()
fukasawa e60969
    which is obsolete.
fukasawa e60969
fukasawa e60969
Version 1.0.3 [January 14, 1999]
fukasawa e60969
  Added makefile.hux, for Hewlett Packard HPUX 10.20 and 11.00 (Jim Rice)
fukasawa e60969
  Added a statement of Y2K compliance in png.h, libpng.3, and Y2KINFO.
fukasawa e60969
fukasawa e60969
Version 1.0.3a [August 12, 1999]
fukasawa e60969
  Added check for PNG_READ_INTERLACE_SUPPORTED in pngread.c; issue a warning
fukasawa e60969
    if an attempt is made to read an interlaced image when it's not supported.
fukasawa e60969
  Added check if png_ptr->trans is defined before freeing it in pngread.c
fukasawa e60969
  Modified the Y2K statement to include versions back to version 0.71
fukasawa e60969
  Fixed a bug in the check for valid IHDR bit_depth/color_types in pngrutil.c
fukasawa e60969
  Modified makefile.wat (added -zp8 flag, ".symbolic", changed some comments)
fukasawa e60969
  Replaced leading blanks with tab characters in makefile.hux
fukasawa e60969
  Changed "dworkin.wustl.edu" to "ccrc.wustl.edu" in various documents.
fukasawa e60969
  Changed (float)red and (float)green to (double)red, (double)green
fukasawa e60969
    in png_set_rgb_to_gray() to avoid "promotion" problems in AIX.
fukasawa e60969
  Fixed a bug in pngconf.h that omitted <stdio.h> when PNG_DEBUG==0 (K Bracey).
fukasawa e60969
  Reformatted libpng.3 and libpngpf.3 with proper fonts (script by J. vanZandt).
fukasawa e60969
  Updated documentation to refer to the PNG-1.2 specification.
fukasawa e60969
  Removed ansi2knr.c and left pointers to the latest source for ansi2knr.c
fukasawa e60969
    in makefile.knr, INSTALL, and README (L. Peter Deutsch)
fukasawa e60969
  Fixed bugs in calculation of the length of rowbytes when adding alpha
fukasawa e60969
    channels to 16-bit images, in pngrtran.c (Chris Nokleberg)
fukasawa e60969
  Added function png_set_user_transform_info() to store user_transform_ptr,
fukasawa e60969
    user_depth, and user_channels into the png_struct, and a function
fukasawa e60969
    png_get_user_transform_ptr() to retrieve the pointer (Chris Nokleberg)
fukasawa e60969
  Added function png_set_empty_plte_permitted() to make libpng useable
fukasawa e60969
    in MNG applications.
fukasawa e60969
  Corrected the typedef for png_free_ptr in png.h (Jesse Jones).
fukasawa e60969
  Correct gamma with srgb is 45455 instead of 45000 in pngrutil.c, to be
fukasawa e60969
    consistent with PNG-1.2, and allow variance of 500 before complaining.
fukasawa e60969
  Added assembler code contributed by Intel in file pngvcrd.c and modified
fukasawa e60969
    makefile.w32 to use it (Nirav Chhatrapati, INTEL Corporation,
fukasawa e60969
    Gilles Vollant)
fukasawa e60969
  Changed "ln -s -f" to "ln -f -s" in the makefiles to make Solaris happy.
fukasawa e60969
  Added some aliases for png_set_expand() in pngrtran.c, namely
fukasawa e60969
    png_set_expand_PLTE(), png_set_expand_depth(), and png_set_expand_tRNS()
fukasawa e60969
    (Greg Roelofs, in "PNG: The Definitive Guide").
fukasawa e60969
  Added makefile.beo for BEOS on X86, contributed by Sander Stok.
fukasawa e60969
fukasawa e60969
Version 1.0.3b [August 26, 1999]
fukasawa e60969
  Replaced 2147483647L several places with PNG_MAX_UINT macro, defined in png.h
fukasawa e60969
  Changed leading blanks to tabs in all makefiles.
fukasawa e60969
  Define PNG_USE_PNGVCRD in makefile.w32, to get MMX assembler code.
fukasawa e60969
  Made alternate versions of  png_set_expand() in pngrtran.c, namely
fukasawa e60969
    png_set_gray_1_2_4_to_8, png_set_palette_to_rgb, and png_set_tRNS_to_alpha
fukasawa e60969
    (Greg Roelofs, in "PNG: The Definitive Guide").  Deleted the 1.0.3a aliases.
fukasawa e60969
  Relocated start of 'extern "C"' block in png.h so it doesn't include pngconf.h
fukasawa e60969
  Revised calculation of num_blocks in pngmem.c to avoid a potentially
fukasawa e60969
    negative shift distance, whose results are undefined in the C language.
fukasawa e60969
  Added a check in pngset.c to prevent writing multiple tIME chunks.
fukasawa e60969
  Added a check in pngwrite.c to detect invalid small window_bits sizes.
fukasawa e60969
fukasawa e60969
Version 1.0.3d [September 4, 1999]
fukasawa e60969
  Fixed type casting of igamma in pngrutil.c
fukasawa e60969
  Added new png_expand functions to scripts/pngdef.pas and pngos2.def
fukasawa e60969
  Added a demo read_user_transform_fn that examines the row filters in pngtest.c
fukasawa e60969
fukasawa e60969
Version 1.0.4 [September 24, 1999, not distributed publicly]
fukasawa e60969
  Define PNG_ALWAYS_EXTERN in pngconf.h if __STDC__ is defined
fukasawa e60969
  Delete #define PNG_INTERNAL and include "png.h" from pngasmrd.h
fukasawa e60969
  Made several minor corrections to pngtest.c
fukasawa e60969
  Renamed the makefiles with longer but more user friendly extensions.
fukasawa e60969
  Copied the PNG copyright and license to a separate LICENSE file.
fukasawa e60969
  Revised documentation, png.h, and example.c to remove reference to
fukasawa e60969
    "viewing_gamma" which no longer appears in the PNG specification.
fukasawa e60969
  Revised pngvcrd.c to use MMX code for interlacing only on the final pass.
fukasawa e60969
  Updated pngvcrd.c to use the faster C filter algorithms from libpng-1.0.1a
fukasawa e60969
  Split makefile.win32vc into two versions, makefile.vcawin32 (uses MMX
fukasawa e60969
    assembler code) and makefile.vcwin32 (doesn't).
fukasawa e60969
  Added a CPU timing report to pngtest.c (enabled by defining PNGTEST_TIMING)
fukasawa e60969
  Added a copy of pngnow.png to the distribution.
fukasawa e60969
fukasawa e60969
Version 1.0.4a [September 25, 1999]
fukasawa e60969
  Increase max_pixel_depth in pngrutil.c if a user transform needs it.
fukasawa e60969
  Changed several division operations to right-shifts in pngvcrd.c
fukasawa e60969
fukasawa e60969
Version 1.0.4b [September 30, 1999]
fukasawa e60969
  Added parentheses in line 3732 of pngvcrd.c
fukasawa e60969
  Added a comment in makefile.linux warning about buggy -O3 in pgcc 2.95.1
fukasawa e60969
fukasawa e60969
Version 1.0.4c [October 1, 1999]
fukasawa e60969
  Added a "png_check_version" function in png.c and pngtest.c that will generate
fukasawa e60969
    a helpful compiler error if an old png.h is found in the search path.
fukasawa e60969
  Changed type of png_user_transform_depth|channels from int to png_byte.
fukasawa e60969
  Added "Libpng is OSI Certified Open Source Software" statement to png.h
fukasawa e60969
fukasawa e60969
Version 1.0.4d [October 6, 1999]
fukasawa e60969
  Changed 0.45 to 0.45455 in png_set_sRGB()
fukasawa e60969
  Removed unused PLTE entries from pngnow.png
fukasawa e60969
  Re-enabled some parts of pngvcrd.c (png_combine_row) that work properly.
fukasawa e60969
fukasawa e60969
Version 1.0.4e [October 10, 1999]
fukasawa e60969
  Fixed sign error in pngvcrd.c (Greg Roelofs)
fukasawa e60969
  Replaced some instances of memcpy with simple assignments in pngvcrd (GR-P)
fukasawa e60969
fukasawa e60969
Version 1.0.4f [October 15, 1999]
fukasawa e60969
  Surrounded example.c code with #if 0 .. #endif to prevent people from
fukasawa e60969
    inadvertently trying to compile it.
fukasawa e60969
  Changed png_get_header_version() from a function to a macro in png.h
fukasawa e60969
  Added type casting mostly in pngrtran.c and pngwtran.c
fukasawa e60969
  Removed some pointless "ptr = NULL" in pngmem.c
fukasawa e60969
  Added a "contrib" directory containing the source code from Greg's book.
fukasawa e60969
fukasawa e60969
Version 1.0.5 [October 15, 1999]
fukasawa e60969
  Minor editing of the INSTALL and README files.
fukasawa e60969
fukasawa e60969
Version 1.0.5a [October 23, 1999]
fukasawa e60969
  Added contrib/pngsuite and contrib/pngminus (Willem van Schaik)
fukasawa e60969
  Fixed a typo in the png_set_sRGB() function call in example.c (Jan Nijtmans)
fukasawa e60969
  Further optimization and bugfix of pngvcrd.c
fukasawa e60969
  Revised pngset.c so that it does not allocate or free memory in the user's
fukasawa e60969
    text_ptr structure.  Instead, it makes its own copy.
fukasawa e60969
  Created separate write_end_info_struct in pngtest.c for a more severe test.
fukasawa e60969
  Added code in pngwrite.c to free info_ptr->text[i].key to stop a memory leak.
fukasawa e60969
fukasawa e60969
Version 1.0.5b [November 23, 1999]
fukasawa e60969
  Moved PNG_FLAG_HAVE_CHUNK_HEADER, PNG_FLAG_BACKGROUND_IS_GRAY and
fukasawa e60969
    PNG_FLAG_WROTE_tIME from flags to mode.
fukasawa e60969
  Added png_write_info_before_PLTE() function.
fukasawa e60969
  Fixed some typecasting in contrib/gregbook/*.c
fukasawa e60969
  Updated scripts/makevms.com and added makevms.com to contrib/gregbook
fukasawa e60969
    and contrib/pngminus (Martin Zinser)
fukasawa e60969
fukasawa e60969
Version 1.0.5c [November 26, 1999]
fukasawa e60969
  Moved png_get_header_version from png.h to png.c, to accommodate ansi2knr.
fukasawa e60969
  Removed all global arrays (according to PNG_NO_GLOBAL_ARRAYS macro), to
fukasawa e60969
    accommodate making DLL's: Moved usr_png_ver from global variable to function
fukasawa e60969
    png_get_header_ver() in png.c.  Moved png_sig to png_sig_bytes in png.c and
fukasawa e60969
    eliminated use of png_sig in pngwutil.c.  Moved the various png_CHNK arrays
fukasawa e60969
    into pngtypes.h.  Eliminated use of global png_pass arrays.  Declared the
fukasawa e60969
    png_CHNK and png_pass arrays to be "const".  Made the global arrays
fukasawa e60969
    available to applications (although none are used in libpng itself) when
fukasawa e60969
    PNG_NO_GLOBAL_ARRAYS is not defined or when PNG_GLOBAL_ARRAYS is defined.
fukasawa e60969
  Removed some extraneous "-I" from contrib/pngminus/makefile.std
fukasawa e60969
  Changed the PNG_sRGB_INTENT macros in png.h to be consistent with PNG-1.2.
fukasawa e60969
  Change PNG_SRGB_INTENT to PNG_sRGB_INTENT in libpng.txt and libpng.3
fukasawa e60969
fukasawa e60969
Version 1.0.5d [November 29, 1999]
fukasawa e60969
  Add type cast (png_const_charp) two places in png.c
fukasawa e60969
  Eliminated pngtypes.h; use macros instead to declare PNG_CHNK arrays.
fukasawa e60969
  Renamed "PNG_GLOBAL_ARRAYS" to "PNG_USE_GLOBAL_ARRAYS" and made available
fukasawa e60969
    to applications a macro "PNG_USE_LOCAL_ARRAYS".
fukasawa e60969
  comment out (with #ifdef) all the new declarations when
fukasawa e60969
    PNG_USE_GLOBAL_ARRAYS is defined.
fukasawa e60969
  Added PNG_EXPORT_VAR macro to accommodate making DLL's.
fukasawa e60969
fukasawa e60969
Version 1.0.5e [November 30, 1999]
fukasawa e60969
  Added iCCP, iTXt, and sPLT support; added "lang" member to the png_text
fukasawa e60969
    structure; refactored the inflate/deflate support to make adding new chunks
fukasawa e60969
    with trailing compressed parts easier in the future, and added new functions
fukasawa e60969
    png_free_iCCP, png_free_pCAL, png_free_sPLT, png_free_text, png_get_iCCP,
fukasawa e60969
    png_get_spalettes, png_set_iCCP, png_set_spalettes (Eric S. Raymond).
fukasawa e60969
  NOTE: Applications that write text chunks MUST define png_text->lang
fukasawa e60969
    before calling png_set_text(). It must be set to NULL if you want to
fukasawa e60969
    write tEXt or zTXt chunks.  If you want your application to be able to
fukasawa e60969
    run with older versions of libpng, use
fukasawa e60969
fukasawa e60969
      #ifdef PNG_iTXt_SUPPORTED
fukasawa e60969
         png_text[i].lang = NULL;
fukasawa e60969
      #endif
fukasawa e60969
fukasawa e60969
  Changed png_get_oFFs() and png_set_oFFs() to use signed rather than unsigned
fukasawa e60969
    offsets (Eric S. Raymond).
fukasawa e60969
  Combined PNG_READ_cHNK_SUPPORTED and PNG_WRITE_cHNK_SUPPORTED macros into
fukasawa e60969
    PNG_cHNK_SUPPORTED and combined the three types of PNG_text_SUPPORTED
fukasawa e60969
    macros, leaving the separate macros also available.
fukasawa e60969
  Removed comments on #endifs at the end of many short, non-nested #if-blocks.
fukasawa e60969
fukasawa e60969
Version 1.0.5f [December 6, 1999]
fukasawa e60969
  Changed makefile.solaris to issue a warning about potential problems when
fukasawa e60969
    the ucb "ld" is in the path ahead of the ccs "ld".
fukasawa e60969
  Removed "- [date]" from the "synopsis" line in libpng.3 and libpngpf.3.
fukasawa e60969
  Added sCAL chunk support (Eric S. Raymond).
fukasawa e60969
fukasawa e60969
Version 1.0.5g [December 7, 1999]
fukasawa e60969
  Fixed "png_free_spallettes" typo in png.h
fukasawa e60969
  Added code to handle new chunks in pngpread.c
fukasawa e60969
  Moved PNG_CHNK string macro definitions outside of PNG_NO_EXTERN block
fukasawa e60969
  Added "translated_key" to png_text structure and png_write_iTXt().
fukasawa e60969
  Added code in pngwrite.c to work around a newly discovered zlib bug.
fukasawa e60969
fukasawa e60969
Version 1.0.5h [December 10, 1999]
fukasawa e60969
  NOTE: regarding the note for version 1.0.5e, the following must also
fukasawa e60969
    be included in your code:
fukasawa e60969
        png_text[i].translated_key = NULL;
fukasawa e60969
  Unknown chunk handling is now supported.
fukasawa e60969
  Option to eliminate all floating point support was added.  Some new
fukasawa e60969
    fixed-point functions such as png_set_gAMA_fixed() were added.
fukasawa e60969
  Expanded tabs and removed trailing blanks in source files.
fukasawa e60969
fukasawa e60969
Version 1.0.5i [December 13, 1999]
fukasawa e60969
  Added some type casts to silence compiler warnings.
fukasawa e60969
  Renamed "png_free_spalette" to "png_free_spalettes" for consistency.
fukasawa e60969
  Removed leading blanks from a #define in pngvcrd.c
fukasawa e60969
  Added some parameters to the new png_set_keep_unknown_chunks() function.
fukasawa e60969
  Added a test for up->location != 0 in the first instance of writing
fukasawa e60969
    unknown chunks in pngwrite.c
fukasawa e60969
  Changed "num" to "i" in png_free_spalettes() and png_free_unknowns() to
fukasawa e60969
    prevent recursion.
fukasawa e60969
  Added png_free_hIST() function.
fukasawa e60969
  Various patches to fix bugs in the sCAL and integer cHRM processing,
fukasawa e60969
    and to add some convenience macros for use with sCAL.
fukasawa e60969
fukasawa e60969
Version 1.0.5j [December 21, 1999]
fukasawa e60969
  Changed "unit" parameter of png_write_sCAL from png_byte to int, to work
fukasawa e60969
    around buggy compilers.
fukasawa e60969
  Added new type "png_fixed_point" for integers that hold float*100000 values
fukasawa e60969
  Restored backward compatibility of tEXt/zTXt chunk processing:
fukasawa e60969
    Restored the first four members of png_text to the same order as v.1.0.5d.
fukasawa e60969
    Added members "lang_key" and "itxt_length" to png_text struct.  Set
fukasawa e60969
    text_length=0 when "text" contains iTXt data.  Use the "compression"
fukasawa e60969
    member to distinguish among tEXt/zTXt/iTXt types.  Added
fukasawa e60969
    PNG_ITXT_COMPRESSION_NONE (1) and PNG_ITXT_COMPRESSION_zTXt(2) macros.
fukasawa e60969
    The "Note" above, about backward incompatibility of libpng-1.0.5e, no
fukasawa e60969
    longer applies.
fukasawa e60969
  Fixed png_read|write_iTXt() to read|write parameters in the right order,
fukasawa e60969
    and to write the iTXt chunk after IDAT if it appears in the end_ptr.
fukasawa e60969
  Added pnggccrd.c, version of pngvcrd.c Intel assembler for gcc (Greg Roelofs)
fukasawa e60969
  Reversed the order of trying to write floating-point and fixed-point gAMA.
fukasawa e60969
fukasawa e60969
Version 1.0.5k [December 27, 1999]
fukasawa e60969
  Added many parentheses, e.g., "if (a && b & c)" becomes "if (a && (b & c))"
fukasawa e60969
  Added png_handle_as_unknown() function (Glenn)
fukasawa e60969
  Added png_free_chunk_list() function and chunk_list and num_chunk_list members
fukasawa e60969
    of png_ptr.
fukasawa e60969
  Eliminated erroneous warnings about multiple sPLT chunks and sPLT-after-PLTE.
fukasawa e60969
  Fixed a libpng-1.0.5h bug in pngrutil.c that was issuing erroneous warnings
fukasawa e60969
    about ignoring incorrect gAMA with sRGB (gAMA was in fact not ignored)
fukasawa e60969
  Added png_free_tRNS(); png_set_tRNS() now malloc's its own trans array (ESR).
fukasawa e60969
  Define png_get_int_32 when oFFs chunk is supported as well as when pCAL is.
fukasawa e60969
  Changed type of proflen from png_int_32 to png_uint_32 in png_get_iCCP().
fukasawa e60969
fukasawa e60969
Version 1.0.5l [January 1, 2000]
fukasawa e60969
  Added functions png_set_read_user_chunk_fn() and png_get_user_chunk_ptr()
fukasawa e60969
    for setting a callback function to handle unknown chunks and for
fukasawa e60969
    retrieving the associated user pointer (Glenn).
fukasawa e60969
fukasawa e60969
Version 1.0.5m [January 7, 2000]
fukasawa e60969
  Added high-level functions png_read_png(), png_write_png(), png_free_pixels().
fukasawa e60969
fukasawa e60969
Version 1.0.5n [January 9, 2000]
fukasawa e60969
  Added png_free_PLTE() function, and modified png_set_PLTE() to malloc its
fukasawa e60969
    own memory for info_ptr->palette.  This makes it safe for the calling
fukasawa e60969
    application to free its copy of the palette any time after it calls
fukasawa e60969
    png_set_PLTE().
fukasawa e60969
fukasawa e60969
Version 1.0.5o [January 20, 2000]
fukasawa e60969
  Cosmetic changes only (removed some trailing blanks and TABs)
fukasawa e60969
fukasawa e60969
Version 1.0.5p [January 31, 2000]
fukasawa e60969
  Renamed pngdll.mak to makefile.bd32
fukasawa e60969
  Cosmetic changes in pngtest.c
fukasawa e60969
fukasawa e60969
Version 1.0.5q [February 5, 2000]
fukasawa e60969
  Relocated the makefile.solaris warning about PATH problems.
fukasawa e60969
  Fixed pngvcrd.c bug by pushing/popping registers in mmxsupport (Bruce Oberg)
fukasawa e60969
  Revised makefile.gcmmx
fukasawa e60969
  Added PNG_SETJMP_SUPPORTED, PNG_SETJMP_NOT_SUPPORTED, and PNG_ABORT() macros
fukasawa e60969
fukasawa e60969
Version 1.0.5r [February 7, 2000]
fukasawa e60969
  Removed superfluous prototype for png_get_itxt from png.h
fukasawa e60969
  Fixed a bug in pngrtran.c that improperly expanded the background color.
fukasawa e60969
  Return *num_text=0 from png_get_text() when appropriate, and fix documentation
fukasawa e60969
    of png_get_text() in libpng.txt/libpng.3.
fukasawa e60969
fukasawa e60969
Version 1.0.5s [February 18, 2000]
fukasawa e60969
  Added "png_jmp_env()" macro to pngconf.h, to help people migrate to the
fukasawa e60969
    new error handler that's planned for the next libpng release, and changed
fukasawa e60969
    example.c, pngtest.c, and contrib programs to use this macro.
fukasawa e60969
  Revised some of the DLL-export macros in pngconf.h (Greg Roelofs)
fukasawa e60969
  Fixed a bug in png_read_png() that caused it to fail to expand some images
fukasawa e60969
    that it should have expanded.
fukasawa e60969
  Fixed some mistakes in the unused and undocumented INCH_CONVERSIONS functions
fukasawa e60969
    in pngget.c
fukasawa e60969
  Changed the allocation of palette, history, and trans arrays back to
fukasawa e60969
    the version 1.0.5 method (linking instead of copying) which restores
fukasawa e60969
    backward compatibility with version 1.0.5.  Added some remarks about
fukasawa e60969
    that in example.c.  Added "free_me" member to info_ptr and png_ptr
fukasawa e60969
    and added png_free_data() function.
fukasawa e60969
  Updated makefile.linux and makefile.gccmmx to make directories conditionally.
fukasawa e60969
  Made cosmetic changes to pngasmrd.h
fukasawa e60969
  Added png_set_rows() and png_get_rows(), for use with png_read|write_png().
fukasawa e60969
  Modified png_read_png() to allocate info_ptr->row_pointers only if it
fukasawa e60969
    hasn't already been allocated.
fukasawa e60969
fukasawa e60969
Version 1.0.5t [March 4, 2000]
fukasawa e60969
  Changed png_jmp_env() migration aiding macro to png_jmpbuf().
fukasawa e60969
  Fixed "interlace" typo (should be "interlaced") in contrib/gregbook/read2-x.c
fukasawa e60969
  Fixed bug with use of PNG_BEFORE_IHDR bit in png_ptr->mode, introduced when
fukasawa e60969
    PNG_FLAG_HAVE_CHUNK_HEADER was moved into png_ptr->mode in version 1.0.5b
fukasawa e60969
  Files in contrib/gregbook were revised to use png_jmpbuf() and to select
fukasawa e60969
    a 24-bit visual if one is available, and to allow abbreviated options.
fukasawa e60969
  Files in contrib/pngminus were revised to use the png_jmpbuf() macro.
fukasawa e60969
  Removed spaces in makefile.linux and makefile.gcmmx, introduced in 1.0.5s
fukasawa e60969
fukasawa e60969
Version 1.0.5u [March 5, 2000]
fukasawa e60969
  Simplified the code that detects old png.h in png.c and pngtest.c
fukasawa e60969
  Renamed png_spalette (_p, _pp) to png_sPLT_t (_tp, _tpp)
fukasawa e60969
  Increased precision of rgb_to_gray calculations from 8 to 15 bits and
fukasawa e60969
    added png_set_rgb_to_gray_fixed() function.
fukasawa e60969
  Added makefile.bc32 (32-bit Borland C++, C mode)
fukasawa e60969
fukasawa e60969
Version 1.0.5v [March 11, 2000]
fukasawa e60969
  Added some parentheses to the png_jmpbuf macro definition.
fukasawa e60969
  Updated references to the zlib home page, which has moved to freesoftware.com.
fukasawa e60969
  Corrected bugs in documentation regarding png_read_row() and png_write_row().
fukasawa e60969
  Updated documentation of png_rgb_to_gray calculations in libpng.3/libpng.txt.
fukasawa e60969
  Renamed makefile.borland,turboc3 back to makefile.bor,tc3 as in version 1.0.3,
fukasawa e60969
    revised borland makefiles; added makefile.ibmvac3 and makefile.gcc (Cosmin)
fukasawa e60969
fukasawa e60969
Version 1.0.6 [March 20, 2000]
fukasawa e60969
  Minor revisions of makefile.bor, libpng.txt, and gregbook/rpng2-win.c
fukasawa e60969
  Added makefile.sggcc (SGI IRIX with gcc)
fukasawa e60969
fukasawa e60969
Version 1.0.6d [April 7, 2000]
fukasawa e60969
  Changed sprintf() to strcpy() in png_write_sCAL_s() to work without STDIO
fukasawa e60969
  Added data_length parameter to png_decompress_chunk() function
fukasawa e60969
  Revised documentation to remove reference to abandoned png_free_chnk functions
fukasawa e60969
  Fixed an error in png_rgb_to_gray_fixed()
fukasawa e60969
  Revised example.c, usage of png_destroy_write_struct().
fukasawa e60969
  Renamed makefile.ibmvac3 to makefile.ibmc, added libpng.icc IBM project file
fukasawa e60969
  Added a check for info_ptr->free_me&PNG_FREE_TEXT when freeing text in png.c
fukasawa e60969
  Simplify png_sig_bytes() function to remove use of non-ISO-C strdup().
fukasawa e60969
fukasawa e60969
Version 1.0.6e [April 9, 2000]
fukasawa e60969
  Added png_data_freer() function.
fukasawa e60969
  In the code that checks for over-length tRNS chunks, added check of
fukasawa e60969
    info_ptr->num_trans as well as png_ptr->num_trans (Matthias Benckmann)
fukasawa e60969
  Minor revisions of libpng.txt/libpng.3.
fukasawa e60969
  Check for existing data and free it if the free_me flag is set, in png_set_*()
fukasawa e60969
    and png_handle_*().
fukasawa e60969
  Only define PNG_WEIGHTED_FILTERS_SUPPORTED when PNG_FLOATING_POINT_SUPPORTED
fukasawa e60969
    is defined.
fukasawa e60969
  Changed several instances of PNG_NO_CONSOLE_ID to PNG_NO_STDIO in pngrutil.c
fukasawa e60969
    and mentioned the purposes of the two macros in libpng.txt/libpng.3.
fukasawa e60969
fukasawa e60969
Version 1.0.6f [April 14, 2000]
fukasawa e60969
  Revised png_set_iCCP() and png_set_rows() to avoid prematurely freeing data.
fukasawa e60969
  Add checks in png_set_text() for NULL members of the input text structure.
fukasawa e60969
  Revised libpng.txt/libpng.3.
fukasawa e60969
  Removed superfluous prototype for png_set_iTXt from png.h
fukasawa e60969
  Removed "else" from pngread.c, after png_error(), and changed "0" to "length".
fukasawa e60969
  Changed several png_errors about malformed ancillary chunks to png_warnings.
fukasawa e60969
fukasawa e60969
Version 1.0.6g [April 24, 2000]
fukasawa e60969
  Added png_pass-* arrays to pnggccrd.c when PNG_USE_LOCAL_ARRAYS is defined.
fukasawa e60969
  Relocated paragraph about png_set_background() in libpng.3/libpng.txt
fukasawa e60969
    and other revisions (Matthias Benckmann)
fukasawa e60969
  Relocated info_ptr->free_me, png_ptr->free_me, and other info_ptr and
fukasawa e60969
    png_ptr members to restore binary compatibility with libpng-1.0.5
fukasawa e60969
    (breaks compatibility with libpng-1.0.6).
fukasawa e60969
fukasawa e60969
Version 1.0.6h [April 24, 2000]
fukasawa e60969
  Changed shared library so-number pattern from 2.x.y.z to xy.z (this builds
fukasawa e60969
    libpng.so.10 & libpng.so.10.6h instead of libpng.so.2 & libpng.so.2.1.0.6h)
fukasawa e60969
    This is a temporary change for test purposes.
fukasawa e60969
fukasawa e60969
Version 1.0.6i [May 2, 2000]
fukasawa e60969
  Rearranged some members at the end of png_info and png_struct, to put
fukasawa e60969
    unknown_chunks_num and free_me within the original size of the png_structs
fukasawa e60969
    and free_me, png_read_user_fn, and png_free_fn within the original png_info,
fukasawa e60969
    because some old applications allocate the structs directly instead of
fukasawa e60969
    using png_create_*().
fukasawa e60969
  Added documentation of user memory functions in libpng.txt/libpng.3
fukasawa e60969
  Modified png_read_png so that it will use user_allocated row_pointers
fukasawa e60969
    if present, unless free_me directs that it be freed, and added description
fukasawa e60969
    of the use of png_set_rows() and png_get_rows() in libpng.txt/libpng.3.
fukasawa e60969
  Added PNG_LEGACY_SUPPORTED macro, and #ifdef out all new (since version
fukasawa e60969
    1.00) members of png_struct and png_info, to regain binary compatibility
fukasawa e60969
    when you define this macro.  Capabilities lost in this event
fukasawa e60969
    are user transforms (new in version 1.0.0),the user transform pointer
fukasawa e60969
    (new in version 1.0.2), rgb_to_gray (new in 1.0.5), iCCP, sCAL, sPLT,
fukasawa e60969
    the high-level interface, and unknown chunks support (all new in 1.0.6).
fukasawa e60969
    This was necessary because of old applications that allocate the structs
fukasawa e60969
    directly as authors were instructed to do in libpng-0.88 and earlier,
fukasawa e60969
    instead of using png_create_*().
fukasawa e60969
  Added modes PNG_CREATED_READ_STRUCT and PNG_CREATED_WRITE_STRUCT which
fukasawa e60969
    can be used to detect codes that directly allocate the structs, and
fukasawa e60969
    code to check these modes in png_read_init() and png_write_init() and
fukasawa e60969
    generate a libpng error if the modes aren't set and PNG_LEGACY_SUPPORTED
fukasawa e60969
    was not defined.
fukasawa e60969
  Added makefile.intel and updated makefile.watcom (Pawel Mrochen)
fukasawa e60969
fukasawa e60969
Version 1.0.6j [May 3, 2000]
fukasawa e60969
  Overloaded png_read_init() and png_write_init() with macros that convert
fukasawa e60969
    calls to png_read_init_2() or png_write_init_2() that check the version
fukasawa e60969
    and structure sizes.
fukasawa e60969
fukasawa e60969
Version 1.0.7beta11 [May 7, 2000]
fukasawa e60969
  Removed the new PNG_CREATED_READ_STRUCT and PNG_CREATED_WRITE_STRUCT modes
fukasawa e60969
    which are no longer used.
fukasawa e60969
  Eliminated the three new members of png_text when PNG_LEGACY_SUPPORTED is
fukasawa e60969
    defined or when neither PNG_READ_iTXt_SUPPORTED nor PNG_WRITE_iTXT_SUPPORTED
fukasawa e60969
    is defined.
fukasawa e60969
  Made PNG_NO_READ|WRITE_iTXt the default setting, to avoid memory
fukasawa e60969
    overrun when old applications fill the info_ptr->text structure directly.
fukasawa e60969
  Added PNGAPI macro, and added it to the definitions of all exported functions.
fukasawa e60969
  Relocated version macro definitions ahead of the includes of zlib.h and
fukasawa e60969
    pngconf.h in png.h.
fukasawa e60969
fukasawa e60969
Version 1.0.7beta12 [May 12, 2000]
fukasawa e60969
  Revised pngset.c to avoid a problem with expanding the png_debug macro.
fukasawa e60969
  Deleted some extraneous defines from pngconf.h
fukasawa e60969
  Made PNG_NO_CONSOLE_IO the default condition when PNG_BUILD_DLL is defined.
fukasawa e60969
  Use MSC _RPTn debugging instead of fprintf if _MSC_VER is defined.
fukasawa e60969
  Added png_access_version_number() function.
fukasawa e60969
  Check for mask&PNG_FREE_CHNK (for TEXT, SCAL, PCAL) in png_free_data().
fukasawa e60969
  Expanded libpng.3/libpng.txt information about png_data_freer().
fukasawa e60969
fukasawa e60969
Version 1.0.7beta14 [May 17, 2000] (beta13 was not published)
fukasawa e60969
  Changed pnggccrd.c and pngvcrd.c to handle bad adaptive filter types as
fukasawa e60969
    warnings instead of errors, as pngrutil.c does.
fukasawa e60969
  Set the PNG_INFO_IDAT valid flag in png_set_rows() so png_write_png()
fukasawa e60969
    will actually write IDATs.
fukasawa e60969
  Made the default PNG_USE_LOCAL_ARRAYS depend on PNG_DLL instead of WIN32.
fukasawa e60969
  Make png_free_data() ignore its final parameter except when freeing data
fukasawa e60969
    that can have multiple instances (text, sPLT, unknowns).
fukasawa e60969
  Fixed a new bug in png_set_rows().
fukasawa e60969
  Removed info_ptr->valid tests from png_free_data(), as in version 1.0.5.
fukasawa e60969
  Added png_set_invalid() function.
fukasawa e60969
  Fixed incorrect illustrations of png_destroy_write_struct() in example.c.
fukasawa e60969
fukasawa e60969
Version 1.0.7beta15 [May 30, 2000]
fukasawa e60969
  Revised the deliberately erroneous Linux setjmp code in pngconf.h to produce
fukasawa e60969
    fewer error messages.
fukasawa e60969
  Rearranged checks for Z_OK to check the most likely path first in pngpread.c
fukasawa e60969
    and pngwutil.c.
fukasawa e60969
  Added checks in pngtest.c for png_create_*() returning NULL, and mentioned
fukasawa e60969
    in libpng.txt/libpng.3 the need for applications to check this.
fukasawa e60969
  Changed names of png_default_*() functions in pngtest to pngtest_*().
fukasawa e60969
  Changed return type of png_get_x|y_offset_*() from png_uint_32 to png_int_32.
fukasawa e60969
  Fixed some bugs in the unused PNG_INCH_CONVERSIONS functions in pngget.c
fukasawa e60969
  Set each pointer to NULL after freeing it in png_free_data().
fukasawa e60969
  Worked around a problem in pngconf.h; AIX's strings.h defines an "index"
fukasawa e60969
    macro that conflicts with libpng's png_color_16.index. (Dimitri
fukasawa e60969
    Papadapoulos)
fukasawa e60969
  Added "msvc" directory with MSVC++ project files (Simon-Pierre Cadieux).
fukasawa e60969
fukasawa e60969
Version 1.0.7beta16 [June 4, 2000]
fukasawa e60969
  Revised the workaround of AIX string.h "index" bug.
fukasawa e60969
  Added a check for overlength PLTE chunk in pngrutil.c.
fukasawa e60969
  Added PNG_NO_POINTER_INDEXING macro to use array-indexing instead of pointer
fukasawa e60969
    indexing in pngrutil.c and pngwutil.c to accommodate a buggy compiler.
fukasawa e60969
  Added a warning in png_decompress_chunk() when it runs out of data, e.g.
fukasawa e60969
    when it tries to read an erroneous PhotoShop iCCP chunk.
fukasawa e60969
  Added PNG_USE_DLL macro.
fukasawa e60969
  Revised the copyright/disclaimer/license notice.
fukasawa e60969
  Added contrib/msvctest directory
fukasawa e60969
fukasawa e60969
Version 1.0.7rc1 [June 9, 2000]
fukasawa e60969
  Corrected the definition of PNG_TRANSFORM_INVERT_ALPHA  (0x0400 not 0x0200)
fukasawa e60969
  Added contrib/visupng directory (Willem van Schaik)
fukasawa e60969
fukasawa e60969
Version 1.0.7beta18 [June 23, 2000]
fukasawa e60969
  Revised PNGAPI definition, and pngvcrd.c to work with __GCC__
fukasawa e60969
    and do not redefine PNGAPI if it is passed in via a compiler directive.
fukasawa e60969
  Revised visupng/PngFile.c to remove returns from within the Try block.
fukasawa e60969
  Removed leading underscores from "_PNG_H" and "_PNG_SAVE_BSD_SOURCE" macros.
fukasawa e60969
  Updated contrib/visupng/cexcept.h to version 1.0.0.
fukasawa e60969
  Fixed bugs in pngwrite.c and pngwutil.c that prevented writing iCCP chunks.
fukasawa e60969
fukasawa e60969
Version 1.0.7rc2 [June 28, 2000]
fukasawa e60969
  Updated license to include disclaimers required by UCITA.
fukasawa e60969
  Fixed "DJBPP" typo in pnggccrd.c introduced in beta18.
fukasawa e60969
fukasawa e60969
Version 1.0.7 [July 1, 2000]
fukasawa e60969
  Revised the definition of "trans_values" in libpng.3/libpng.txt
fukasawa e60969
fukasawa e60969
Version 1.0.8beta1 [July 8, 2000]
fukasawa e60969
  Added png_free(png_ptr, key) two places in pngpread.c to stop memory leaks.
fukasawa e60969
  Changed PNG_NO_STDIO to PNG_NO_CONSOLE_IO, several places in pngrutil.c and
fukasawa e60969
    pngwutil.c.
fukasawa e60969
  Changed PNG_EXPORT_VAR to use PNG_IMPEXP, in pngconf.h.
fukasawa e60969
  Removed unused "#include <assert.h>" from png.c
fukasawa e60969
  Added WindowsCE support.
fukasawa e60969
  Revised pnggccrd.c to work with gcc-2.95.2 and in the Cygwin environment.
fukasawa e60969
fukasawa e60969
Version 1.0.8beta2 [July 10, 2000]
fukasawa e60969
  Added project files to the wince directory and made further revisions
fukasawa e60969
    of pngtest.c, pngrio.c, and pngwio.c in support of WindowsCE.
fukasawa e60969
fukasawa e60969
Version 1.0.8beta3 [July 11, 2000]
fukasawa e60969
  Only set the PNG_FLAG_FREE_TRNS or PNG_FREE_TRNS flag in png_handle_tRNS()
fukasawa e60969
    for indexed-color input files to avoid potential double-freeing trans array
fukasawa e60969
    under some unusual conditions; problem was introduced in version 1.0.6f.
fukasawa e60969
  Further revisions to pngtest.c and files in the wince subdirectory.
fukasawa e60969
fukasawa e60969
Version 1.0.8beta4 [July 14, 2000]
fukasawa e60969
  Added the files pngbar.png and pngbar.jpg to the distribution.
fukasawa e60969
  Added makefile.cygwin, and cygwin support in pngconf.h
fukasawa e60969
  Added PNG_NO_ZALLOC_ZERO macro (makes png_zalloc skip zeroing memory)
fukasawa e60969
fukasawa e60969
Version 1.0.8rc1 [July 16, 2000]
fukasawa e60969
  Revised png_debug() macros and statements to eliminate compiler warnings.
fukasawa e60969
fukasawa e60969
Version 1.0.8 [July 24, 2000]
fukasawa e60969
  Added png_flush() in pngwrite.c, after png_write_IEND().
fukasawa e60969
  Updated makefile.hpux to build a shared library.
fukasawa e60969
fukasawa e60969
Version 1.0.9beta1 [November 10, 2000]
fukasawa e60969
  Fixed typo in scripts/makefile.hpux
fukasawa e60969
  Updated makevms.com in scripts and contrib/* and contrib/* (Martin Zinser)
fukasawa e60969
  Fixed seqence-point bug in contrib/pngminus/png2pnm (Martin Zinser)
fukasawa e60969
  Changed "cdrom.com" in documentation to "libpng.org"
fukasawa e60969
  Revised pnggccrd.c to get it all working, and updated makefile.gcmmx (Greg).
fukasawa e60969
  Changed type of "params" from voidp to png_voidp in png_read|write_png().
fukasawa e60969
  Make sure PNGAPI and PNG_IMPEXP are defined in pngconf.h.
fukasawa e60969
  Revised the 3 instances of WRITEFILE in pngtest.c.
fukasawa e60969
  Relocated "msvc" and "wince" project subdirectories into "dll" subdirectory.
fukasawa e60969
  Updated png.rc in dll/msvc project
fukasawa e60969
  Revised makefile.dec to define and use LIBPATH and INCPATH
fukasawa e60969
  Increased size of global png_libpng_ver[] array from 12 to 18 chars.
fukasawa e60969
  Made global png_libpng_ver[], png_sig[] and png_pass_*[] arrays const.
fukasawa e60969
  Removed duplicate png_crc_finish() from png_handle_bKGD() function.
fukasawa e60969
  Added a warning when application calls png_read_update_info() multiple times.
fukasawa e60969
  Revised makefile.cygwin
fukasawa e60969
  Fixed bugs in iCCP support in pngrutil.c and pngwutil.c.
fukasawa e60969
  Replaced png_set_empty_plte_permitted() with png_permit_mng_features().
fukasawa e60969
fukasawa e60969
Version 1.0.9beta2 [November 19, 2000]
fukasawa e60969
  Renamed the "dll" subdirectory "projects".
fukasawa e60969
  Added borland project files to "projects" subdirectory.
fukasawa e60969
  Set VS_FF_PRERELEASE and VS_FF_PATCHED flags in msvc/png.rc when appropriate.
fukasawa e60969
  Add error message in png_set_compression_buffer_size() when malloc fails.
fukasawa e60969
fukasawa e60969
Version 1.0.9beta3 [November 23, 2000]
fukasawa e60969
  Revised PNG_LIBPNG_BUILD_TYPE macro in png.h, used in the msvc project.
fukasawa e60969
  Removed the png_flush() in pngwrite.c that crashes some applications
fukasawa e60969
    that don't set png_output_flush_fn.
fukasawa e60969
  Added makefile.macosx and makefile.aix to scripts directory.
fukasawa e60969
fukasawa e60969
Version 1.0.9beta4 [December 1, 2000]
fukasawa e60969
  Change png_chunk_warning to png_warning in png_check_keyword().
fukasawa e60969
  Increased the first part of msg buffer from 16 to 18 in png_chunk_error().
fukasawa e60969
fukasawa e60969
Version 1.0.9beta5 [December 15, 2000]
fukasawa e60969
  Added support for filter method 64 (for PNG datastreams embedded in MNG).
fukasawa e60969
fukasawa e60969
Version 1.0.9beta6 [December 18, 2000]
fukasawa e60969
  Revised png_set_filter() to accept filter method 64 when appropriate.
fukasawa e60969
  Added new PNG_HAVE_PNG_SIGNATURE bit to png_ptr->mode and use it to
fukasawa e60969
    help prevent applications from using MNG features in PNG datastreams.
fukasawa e60969
  Added png_permit_mng_features() function.
fukasawa e60969
  Revised libpng.3/libpng.txt.  Changed "filter type" to "filter method".
fukasawa e60969
fukasawa e60969
Version 1.0.9rc1 [December 23, 2000]
fukasawa e60969
  Revised test for PNG_HAVE_PNG_SIGNATURE in pngrutil.c
fukasawa e60969
  Fixed error handling of unknown compression type in png_decompress_chunk().
fukasawa e60969
  In pngconf.h, define __cdecl when _MSC_VER is defined.
fukasawa e60969
fukasawa e60969
Version 1.0.9beta7 [December 28, 2000]
fukasawa e60969
  Changed PNG_TEXT_COMPRESSION_zTXt to PNG_COMPRESSION_TYPE_BASE several places.
fukasawa e60969
  Revised memory management in png_set_hIST and png_handle_hIST in a backward
fukasawa e60969
    compatible manner.  PLTE and tRNS were revised similarly.
fukasawa e60969
  Revised the iCCP chunk reader to ignore trailing garbage.
fukasawa e60969
fukasawa e60969
Version 1.0.9beta8 [January 12, 2001]
fukasawa e60969
  Moved pngasmrd.h into pngconf.h.
fukasawa e60969
  Improved handling of out-of-spec garbage iCCP chunks generated by PhotoShop.
fukasawa e60969
fukasawa e60969
Version 1.0.9beta9 [January 15, 2001]
fukasawa e60969
  Added png_set_invalid, png_permit_mng_features, and png_mmx_supported to
fukasawa e60969
    wince and msvc project module definition files.
fukasawa e60969
  Minor revision of makefile.cygwin.
fukasawa e60969
  Fixed bug with progressive reading of narrow interlaced images in pngpread.c
fukasawa e60969
fukasawa e60969
Version 1.0.9beta10 [January 16, 2001]
fukasawa e60969
  Do not typedef png_FILE_p in pngconf.h when PNG_NO_STDIO is defined.
fukasawa e60969
  Fixed "png_mmx_supported" typo in project definition files.
fukasawa e60969
fukasawa e60969
Version 1.0.9beta11 [January 19, 2001]
fukasawa e60969
  Updated makefile.sgi to make shared library.
fukasawa e60969
  Removed png_mmx_support() function and disabled PNG_MNG_FEATURES_SUPPORTED
fukasawa e60969
    by default, for the benefit of DLL forward compatibility.  These will
fukasawa e60969
    be re-enabled in version 1.2.0.
fukasawa e60969
fukasawa e60969
Version 1.0.9rc2 [January 22, 2001]
fukasawa e60969
  Revised cygwin support.
fukasawa e60969
fukasawa e60969
Version 1.0.9 [January 31, 2001]
fukasawa e60969
  Added check of cygwin's ALL_STATIC in pngconf.h
fukasawa e60969
  Added "-nommx" parameter to contrib/gregbook/rpng2-win and rpng2-x demos.
fukasawa e60969
fukasawa e60969
Version 1.0.10beta1 [March 14, 2001]
fukasawa e60969
  Revised makefile.dec, makefile.sgi, and makefile.sggcc; added makefile.hpgcc.
fukasawa e60969
  Reformatted libpng.3 to eliminate bad line breaks.
fukasawa e60969
  Added checks for _mmx_supported in the read_filter_row function of pnggccrd.c
fukasawa e60969
  Added prototype for png_mmx_support() near the top of pnggccrd.c
fukasawa e60969
  Moved some error checking from png_handle_IHDR to png_set_IHDR.
fukasawa e60969
  Added PNG_NO_READ_SUPPORTED and PNG_NO_WRITE_SUPPORTED macros.
fukasawa e60969
  Revised png_mmx_support() function in pnggccrd.c
fukasawa e60969
  Restored version 1.0.8 PNG_WRITE_EMPTY_PLTE_SUPPORTED behavior in pngwutil.c
fukasawa e60969
  Fixed memory leak in contrib/visupng/PngFile.c
fukasawa e60969
  Fixed bugs in png_combine_row() in pnggccrd.c and pngvcrd.c (C version)
fukasawa e60969
  Added warnings when retrieving or setting gamma=0.
fukasawa e60969
  Increased the first part of msg buffer from 16 to 18 in png_chunk_warning().
fukasawa e60969
fukasawa e60969
Version 1.0.10rc1 [March 23, 2001]
fukasawa e60969
  Changed all instances of memcpy, strcpy, and strlen to png_memcpy, png_strcpy,
fukasawa e60969
    and png_strlen.
fukasawa e60969
  Revised png_mmx_supported() function in pnggccrd.c to return proper value.
fukasawa e60969
  Fixed bug in progressive reading (pngpread.c) with small images (height < 8).
fukasawa e60969
fukasawa e60969
Version 1.0.10 [March 30, 2001]
fukasawa e60969
  Deleted extraneous space (introduced in 1.0.9) from line 42 of makefile.cygwin
fukasawa e60969
  Added beos project files (Chris Herborth)
fukasawa e60969
fukasawa e60969
Version 1.0.11beta1 [April 3, 2001]
fukasawa e60969
  Added type casts on several png_malloc() calls (Dimitri Papadapoulos).
fukasawa e60969
  Removed a no-longer needed AIX work-around from pngconf.h
fukasawa e60969
  Changed several "//" single-line comments to C-style in pnggccrd.c
fukasawa e60969
fukasawa e60969
Version 1.0.11beta2 [April 11, 2001]
fukasawa e60969
  Removed PNGAPI from several functions whose prototypes did not have PNGAPI.
fukasawa e60969
  Updated scripts/pngos2.def
fukasawa e60969
fukasawa e60969
Version 1.0.11beta3 [April 14, 2001]
fukasawa e60969
  Added checking the results of many instances of png_malloc() for NULL
fukasawa e60969
fukasawa e60969
Version 1.0.11beta4 [April 20, 2001]
fukasawa e60969
  Undid the changes from version 1.0.11beta3.  Added a check for NULL return
fukasawa e60969
    from user's malloc_fn().
fukasawa e60969
  Removed some useless type casts of the NULL pointer.
fukasawa e60969
  Added makefile.netbsd
fukasawa e60969
fukasawa e60969
Version 1.0.11 [April 27, 2001]
fukasawa e60969
  Revised makefile.netbsd
fukasawa e60969
fukasawa e60969
Version 1.0.12beta1 [May 14, 2001]
fukasawa e60969
  Test for Windows platform in pngconf.h when including malloc.h (Emmanuel Blot)
fukasawa e60969
  Updated makefile.cygwin and handling of Cygwin's ALL_STATIC in pngconf.h
fukasawa e60969
  Added some never-to-be-executed code in pnggccrd.c to quiet compiler warnings.
fukasawa e60969
  Eliminated the png_error about apps using png_read|write_init().  Instead,
fukasawa e60969
    libpng will reallocate the png_struct and info_struct if they are too small.
fukasawa e60969
    This retains future binary compatibility for old applications written for
fukasawa e60969
    libpng-0.88 and earlier.
fukasawa e60969
fukasawa e60969
Version 1.2.0beta1 [May 6, 2001]
fukasawa e60969
  Bumped DLLNUM to 2.
fukasawa e60969
  Re-enabled PNG_MNG_FEATURES_SUPPORTED and enabled PNG_ASSEMBLER_CODE_SUPPORTED
fukasawa e60969
    by default.
fukasawa e60969
  Added runtime selection of MMX features.
fukasawa e60969
  Added png_set_strip_error_numbers function and related macros.
fukasawa e60969
fukasawa e60969
Version 1.2.0beta2 [May 7, 2001]
fukasawa e60969
  Finished merging 1.2.0beta1 with version 1.0.11
fukasawa e60969
  Added a check for attempts to read or write PLTE in grayscale PNG datastreams.
fukasawa e60969
fukasawa e60969
Version 1.2.0beta3 [May 17, 2001]
fukasawa e60969
  Enabled user memory function by default.
fukasawa e60969
  Modified png_create_struct so it passes user mem_ptr to user memory allocator.
fukasawa e60969
  Increased png_mng_features flag from png_byte to png_uint_32.
fukasawa e60969
  Bumped shared-library (so-number) and dll-number to 3.
fukasawa e60969
fukasawa e60969
Version 1.2.0beta4 [June 23, 2001]
fukasawa e60969
  Check for missing profile length field in iCCP chunk and free chunk_data
fukasawa e60969
    in case of truncated iCCP chunk.
fukasawa e60969
  Bumped shared-library number to 3 in makefile.sgi and makefile.sggcc
fukasawa e60969
  Bumped dll-number from 2 to 3 in makefile.cygwin
fukasawa e60969
  Revised contrib/gregbook/rpng*-x.c to avoid a memory leak and to exit cleanly
fukasawa e60969
    if user attempts to run it on an 8-bit display.
fukasawa e60969
  Updated contrib/gregbook
fukasawa e60969
  Use png_malloc instead of png_zalloc to allocate palette in pngset.c
fukasawa e60969
  Updated makefile.ibmc
fukasawa e60969
  Added some typecasts to eliminate gcc 3.0 warnings.  Changed prototypes
fukasawa e60969
    of png_write_oFFS width and height from png_uint_32 to png_int_32.
fukasawa e60969
  Updated example.c
fukasawa e60969
  Revised prototypes for png_debug_malloc and png_debug_free in pngtest.c
fukasawa e60969
fukasawa e60969
Version 1.2.0beta5 [August 8, 2001]
fukasawa e60969
  Revised contrib/gregbook
fukasawa e60969
  Revised makefile.gcmmx
fukasawa e60969
  Revised pnggccrd.c to conditionally compile some thread-unsafe code only
fukasawa e60969
    when PNG_THREAD_UNSAFE_OK is defined.
fukasawa e60969
  Added tests to prevent pngwutil.c from writing a bKGD or tRNS chunk with
fukasawa e60969
    value exceeding 2^bit_depth-1
fukasawa e60969
  Revised makefile.sgi and makefile.sggcc
fukasawa e60969
  Replaced calls to fprintf(stderr,...) with png_warning() in pnggccrd.c
fukasawa e60969
  Removed restriction that do_invert_mono only operate on 1-bit opaque files
fukasawa e60969
fukasawa e60969
Version 1.2.0 [September 1, 2001]
fukasawa e60969
  Changed a png_warning() to png_debug() in pnggccrd.c
fukasawa e60969
  Fixed contrib/gregbook/rpng-x.c, rpng2-x.c to avoid crash with XFreeGC().
fukasawa e60969
fukasawa e60969
Version 1.2.1beta1 [October 19, 2001]
fukasawa e60969
  Revised makefile.std in contrib/pngminus
fukasawa e60969
  Include background_1 in png_struct regardless of gamma support.
fukasawa e60969
  Revised makefile.netbsd and makefile.macosx, added makefile.darwin.
fukasawa e60969
  Revised example.c to provide more details about using row_callback().
fukasawa e60969
fukasawa e60969
Version 1.2.1beta2 [October 25, 2001]
fukasawa e60969
  Added type cast to each NULL appearing in a function call, except for
fukasawa e60969
    WINCE functions.
fukasawa e60969
  Added makefile.so9.
fukasawa e60969
fukasawa e60969
Version 1.2.1beta3 [October 27, 2001]
fukasawa e60969
  Removed type casts from all NULLs.
fukasawa e60969
  Simplified png_create_struct_2().
fukasawa e60969
fukasawa e60969
Version 1.2.1beta4 [November 7, 2001]
fukasawa e60969
  Revised png_create_info_struct() and png_creat_struct_2().
fukasawa e60969
  Added error message if png_write_info() was omitted.
fukasawa e60969
  Type cast NULLs appearing in function calls when _NO_PROTO or
fukasawa e60969
    PNG_TYPECAST_NULL is defined.
fukasawa e60969
fukasawa e60969
Version 1.2.1rc1 [November 24, 2001]
fukasawa e60969
  Type cast NULLs appearing in function calls except when PNG_NO_TYPECAST_NULL
fukasawa e60969
    is defined.
fukasawa e60969
  Changed typecast of "size" argument to png_size_t in pngmem.c calls to
fukasawa e60969
    the user malloc_fn, to agree with the prototype in png.h
fukasawa e60969
  Added a pop/push operation to pnggccrd.c, to preserve Eflag (Maxim Sobolev)
fukasawa e60969
  Updated makefile.sgi to recognize LIBPATH and INCPATH.
fukasawa e60969
  Updated various makefiles so "make clean" does not remove previous major
fukasawa e60969
    version of the shared library.
fukasawa e60969
fukasawa e60969
Version 1.2.1rc2 [December 4, 2001]
fukasawa e60969
  Always allocate 256-entry internal palette, hist, and trans arrays, to
fukasawa e60969
    avoid out-of-bounds memory reference caused by invalid PNG datastreams.
fukasawa e60969
  Added a check for prefix_length > data_length in iCCP chunk handler.
fukasawa e60969
fukasawa e60969
Version 1.2.1 [December 7, 2001]
fukasawa e60969
  None.
fukasawa e60969
fukasawa e60969
Version 1.2.2beta1 [February 22, 2002]
fukasawa e60969
  Fixed a bug with reading the length of iCCP profiles (Larry Reeves).
fukasawa e60969
  Revised makefile.linux, makefile.gcmmx, and makefile.sgi to generate
fukasawa e60969
    libpng.a, libpng12.so (not libpng.so.3), and libpng12/png.h
fukasawa e60969
  Revised makefile.darwin to remove "-undefined suppress" option.
fukasawa e60969
  Added checks for gamma and chromaticity values over 21474.83, which exceed
fukasawa e60969
    the limit for PNG unsigned 32-bit integers when encoded.
fukasawa e60969
  Revised calls to png_create_read_struct() and png_create_write_struct()
fukasawa e60969
    for simpler debugging.
fukasawa e60969
  Revised png_zalloc() so zlib handles errors (uses PNG_FLAG_MALLOC_NULL_MEM_OK)
fukasawa e60969
fukasawa e60969
Version 1.2.2beta2 [February 23, 2002]
fukasawa e60969
  Check chunk_length and idat_size for invalid (over PNG_MAX_UINT) lengths.
fukasawa e60969
  Check for invalid image dimensions in png_get_IHDR.
fukasawa e60969
  Added missing "fi;" in the install target of the SGI makefiles.
fukasawa e60969
  Added install-static to all makefiles that make shared libraries.
fukasawa e60969
  Always do gamma compensation when image is partially transparent.
fukasawa e60969
fukasawa e60969
Version 1.2.2beta3 [March 7, 2002]
fukasawa e60969
  Compute background.gray and background_1.gray even when color_type is RGB
fukasawa e60969
    in case image gets reduced to gray later.
fukasawa e60969
  Modified shared-library makefiles to install pkgconfig/libpngNN.pc.
fukasawa e60969
  Export (with PNGAPI) png_zalloc, png_zfree, and png_handle_as_unknown
fukasawa e60969
  Removed unused png_write_destroy_info prototype from png.h
fukasawa e60969
  Eliminated incorrect use of width_mmx from pnggccrd.c in pixel_bytes == 8 case
fukasawa e60969
  Added install-shared target to all makefiles that make shared libraries.
fukasawa e60969
  Stopped a double free of palette, hist, and trans when not using free_me.
fukasawa e60969
  Added makefile.32sunu for Sun Ultra 32 and makefile.64sunu for Sun Ultra 64.
fukasawa e60969
fukasawa e60969
Version 1.2.2beta4 [March 8, 2002]
fukasawa e60969
  Compute background.gray and background_1.gray even when color_type is RGB
fukasawa e60969
    in case image gets reduced to gray later (Jason Summers).
fukasawa e60969
  Relocated a misplaced /bin/rm in the "install-shared" makefile targets
fukasawa e60969
  Added PNG_1_0_X macro which can be used to build a 1.0.x-compatible library.
fukasawa e60969
fukasawa e60969
Version 1.2.2beta5 [March 26, 2002]
fukasawa e60969
  Added missing PNGAPI to several function definitions.
fukasawa e60969
  Check for invalid bit_depth or color_type in png_get_IHDR(), and
fukasawa e60969
    check for missing PLTE or IHDR in png_push_read_chunk() (Matthias Clasen).
fukasawa e60969
  Revised iTXt support to accept NULL for lang and lang_key.
fukasawa e60969
  Compute gamma for color components of background even when color_type is gray.
fukasawa e60969
  Changed "()" to "{}" in scripts/libpng.pc.in.
fukasawa e60969
  Revised makefiles to put png.h and pngconf.h only in $prefix/include/libpngNN
fukasawa e60969
  Revised makefiles to make symlink to libpng.so.NN in addition to libpngNN.so
fukasawa e60969
fukasawa e60969
Version 1.2.2beta6 [March 31, 2002]
fukasawa e60969
fukasawa e60969
Version 1.0.13beta1 [March 31, 2002]
fukasawa e60969
  Prevent png_zalloc() from trying to memset memory that it failed to acquire.
fukasawa e60969
  Add typecasts of PNG_MAX_UINT in pngset_cHRM_fixed() (Matt Holgate).
fukasawa e60969
  Ensure that the right function (user or default) is used to free the
fukasawa e60969
    png_struct after an error in png_create_read_struct_2().
fukasawa e60969
fukasawa e60969
Version 1.2.2rc1 [April 7, 2002]
fukasawa e60969
fukasawa e60969
Version 1.0.13rc1 [April 7, 2002]
fukasawa e60969
  Save the ebx register in pnggccrd.c (Sami Farin)
fukasawa e60969
  Add "mem_ptr = png_ptr->mem_ptr" in png_destroy_write_struct() (Paul Gardner).
fukasawa e60969
  Updated makefiles to put headers in include/libpng and remove old include/*.h.
fukasawa e60969
fukasawa e60969
Version 1.2.2 [April 15, 2002]
fukasawa e60969
fukasawa e60969
Version 1.0.13 [April 15, 2002]
fukasawa e60969
  Revised description of png_set_filter() in libpng.3/libpng.txt.
fukasawa e60969
  Revised makefile.netbsd and added makefile.neNNbsd and makefile.freebsd
fukasawa e60969
fukasawa e60969
Version 1.0.13patch01 [April 17, 2002]
fukasawa e60969
fukasawa e60969
Version 1.2.2patch01 [April 17, 2002]
fukasawa e60969
  Changed ${PNGMAJ}.${PNGVER} bug to ${PNGVER} in makefile.sgi and
fukasawa e60969
    makefile.sggcc
fukasawa e60969
  Fixed VER -> PNGVER typo in makefile.macosx and added install-static to
fukasawa e60969
    install
fukasawa e60969
  Added install: target to makefile.32sunu and makefile.64sunu
fukasawa e60969
fukasawa e60969
Version 1.0.13patch03 [April 18, 2002]
fukasawa e60969
fukasawa e60969
Version 1.2.2patch03 [April 18, 2002]
fukasawa e60969
  Revised 15 makefiles to link libpng.a to libpngNN.a and the include libpng
fukasawa e60969
  subdirectory to libpngNN subdirectory without the full pathname.
fukasawa e60969
  Moved generation of libpng.pc from "install" to "all" in 15 makefiles.
fukasawa e60969
fukasawa e60969
Version 1.2.3rc1 [April 28, 2002]
fukasawa e60969
  Added install-man target to 15 makefiles (Dimitri Papadopolous-Orfanos).
fukasawa e60969
  Added $(DESTDIR) feature to 24 makefiles (Tim Mooney)
fukasawa e60969
  Fixed bug with $prefix, should be $(prefix) in makefile.hpux.
fukasawa e60969
  Updated cygwin-specific portion of pngconf.h and revised makefile.cygwin
fukasawa e60969
  Added a link from libpngNN.pc to libpng.pc in 15 makefiles.
fukasawa e60969
  Added links from include/libpngNN/*.h to include/*.h in 24 makefiles.
fukasawa e60969
  Revised makefile.darwin to make relative links without full pathname.
fukasawa e60969
  Added setjmp() at the end of png_create_*_struct_2() in case user forgets
fukasawa e60969
    to put one in their application.
fukasawa e60969
  Restored png_zalloc() and png_zfree() prototypes to version 1.2.1 and
fukasawa e60969
    removed them from module definition files.
fukasawa e60969
fukasawa e60969
Version 1.2.3rc2 [May 1, 2002]
fukasawa e60969
  Fixed bug in reporting number of channels in pngget.c and pngset.c,
fukasawa e60969
    that was introduced in version 1.2.2beta5.
fukasawa e60969
  Exported png_zalloc(), png_zfree(), png_default_read(), png_default_write(),
fukasawa e60969
    png_default_flush(), and png_push_fill_buffer() and included them in
fukasawa e60969
    module definition files.
fukasawa e60969
  Added "libpng.pc" dependency to the "install-shared" target in 15 makefiles.
fukasawa e60969
fukasawa e60969
Version 1.2.3rc3 [May 1, 2002]
fukasawa e60969
  Revised prototype for png_default_flush()
fukasawa e60969
  Remove old libpng.pc and libpngNN.pc before installing new ones.
fukasawa e60969
fukasawa e60969
Version 1.2.3rc4 [May 2, 2002]
fukasawa e60969
  Typos in *.def files (png_default_read|write -> png_default_read|write_data)
fukasawa e60969
  In makefiles, changed rm libpng.NN.pc to rm libpngNN.pc
fukasawa e60969
  Added libpng-config and libpngNN-config and modified makefiles to install
fukasawa e60969
    them.
fukasawa e60969
  Changed $(MANPATH) to $(DESTDIR)$(MANPATH) in makefiles
fukasawa e60969
  Added "Win32 DLL VB" configuration to projects/msvc/libpng.dsp
fukasawa e60969
fukasawa e60969
Version 1.2.3rc5 [May 11, 2002]
fukasawa e60969
  Changed "error" and "message" in prototypes to "error_message" and
fukasawa e60969
    "warning_message" to avoid namespace conflict.
fukasawa e60969
  Revised 15 makefiles to build libpng-config from libpng-config-*.in
fukasawa e60969
  Once more restored png_zalloc and png_zfree to regular nonexported form.
fukasawa e60969
  Restored png_default_read|write_data, png_default_flush, png_read_fill_buffer
fukasawa e60969
    to nonexported form, but with PNGAPI, and removed them from module def
fukasawa e60969
    files.
fukasawa e60969
fukasawa e60969
Version 1.2.3rc6 [May 14, 2002]
fukasawa e60969
  Removed "PNGAPI" from png_zalloc() and png_zfree() in png.c
fukasawa e60969
  Changed "Gz" to "Gd" in projects/msvc/libpng.dsp and zlib.dsp.
fukasawa e60969
  Removed leftover libpng-config "sed" script from four makefiles.
fukasawa e60969
  Revised libpng-config creating script in 16 makefiles.
fukasawa e60969
fukasawa e60969
Version 1.2.3 [May 22, 2002]
fukasawa e60969
  Revised libpng-config target in makefile.cygwin.
fukasawa e60969
  Removed description of png_set_mem_fn() from documentation.
fukasawa e60969
  Revised makefile.freebsd.
fukasawa e60969
  Minor cosmetic changes to 15 makefiles, e.g., $(DI) = $(DESTDIR)/$(INCDIR).
fukasawa e60969
  Revised projects/msvc/README.txt
fukasawa e60969
  Changed -lpng to -lpngNN in LDFLAGS in several makefiles.
fukasawa e60969
fukasawa e60969
Version 1.2.4beta1 [May 24, 2002]
fukasawa e60969
  Added libpng.pc and libpng-config to "all:" target in 16 makefiles.
fukasawa e60969
  Fixed bug in 16 makefiles: $(DESTDIR)/$(LIBPATH) to $(DESTDIR)$(LIBPATH)
fukasawa e60969
  Added missing "\" before closing double quote in makefile.gcmmx.
fukasawa e60969
  Plugged various memory leaks; added png_malloc_warn() and png_set_text_2()
fukasawa e60969
    functions.
fukasawa e60969
fukasawa e60969
Version 1.2.4beta2 [June 25, 2002]
fukasawa e60969
  Plugged memory leak of png_ptr->current_text (Matt Holgate).
fukasawa e60969
  Check for buffer overflow before reading CRC in pngpread.c (Warwick Allison)
fukasawa e60969
  Added -soname to the loader flags in makefile.dec, makefile.sgi, and
fukasawa e60969
    makefile.sggcc.
fukasawa e60969
  Added "test-installed" target to makefile.linux, makefile.gcmmx,
fukasawa e60969
    makefile.sgi, and makefile.sggcc.
fukasawa e60969
fukasawa e60969
Version 1.2.4beta3 [June 28, 2002]
fukasawa e60969
  Plugged memory leak of row_buf in pngtest.c when there is a png_error().
fukasawa e60969
  Detect buffer overflow in pngpread.c when IDAT is corrupted with extra data.
fukasawa e60969
  Added "test-installed" target to makefile.32sunu, makefile.64sunu,
fukasawa e60969
    makefile.beos, makefile.darwin, makefile.dec, makefile.macosx,
fukasawa e60969
    makefile.solaris, makefile.hpux, makefile.hpgcc, and makefile.so9.
fukasawa e60969
fukasawa e60969
Version 1.2.4rc1 and 1.0.14rc1 [July 2, 2002]
fukasawa e60969
  Added "test-installed" target to makefile.cygwin and makefile.sco.
fukasawa e60969
  Revised pnggccrd.c to be able to back out version 1.0.x via PNG_1_0_X macro.
fukasawa e60969
fukasawa e60969
Version 1.2.4 and 1.0.14 [July 8, 2002]
fukasawa e60969
  Changed png_warning() to png_error() when width is too large to process.
fukasawa e60969
fukasawa e60969
Version 1.2.4patch01 [July 20, 2002]
fukasawa e60969
  Revised makefile.cygwin to use DLL number 12 instead of 13.
fukasawa e60969
fukasawa e60969
Version 1.2.5beta1 [August 6, 2002]
fukasawa e60969
  Added code to contrib/gregbook/readpng2.c to ignore unused chunks.
fukasawa e60969
  Replaced toucan.png in contrib/gregbook (it has been corrupt since 1.0.11)
fukasawa e60969
  Removed some stray *.o files from contrib/gregbook.
fukasawa e60969
  Changed png_error() to png_warning() about "Too much data" in pngpread.c
fukasawa e60969
    and about "Extra compressed data" in pngrutil.c.
fukasawa e60969
  Prevent png_ptr->pass from exceeding 7 in png_push_finish_row().
fukasawa e60969
  Updated makefile.hpgcc
fukasawa e60969
  Updated png.c and pnggccrd.c handling of return from png_mmx_support()
fukasawa e60969
fukasawa e60969
Version 1.2.5beta2 [August 15, 2002]
fukasawa e60969
  Only issue png_warning() about "Too much data" in pngpread.c when avail_in
fukasawa e60969
    is nonzero.
fukasawa e60969
  Updated makefiles to install a separate libpng.so.3 with its own rpath.
fukasawa e60969
fukasawa e60969
Version 1.2.5rc1 and 1.0.15rc1 [August 24, 2002]
fukasawa e60969
  Revised makefiles to not remove previous minor versions of shared libraries.
fukasawa e60969
fukasawa e60969
Version 1.2.5rc2 and 1.0.15rc2 [September 16, 2002]
fukasawa e60969
  Revised 13 makefiles to remove "-lz" and "-L$(ZLIBLIB)", etc., from shared
fukasawa e60969
    library loader directive.
fukasawa e60969
  Added missing "$OBJSDLL" line to makefile.gcmmx.
fukasawa e60969
  Added missing "; fi" to makefile.32sunu.
fukasawa e60969
fukasawa e60969
Version 1.2.5rc3 and 1.0.15rc3 [September 18, 2002]
fukasawa e60969
  Revised libpng-config script.
fukasawa e60969
fukasawa e60969
Version 1.2.5 and 1.0.15 [October 3, 2002]
fukasawa e60969
  Revised makefile.macosx, makefile.darwin, makefile.hpgcc, and makefile.hpux,
fukasawa e60969
    and makefile.aix.
fukasawa e60969
  Relocated two misplaced PNGAPI lines in pngtest.c
fukasawa e60969
fukasawa e60969
Version 1.2.6beta1 [October 22, 2002]
fukasawa e60969
  Commented out warning about uninitialized mmx_support in pnggccrd.c.
fukasawa e60969
  Changed "IBMCPP__" flag to "__IBMCPP__" in pngconf.h.
fukasawa e60969
  Relocated two more misplaced PNGAPI lines in pngtest.c
fukasawa e60969
  Fixed memory overrun bug in png_do_read_filler() with 16-bit datastreams,
fukasawa e60969
    introduced in version 1.0.2.
fukasawa e60969
  Revised makefile.macosx, makefile.dec, makefile.aix, and makefile.32sunu.
fukasawa e60969
fukasawa e60969
Version 1.2.6beta2 [November 1, 2002]
fukasawa e60969
  Added libpng-config "--ldopts" output.
fukasawa e60969
  Added "AR=ar" and "ARFLAGS=rc" and changed "ar rc" to "$(AR) $(ARFLAGS)"
fukasawa e60969
    in makefiles.
fukasawa e60969
fukasawa e60969
Version 1.2.6beta3 [July 18, 2004]
fukasawa e60969
  Reverted makefile changes from version 1.2.6beta2 and some of the changes
fukasawa e60969
    from version 1.2.6beta1; these will be postponed until version 1.2.7.
fukasawa e60969
    Version 1.2.6 is going to be a simple bugfix release.
fukasawa e60969
  Changed the one instance of "ln -sf" to "ln -f -s" in each Sun makefile.
fukasawa e60969
  Fixed potential overrun in pngerror.c by using strncpy instead of memcpy.
fukasawa e60969
  Added "#!/bin/sh" at the top of configure, for recognition of the
fukasawa e60969
    'x' flag under Cygwin (Cosmin).
fukasawa e60969
  Optimized vacuous tests that silence compiler warnings, in png.c (Cosmin).
fukasawa e60969
  Added support for PNG_USER_CONFIG, in pngconf.h (Cosmin).
fukasawa e60969
  Fixed the special memory handler for Borland C under DOS, in pngmem.c
fukasawa e60969
    (Cosmin).
fukasawa e60969
  Removed some spurious assignments in pngrutil.c (Cosmin).
fukasawa e60969
  Replaced 65536 with 65536L, and 0xffff with 0xffffL, to silence warnings
fukasawa e60969
    on 16-bit platforms (Cosmin).
fukasawa e60969
  Enclosed shift op expressions in parentheses, to silence warnings (Cosmin).
fukasawa e60969
  Used proper type png_fixed_point, to avoid problems on 16-bit platforms,
fukasawa e60969
    in png_handle_sRGB() (Cosmin).
fukasawa e60969
  Added compression_type to png_struct, and optimized the window size
fukasawa e60969
    inside the deflate stream (Cosmin).
fukasawa e60969
  Fixed definition of isnonalpha(), in pngerror.c and pngrutil.c (Cosmin).
fukasawa e60969
  Fixed handling of unknown chunks that come after IDAT (Cosmin).
fukasawa e60969
  Allowed png_error() and png_warning() to work even if png_ptr == NULL
fukasawa e60969
    (Cosmin).
fukasawa e60969
  Replaced row_info->rowbytes with row_bytes in png_write_find_filter()
fukasawa e60969
    (Cosmin).
fukasawa e60969
  Fixed definition of PNG_LIBPNG_VER_DLLNUM (Simon-Pierre).
fukasawa e60969
  Used PNG_LIBPNG_VER and PNG_LIBPNG_VER_STRING instead of the hardcoded
fukasawa e60969
    values in png.c (Simon-Pierre, Cosmin).
fukasawa e60969
  Initialized png_libpng_ver[] with PNG_LIBPNG_VER_STRING (Simon-Pierre).
fukasawa e60969
  Replaced PNG_LIBPNG_VER_MAJOR with PNG_LIBPNG_VER_DLLNUM in png.rc
fukasawa e60969
    (Simon-Pierre).
fukasawa e60969
  Moved the definition of PNG_HEADER_VERSION_STRING near the definitions
fukasawa e60969
    of the other PNG_LIBPNG_VER_... symbols in png.h (Cosmin).
fukasawa e60969
  Relocated #ifndef PNGAPI guards in pngconf.h (Simon-Pierre, Cosmin).
fukasawa e60969
  Updated scripts/makefile.vc(a)win32 (Cosmin).
fukasawa e60969
  Updated the MSVC project (Simon-Pierre, Cosmin).
fukasawa e60969
  Updated the Borland C++ Builder project (Cosmin).
fukasawa e60969
  Avoided access to asm_flags in pngvcrd.c, if PNG_1_0_X is defined (Cosmin).
fukasawa e60969
  Commented out warning about uninitialized mmx_support in pngvcrd.c (Cosmin).
fukasawa e60969
  Removed scripts/makefile.bd32 and scripts/pngdef.pas (Cosmin).
fukasawa e60969
  Added extra guard around inclusion of Turbo C memory headers, in pngconf.h
fukasawa e60969
    (Cosmin).
fukasawa e60969
  Renamed projects/msvc/ to projects/visualc6/, and projects/borland/ to
fukasawa e60969
    projects/cbuilder5/ (Cosmin).
fukasawa e60969
  Moved projects/visualc6/png32ms.def to scripts/pngw32.def,
fukasawa e60969
    and projects/visualc6/png.rc to scripts/pngw32.rc (Cosmin).
fukasawa e60969
  Added projects/visualc6/pngtest.dsp; removed contrib/msvctest/ (Cosmin).
fukasawa e60969
  Changed line endings to DOS style in cbuilder5 and visualc6 files, even
fukasawa e60969
    in the tar.* distributions (Cosmin).
fukasawa e60969
  Updated contrib/visupng/VisualPng.dsp (Cosmin).
fukasawa e60969
  Updated contrib/visupng/cexcept.h to version 2.0.0 (Cosmin).
fukasawa e60969
  Added a separate distribution with "configure" and supporting files (Junichi).
fukasawa e60969
fukasawa e60969
Version 1.2.6beta4 [July 28, 2004]
fukasawa e60969
  Added user ability to change png_size_t via a PNG_SIZE_T macro.
fukasawa e60969
  Added png_sizeof() and png_convert_size() functions.
fukasawa e60969
  Added PNG_SIZE_MAX (maximum value of a png_size_t variable.
fukasawa e60969
  Added check in png_malloc_default() for (size_t)size != (png_uint_32)size
fukasawa e60969
    which would indicate an overflow.
fukasawa e60969
  Changed sPLT failure action from png_error to png_warning and abandon chunk.
fukasawa e60969
  Changed sCAL and iCCP failures from png_error to png_warning and abandon.
fukasawa e60969
  Added png_get_uint_31(png_ptr, buf) function.
fukasawa e60969
  Added PNG_UINT_32_MAX macro.
fukasawa e60969
  Renamed PNG_MAX_UINT to PNG_UINT_31_MAX.
fukasawa e60969
  Made png_zalloc() issue a png_warning and return NULL on potential
fukasawa e60969
    overflow.
fukasawa e60969
  Turn on PNG_NO_ZALLOC_ZERO by default in version 1.2.x
fukasawa e60969
  Revised "clobber list" in pnggccrd.c so it will compile under gcc-3.4.
fukasawa e60969
  Revised Borland portion of png_malloc() to return NULL or issue
fukasawa e60969
    png_error() according to setting of PNG_FLAG_MALLOC_NULL_MEM_OK.
fukasawa e60969
  Added PNG_NO_SEQUENTIAL_READ_SUPPORTED macro to conditionally remove
fukasawa e60969
    sequential read support.
fukasawa e60969
  Added some "#if PNG_WRITE_SUPPORTED" blocks.
fukasawa e60969
  Added #ifdef to remove some redundancy in png_malloc_default().
fukasawa e60969
  Use png_malloc instead of png_zalloc to allocate the pallete.
fukasawa e60969
fukasawa e60969
Version 1.0.16rc1 and 1.2.6rc1 [August 4, 2004]
fukasawa e60969
  Fixed buffer overflow vulnerability (CVE-2004-0597) in png_handle_tRNS().
fukasawa e60969
  Fixed NULL dereference vulnerability (CVE-2004-0598) in png_handle_iCCP().
fukasawa e60969
  Fixed integer overflow vulnerability (CVE-2004-0599) in png_read_png().
fukasawa e60969
  Fixed some harmless bugs in png_handle_sBIT, etc, that would cause
fukasawa e60969
    duplicate chunk types to go undetected.
fukasawa e60969
  Fixed some timestamps in the -config version
fukasawa e60969
  Rearranged order of processing of color types in png_handle_tRNS().
fukasawa e60969
  Added ROWBYTES macro to calculate rowbytes without integer overflow.
fukasawa e60969
  Updated makefile.darwin and removed makefile.macosx from scripts directory.
fukasawa e60969
  Imposed default one million column, one-million row limits on the image
fukasawa e60969
    dimensions, and added png_set_user_limits() function to override them.
fukasawa e60969
  Revised use of PNG_SET_USER_LIMITS_SUPPORTED macro.
fukasawa e60969
  Fixed wrong cast of returns from png_get_user_width|height_max().
fukasawa e60969
  Changed some "keep the compiler happy" from empty statements to returns,
fukasawa e60969
  Revised libpng.txt to remove 1.2.x stuff from the 1.0.x distribution
fukasawa e60969
fukasawa e60969
Version 1.0.16rc2 and 1.2.6rc2 [August 7, 2004]
fukasawa e60969
  Revised makefile.darwin and makefile.solaris.  Removed makefile.macosx.
fukasawa e60969
  Revised pngtest's png_debug_malloc() to use png_malloc() instead of
fukasawa e60969
    png_malloc_default() which is not supposed to be exported.
fukasawa e60969
  Fixed off-by-one error in one of the conversions to PNG_ROWBYTES() in
fukasawa e60969
    pngpread.c.  Bug was introduced in 1.2.6rc1.
fukasawa e60969
  Fixed bug in RGB to RGBX transformation introduced in 1.2.6rc1.
fukasawa e60969
  Fixed old bug in RGB to Gray transformation.
fukasawa e60969
  Fixed problem with 64-bit compilers by casting arguments to abs()
fukasawa e60969
    to png_int_32.
fukasawa e60969
  Changed "ln -sf" to "ln -f -s" in three makefiles (solaris, sco, so9).
fukasawa e60969
  Changed "HANDLE_CHUNK_*" to "PNG_HANDLE_CHUNK_*" (Cosmin)
fukasawa e60969
  Added "-@/bin/rm -f $(DL)/$(LIBNAME).so.$(PNGMAJ)" to 15 *NIX makefiles.
fukasawa e60969
  Added code to update the row_info->colortype in png_do_read_filler() (MSB).
fukasawa e60969
fukasawa e60969
Version 1.0.16rc3 and 1.2.6rc3 [August 9, 2004]
fukasawa e60969
  Eliminated use of "abs()" in testing cHRM and gAMA values, to avoid
fukasawa e60969
    trouble with some 64-bit compilers.  Created PNG_OUT_OF_RANGE() macro.
fukasawa e60969
  Revised documentation of png_set_keep_unknown_chunks().
fukasawa e60969
  Check handle_as_unknown status in pngpread.c, as in pngread.c previously.
fukasawa e60969
  Moved  "PNG_HANDLE_CHUNK_*" macros out of PNG_INTERNAL section of png.h
fukasawa e60969
  Added "rim" definitions for CONST4 and CONST6 in pnggccrd.c
fukasawa e60969
fukasawa e60969
Version 1.0.16rc4 and 1.2.6rc4 [August 10, 2004]
fukasawa e60969
  Fixed mistake in pngtest.c introduced in 1.2.6rc2 (declaration of
fukasawa e60969
    "pinfo" was out of place).
fukasawa e60969
fukasawa e60969
Version 1.0.16rc5 and 1.2.6rc5 [August 10, 2004]
fukasawa e60969
  Moved  "PNG_HANDLE_CHUNK_*" macros out of PNG_ASSEMBLER_CODE_SUPPORTED
fukasawa e60969
    section of png.h where they were inadvertently placed in version rc3.
fukasawa e60969
fukasawa e60969
Version 1.2.6 and 1.0.16 [August 15, 2004]
fukasawa e60969
  Revised pngtest so memory allocation testing is only done when PNG_DEBUG==1.
fukasawa e60969
fukasawa e60969
Version 1.2.7beta1 [August 26, 2004]
fukasawa e60969
  Removed unused pngasmrd.h file.
fukasawa e60969
  Removed references to uu.net for archived files.  Added references to
fukasawa e60969
    PNG Spec (second edition) and the PNG ISO/IEC Standard.
fukasawa e60969
  Added "test-dd" target in 15 makefiles, to run pngtest in DESTDIR.
fukasawa e60969
  Fixed bug with "optimized window size" in the IDAT datastream, that
fukasawa e60969
    causes libpng to write PNG files with incorrect zlib header bytes.
fukasawa e60969
fukasawa e60969
Version 1.2.7beta2 [August 28, 2004]
fukasawa e60969
  Fixed bug with sCAL chunk and big-endian machines (David Munro).
fukasawa e60969
  Undid new code added in 1.2.6rc2 to update the color_type in
fukasawa e60969
    png_set_filler().
fukasawa e60969
  Added png_set_add_alpha() that updates color type.
fukasawa e60969
fukasawa e60969
Version 1.0.17rc1 and 1.2.7rc1 [September 4, 2004]
fukasawa e60969
  Revised png_set_strip_filler() to not remove alpha if color_type has alpha.
fukasawa e60969
fukasawa e60969
Version 1.2.7 and 1.0.17 [September 12, 2004]
fukasawa e60969
  Added makefile.hp64
fukasawa e60969
  Changed projects/msvc/png32ms.def to scripts/png32ms.def in makefile.cygwin
fukasawa e60969
fukasawa e60969
Version 1.2.8beta1 [November 1, 2004]
fukasawa e60969
  Fixed bug in png_text_compress() that would fail to complete a large block.
fukasawa e60969
  Fixed bug, introduced in libpng-1.2.7, that overruns a buffer during
fukasawa e60969
    strip alpha operation in png_do_strip_filler().
fukasawa e60969
  Added PNG_1_2_X definition in pngconf.h
fukasawa e60969
  Use #ifdef to comment out png_info_init in png.c and png_read_init in
fukasawa e60969
    pngread.c (as of 1.3.0)
fukasawa e60969
fukasawa e60969
Version 1.2.8beta2 [November 2, 2004]
fukasawa e60969
  Reduce color_type to a nonalpha type after strip alpha operation in
fukasawa e60969
    png_do_strip_filler().
fukasawa e60969
fukasawa e60969
Version 1.2.8beta3 [November 3, 2004]
fukasawa e60969
  Revised definitions of PNG_MAX_UINT_32, PNG_MAX_SIZE, and PNG_MAXSUM
fukasawa e60969
fukasawa e60969
Version 1.2.8beta4 [November 12, 2004]
fukasawa e60969
  Fixed (again) definition of PNG_LIBPNG_VER_DLLNUM in png.h (Cosmin).
fukasawa e60969
  Added PNG_LIBPNG_BUILD_PRIVATE in png.h (Cosmin).
fukasawa e60969
  Set png_ptr->zstream.data_type to Z_BINARY, to avoid unnecessary detection
fukasawa e60969
    of data type in deflate (Cosmin).
fukasawa e60969
  Deprecated but continue to support SPECIALBUILD and PRIVATEBUILD in favor of
fukasawa e60969
    PNG_LIBPNG_BUILD_SPECIAL_STRING and PNG_LIBPNG_BUILD_PRIVATE_STRING.
fukasawa e60969
fukasawa e60969
Version 1.2.8beta5 [November 20, 2004]
fukasawa e60969
  Use png_ptr->flags instead of png_ptr->transformations to pass
fukasawa e60969
    PNG_STRIP_ALPHA info to png_do_strip_filler(), to preserve ABI
fukasawa e60969
    compatibility.
fukasawa e60969
  Revised handling of SPECIALBUILD, PRIVATEBUILD,
fukasawa e60969
    PNG_LIBPNG_BUILD_SPECIAL_STRING and PNG_LIBPNG_BUILD_PRIVATE_STRING.
fukasawa e60969
fukasawa e60969
Version 1.2.8rc1 [November 24, 2004]
fukasawa e60969
  Moved handling of BUILD macros from pngconf.h to png.h
fukasawa e60969
  Added definition of PNG_LIBPNG_BASE_TYPE in png.h, inadvertently
fukasawa e60969
    omitted from beta5.
fukasawa e60969
  Revised scripts/pngw32.rc
fukasawa e60969
  Despammed mailing addresses by masking "@" with "at".
fukasawa e60969
  Inadvertently installed a supposedly faster test version of pngrutil.c
fukasawa e60969
fukasawa e60969
Version 1.2.8rc2 [November 26, 2004]
fukasawa e60969
  Added two missing "\" in png.h
fukasawa e60969
  Change tests in pngread.c and pngpread.c to
fukasawa e60969
    if (png_ptr->transformations || (png_ptr->flags&PNG_FLAG_STRIP_ALPHA))
fukasawa e60969
       png_do_read_transformations(png_ptr);
fukasawa e60969
fukasawa e60969
Version 1.2.8rc3 [November 28, 2004]
fukasawa e60969
  Reverted pngrutil.c to version libpng-1.2.8beta5.
fukasawa e60969
  Added scripts/makefile.elf with supporting code in pngconf.h for symbol
fukasawa e60969
    versioning (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.2.8rc4 [November 29, 2004]
fukasawa e60969
  Added projects/visualc7 (Simon-pierre).
fukasawa e60969
fukasawa e60969
Version 1.2.8rc5 [November 29, 2004]
fukasawa e60969
  Fixed new typo in scripts/pngw32.rc
fukasawa e60969
fukasawa e60969
Version 1.2.8 [December 3, 2004]
fukasawa e60969
  Removed projects/visualc7, added projects/visualc71.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta1 [February 21, 2006]
fukasawa e60969
  Initialized some structure members in pngwutil.c to avoid gcc-4.0.0 complaints
fukasawa e60969
  Revised man page and libpng.txt to make it clear that one should not call
fukasawa e60969
    png_read_end or png_write_end after png_read_png or png_write_png.
fukasawa e60969
  Updated references to png-mng-implement mailing list.
fukasawa e60969
  Fixed an incorrect typecast in pngrutil.c
fukasawa e60969
  Added PNG_NO_READ_SUPPORTED conditional for making a write-only library.
fukasawa e60969
  Added PNG_NO_WRITE_INTERLACING_SUPPORTED conditional.
fukasawa e60969
  Optimized alpha-inversion loops in pngwtran.c
fukasawa e60969
  Moved test for nonzero gamma outside of png_build_gamma_table() in pngrtran.c
fukasawa e60969
  Make sure num_trans is <= 256 before copying data in png_set_tRNS().
fukasawa e60969
  Make sure num_palette is <= 256 before copying data in png_set_PLTE().
fukasawa e60969
  Interchanged order of write_swap_alpha and write_invert_alpha transforms.
fukasawa e60969
  Added parentheses in the definition of PNG_LIBPNG_BUILD_TYPE (Cosmin).
fukasawa e60969
  Optimized zlib window flag (CINFO) in contrib/pngsuite/*.png (Cosmin).
fukasawa e60969
  Updated scripts/makefile.bc32 for Borland C++ 5.6 (Cosmin).
fukasawa e60969
  Exported png_get_uint_32, png_save_uint_32, png_get_uint_16, png_save_uint_16,
fukasawa e60969
    png_get_int_32, png_save_int_32, png_get_uint_31 (Cosmin).
fukasawa e60969
  Added type cast (png_byte) in png_write_sCAL() (Cosmin).
fukasawa e60969
  Fixed scripts/makefile.cygwin (Christian Biesinger, Cosmin).
fukasawa e60969
  Default iTXt support was inadvertently enabled.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta2 [February 21, 2006]
fukasawa e60969
  Check for png_rgb_to_gray and png_gray_to_rgb read transformations before
fukasawa e60969
    checking for png_read_dither in pngrtran.c
fukasawa e60969
  Revised checking of chromaticity limits to accommodate extended RGB
fukasawa e60969
    colorspace (John Denker).
fukasawa e60969
  Changed line endings in some of the project files to CRLF, even in the
fukasawa e60969
    "Unix" tar distributions (Cosmin).
fukasawa e60969
  Made png_get_int_32 and png_save_int_32 always available (Cosmin).
fukasawa e60969
  Updated scripts/pngos2.def, scripts/pngw32.def and projects/wince/png32ce.def
fukasawa e60969
    with the newly exported functions.
fukasawa e60969
  Eliminated distributions without the "configure" script.
fukasawa e60969
  Updated INSTALL instructions.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta3 [February 24, 2006]
fukasawa e60969
  Fixed CRCRLF line endings in contrib/visupng/VisualPng.dsp
fukasawa e60969
  Made libpng.pc respect EXEC_PREFIX (D. P. Kreil, J. Bowler)
fukasawa e60969
  Removed reference to pngasmrd.h from Makefile.am
fukasawa e60969
  Renamed CHANGES to ChangeLog.
fukasawa e60969
  Renamed LICENSE to COPYING.
fukasawa e60969
  Renamed ANNOUNCE to NEWS.
fukasawa e60969
  Created AUTHORS file.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta4 [March 3, 2006]
fukasawa e60969
  Changed definition of PKGCONFIG from $prefix/lib to $libdir in configure.ac
fukasawa e60969
  Reverted to filenames LICENSE and ANNOUNCE; removed AUTHORS and COPYING.
fukasawa e60969
  Removed newline from the end of some error and warning messages.
fukasawa e60969
  Removed test for sqrt() from configure.ac and configure.
fukasawa e60969
  Made swap tables in pngtrans.c PNG_CONST (Carlo Bramix).
fukasawa e60969
  Disabled default iTXt support that was inadvertently enabled in
fukasawa e60969
    libpng-1.2.9beta1.
fukasawa e60969
  Added "OS2" to list of systems that don't need underscores, in pnggccrd.c
fukasawa e60969
  Removed libpng version and date from *.c files.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta5 [March 4, 2006]
fukasawa e60969
  Removed trailing blanks from source files.
fukasawa e60969
  Put version and date of latest change in each source file, and changed
fukasawa e60969
    copyright year accordingly.
fukasawa e60969
  More cleanup of configure.ac, Makefile.am, and associated scripts.
fukasawa e60969
  Restored scripts/makefile.elf which was inadvertently deleted.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta6 [March 6, 2006]
fukasawa e60969
  Fixed typo (RELEASE) in configuration files.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta7 [March 7, 2006]
fukasawa e60969
  Removed libpng.vers and libpng.sym from libpng12_la_SOURCES in Makefile.am
fukasawa e60969
  Fixed inconsistent #ifdef's around png_sig_bytes() and png_set_sCAL_s()
fukasawa e60969
    in png.h.
fukasawa e60969
  Updated makefile.elf as suggested by debian.
fukasawa e60969
  Made cosmetic changes to some makefiles, adding LN_SF and other macros.
fukasawa e60969
  Made some makefiles accept "exec_prefix".
fukasawa e60969
fukasawa e60969
Version 1.2.9beta8 [March 9, 2006]
fukasawa e60969
  Fixed some "#if defined (..." which should be "#if defined(..."
fukasawa e60969
    Bug introduced in libpng-1.2.8.
fukasawa e60969
  Fixed inconsistency in definition of png_default_read_data()
fukasawa e60969
  Restored blank that was lost from makefile.sggcc "clean" target in beta7.
fukasawa e60969
  Revised calculation of "current" and "major" for irix in ltmain.sh
fukasawa e60969
  Changed "mkdir" to "MKDIR_P" in some makefiles.
fukasawa e60969
  Separated PNG_EXPAND and PNG_EXPAND_tRNS.
fukasawa e60969
  Added png_set_expand_gray_1_2_4_to_8() and deprecated
fukasawa e60969
    png_set_gray_1_2_4_to_8() which also expands tRNS to alpha.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta9 [March 10, 2006]
fukasawa e60969
  Include "config.h" in pngconf.h when available.
fukasawa e60969
  Added some checks for NULL png_ptr or NULL info_ptr (timeless)
fukasawa e60969
fukasawa e60969
Version 1.2.9beta10 [March 20, 2006]
fukasawa e60969
  Removed extra CR from contrib/visualpng/VisualPng.dsw (Cosmin)
fukasawa e60969
  Made pnggccrd.c PIC-compliant (Christian Aichinger).
fukasawa e60969
  Added makefile.mingw (Wolfgang Glas).
fukasawa e60969
  Revised pngconf.h MMX checking.
fukasawa e60969
fukasawa e60969
Version 1.2.9beta11 [March 22, 2006]
fukasawa e60969
  Fixed out-of-order declaration in pngwrite.c that was introduced in beta9
fukasawa e60969
  Simplified some makefiles by using LIBSO, LIBSOMAJ, and LIBSOVER macros.
fukasawa e60969
fukasawa e60969
Version 1.2.9rc1 [March 31, 2006]
fukasawa e60969
  Defined PNG_USER_PRIVATEBUILD when including "pngusr.h" (Cosmin).
fukasawa e60969
  Removed nonsensical assertion check from pngtest.c (Cosmin).
fukasawa e60969
fukasawa e60969
Version 1.2.9 [April 14, 2006]
fukasawa e60969
  Revised makefile.beos and added "none" selector in ltmain.sh
fukasawa e60969
fukasawa e60969
Version 1.2.10beta1 [April 15, 2006]
fukasawa e60969
  Renamed "config.h" to "png_conf.h" and revised Makefile.am to add
fukasawa e60969
    -DPNG_BUILDING_LIBPNG to compile directive, and modified pngconf.h
fukasawa e60969
    to include png_conf.h only when PNG_BUILDING_LIBPNG is defined.
fukasawa e60969
fukasawa e60969
Version 1.2.10beta2 [April 15, 2006]
fukasawa e60969
  Manually updated Makefile.in and configure.  Changed png_conf.h.in
fukasawa e60969
    back to config.h.
fukasawa e60969
fukasawa e60969
Version 1.2.10beta3 [April 15, 2006]
fukasawa e60969
  Change png_conf.h back to config.h in pngconf.h.
fukasawa e60969
fukasawa e60969
Version 1.2.10beta4 [April 16, 2006]
fukasawa e60969
  Change PNG_BUILDING_LIBPNG to PNG_CONFIGURE_LIBPNG in config/Makefile*.
fukasawa e60969
fukasawa e60969
Version 1.2.10beta5 [April 16, 2006]
fukasawa e60969
  Added a configure check for compiling assembler code in pnggccrd.c
fukasawa e60969
fukasawa e60969
Version 1.2.10beta6 [April 17, 2006]
fukasawa e60969
  Revised the configure check for pnggccrd.c
fukasawa e60969
  Moved -DPNG_CONFIGURE_LIBPNG into @LIBPNG_DEFINES@
fukasawa e60969
  Added @LIBPNG_DEFINES@ to arguments when building libpng.sym
fukasawa e60969
fukasawa e60969
Version 1.2.10beta7 [April 18, 2006]
fukasawa e60969
  Change "exec_prefix=$prefix" to "exec_prefix=$(prefix)" in makefiles.
fukasawa e60969
fukasawa e60969
Version 1.2.10rc1 [April 19, 2006]
fukasawa e60969
  Ensure pngconf.h doesn't define both PNG_USE_PNGGCCRD and PNG_USE_PNGVCRD
fukasawa e60969
  Fixed "LN_FS" typo in makefile.sco and makefile.solaris.
fukasawa e60969
fukasawa e60969
Version 1.2.10rc2 [April 20, 2006]
fukasawa e60969
  Added a backslash between -DPNG_CONFIGURE_LIBPNG and -DPNG_NO_ASSEMBLER_CODE
fukasawa e60969
   in configure.ac and configure
fukasawa e60969
  Made the configure warning about versioned symbols less arrogant.
fukasawa e60969
fukasawa e60969
Version 1.2.10rc3 [April 21, 2006]
fukasawa e60969
  Added a note in libpng.txt that png_set_sig_bytes(8) can be used when
fukasawa e60969
    writing an embedded PNG without the 8-byte signature.
fukasawa e60969
  Revised makefiles and configure to avoid making links to libpng.so.*
fukasawa e60969
fukasawa e60969
Version 1.2.10 [April 23, 2006]
fukasawa e60969
  Reverted configure to "rc2" state.
fukasawa e60969
fukasawa e60969
Version 1.2.11beta1 [May 31, 2006]
fukasawa e60969
  scripts/libpng.pc.in contained "configure" style version info and would
fukasawa e60969
    not work with makefiles.
fukasawa e60969
  The shared-library makefiles were linking to libpng.so.0 instead of
fukasawa e60969
    libpng.so.3 compatibility as the library.
fukasawa e60969
fukasawa e60969
Version 1.2.11beta2 [June 2, 2006]
fukasawa e60969
  Increased sprintf buffer from 50 to 52 chars in pngrutil.c to avoid
fukasawa e60969
    buffer overflow.
fukasawa e60969
  Fixed bug in example.c (png_set_palette_rgb -> png_set_palette_to_rgb)
fukasawa e60969
fukasawa e60969
Version 1.2.11beta3 [June 5, 2006]
fukasawa e60969
  Prepended "#! /bin/sh" to ltmail.sh and contrib/pngminus/*.sh (Cosmin).
fukasawa e60969
  Removed the accidental leftover Makefile.in~ (Cosmin).
fukasawa e60969
  Avoided potential buffer overflow and optimized buffer in
fukasawa e60969
    png_write_sCAL(), png_write_sCAL_s() (Cosmin).
fukasawa e60969
  Removed the include directories and libraries from CFLAGS and LDFLAGS
fukasawa e60969
    in scripts/makefile.gcc (Nelson A. de Oliveira, Cosmin).
fukasawa e60969
fukasawa e60969
Version 1.2.11beta4 [June 6, 2006]
fukasawa e60969
  Allow zero-length IDAT chunks after the entire zlib datastream, but not
fukasawa e60969
    after another intervening chunk type.
fukasawa e60969
fukasawa e60969
Version 1.0.19rc1, 1.2.11rc1 [June 13, 2006]
fukasawa e60969
  Deleted extraneous square brackets from [config.h] in configure.ac
fukasawa e60969
fukasawa e60969
Version 1.0.19rc2, 1.2.11rc2 [June 14, 2006]
fukasawa e60969
  Added prototypes for PNG_INCH_CONVERSIONS functions to png.h
fukasawa e60969
  Revised INSTALL and autogen.sh
fukasawa e60969
  Fixed typo in several makefiles (-W1 should be -Wl)
fukasawa e60969
  Added typedef for png_int_32 and png_uint_32 on 64-bit systems.
fukasawa e60969
fukasawa e60969
Version 1.0.19rc3, 1.2.11rc3 [June 15, 2006]
fukasawa e60969
  Removed the new typedefs for 64-bit systems (delay until version 1.4.0)
fukasawa e60969
  Added one zero element to png_gamma_shift[] array in pngrtran.c to avoid
fukasawa e60969
    reading out of bounds.
fukasawa e60969
fukasawa e60969
Version 1.0.19rc4, 1.2.11rc4 [June 15, 2006]
fukasawa e60969
  Really removed the new typedefs for 64-bit systems.
fukasawa e60969
fukasawa e60969
Version 1.0.19rc5, 1.2.11rc5 [June 22, 2006]
fukasawa e60969
  Removed png_sig_bytes entry from scripts/pngw32.def
fukasawa e60969
fukasawa e60969
Version 1.0.19, 1.2.11 [June 26, 2006]
fukasawa e60969
  None.
fukasawa e60969
fukasawa e60969
Version 1.0.20, 1.2.12 [June 27, 2006]
fukasawa e60969
  Really increased sprintf buffer from 50 to 52 chars in pngrutil.c to avoid
fukasawa e60969
    buffer overflow.
fukasawa e60969
fukasawa e60969
Version 1.2.13beta1 [October 2, 2006]
fukasawa e60969
  Removed AC_FUNC_MALLOC from configure.ac
fukasawa e60969
  Work around Intel-Mac compiler bug by setting PNG_NO_MMX_CODE in pngconf.h
fukasawa e60969
  Change "logical" to "bitwise" throughout documentation.
fukasawa e60969
  Detect and fix attempt to write wrong iCCP profile length (CVE-2006-7244)
fukasawa e60969
fukasawa e60969
Version 1.0.21, 1.2.13 [November 14, 2006]
fukasawa e60969
  Fix potential buffer overflow in sPLT chunk handler.
fukasawa e60969
  Fix Makefile.am to not try to link to noexistent files.
fukasawa e60969
  Check all exported functions for NULL png_ptr.
fukasawa e60969
fukasawa e60969
Version 1.2.14beta1 [November 17, 2006]
fukasawa e60969
  Relocated three misplaced tests for NULL png_ptr.
fukasawa e60969
  Built Makefile.in with automake-1.9.6 instead of 1.9.2.
fukasawa e60969
  Build configure with autoconf-2.60 instead of 2.59
fukasawa e60969
fukasawa e60969
Version 1.2.14beta2 [November 17, 2006]
fukasawa e60969
  Added some typecasts in png_zalloc().
fukasawa e60969
fukasawa e60969
Version 1.2.14rc1 [November 20, 2006]
fukasawa e60969
  Changed "strtod" to "png_strtod" in pngrutil.c
fukasawa e60969
fukasawa e60969
Version 1.0.22, 1.2.14    [November 27, 2006]
fukasawa e60969
  Added missing "$(srcdir)" in Makefile.am and Makefile.in
fukasawa e60969
fukasawa e60969
Version 1.2.15beta1 [December 3, 2006]
fukasawa e60969
  Generated configure with autoconf-2.61 instead of 2.60
fukasawa e60969
  Revised configure.ac to update libpng.pc and libpng-config.
fukasawa e60969
fukasawa e60969
Version 1.2.15beta2 [December 3, 2006]
fukasawa e60969
  Always export MMX asm functions, just stubs if not building pnggccrd.c
fukasawa e60969
fukasawa e60969
Version 1.2.15beta3 [December 4, 2006]
fukasawa e60969
  Add "png_bytep" typecast to profile while calculating length in pngwutil.c
fukasawa e60969
fukasawa e60969
Version 1.2.15beta4 [December 7, 2006]
fukasawa e60969
  Added scripts/CMakeLists.txt
fukasawa e60969
  Changed PNG_NO_ASSEMBLER_CODE to PNG_NO_MMX_CODE in scripts, like 1.4.0beta
fukasawa e60969
fukasawa e60969
Version 1.2.15beta5 [December 7, 2006]
fukasawa e60969
  Changed some instances of PNG_ASSEMBLER_* to PNG_MMX_* in pnggccrd.c
fukasawa e60969
  Revised scripts/CMakeLists.txt
fukasawa e60969
fukasawa e60969
Version 1.2.15beta6 [December 13, 2006]
fukasawa e60969
  Revised scripts/CMakeLists.txt and configure.ac
fukasawa e60969
fukasawa e60969
Version 1.2.15rc1 [December 18, 2006]
fukasawa e60969
  Revised scripts/CMakeLists.txt
fukasawa e60969
fukasawa e60969
Version 1.2.15rc2 [December 21, 2006]
fukasawa e60969
  Added conditional #undef jmpbuf in pngtest.c to undo #define in AIX headers.
fukasawa e60969
  Added scripts/makefile.nommx
fukasawa e60969
fukasawa e60969
Version 1.2.15rc3 [December 25, 2006]
fukasawa e60969
  Fixed shared library numbering error that was introduced in 1.2.15beta6.
fukasawa e60969
fukasawa e60969
Version 1.2.15rc4 [December 27, 2006]
fukasawa e60969
  Fixed handling of rgb_to_gray when png_ptr->color.gray isn't set.
fukasawa e60969
fukasawa e60969
Version 1.2.15rc5 [December 31, 2006]
fukasawa e60969
  Revised handling of rgb_to_gray.
fukasawa e60969
fukasawa e60969
Version 1.2.15 [January 5, 2007]
fukasawa e60969
  Added some (unsigned long) typecasts in pngtest.c to avoid printing errors.
fukasawa e60969
fukasawa e60969
Version 1.2.16beta1 [January 6, 2007]
fukasawa e60969
  Fix bugs in makefile.nommx
fukasawa e60969
fukasawa e60969
Version 1.2.16beta2 [January 16, 2007]
fukasawa e60969
  Revised scripts/CMakeLists.txt
fukasawa e60969
fukasawa e60969
Version 1.2.16 [January 31, 2007]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.2.17beta1 [March 6, 2007]
fukasawa e60969
  Revised scripts/CMakeLists.txt to install both shared and static libraries.
fukasawa e60969
  Deleted a redundant line from pngset.c.
fukasawa e60969
fukasawa e60969
Version 1.2.17beta2 [April 26, 2007]
fukasawa e60969
  Relocated misplaced test for png_ptr == NULL in pngpread.c
fukasawa e60969
  Change "==" to "&" for testing PNG_RGB_TO_GRAY_ERR & PNG_RGB_TO_GRAY_WARN
fukasawa e60969
    flags.
fukasawa e60969
  Changed remaining instances of PNG_ASSEMBLER_* to PNG_MMX_*
fukasawa e60969
  Added pngerror() when write_IHDR fails in deflateInit2().
fukasawa e60969
  Added "const" to some array declarations.
fukasawa e60969
  Mention examples of libpng usage in the libpng*.txt and libpng.3 documents.
fukasawa e60969
fukasawa e60969
Version 1.2.17rc1 [May 4, 2007]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.2.17rc2 [May 8, 2007]
fukasawa e60969
  Moved several PNG_HAVE_* macros out of PNG_INTERNAL because applications
fukasawa e60969
    calling set_unknown_chunk_location() need them.
fukasawa e60969
  Changed transformation flag from PNG_EXPAND_tRNS to PNG_EXPAND in
fukasawa e60969
    png_set_expand_gray_1_2_4_to_8().
fukasawa e60969
  Added png_ptr->unknown_chunk to hold working unknown chunk data, so it
fukasawa e60969
    can be free'ed in case of error.  Revised unknown chunk handling in
fukasawa e60969
    pngrutil.c and pngpread.c to use this structure.
fukasawa e60969
fukasawa e60969
Version 1.2.17rc3 [May 8, 2007]
fukasawa e60969
  Revised symbol-handling in configure script.
fukasawa e60969
fukasawa e60969
Version 1.2.17rc4 [May 10, 2007]
fukasawa e60969
  Revised unknown chunk handling to avoid storing unknown critical chunks.
fukasawa e60969
fukasawa e60969
Version 1.0.25 [May 15, 2007]
fukasawa e60969
Version 1.2.17 [May 15, 2007]
fukasawa e60969
  Added "png_ptr->num_trans=0" before error return in png_handle_tRNS,
fukasawa e60969
    to eliminate a vulnerability (CVE-2007-2445, CERT VU#684664)
fukasawa e60969
fukasawa e60969
Version 1.0.26 [May 15, 2007]
fukasawa e60969
Version 1.2.18 [May 15, 2007]
fukasawa e60969
  Reverted the libpng-1.2.17rc3 change to symbol-handling in configure script
fukasawa e60969
fukasawa e60969
Version 1.2.19beta1 [May 18, 2007]
fukasawa e60969
  Changed "const static" to "static PNG_CONST" everywhere, mostly undoing
fukasawa e60969
    change of libpng-1.2.17beta2.  Changed other "const" to "PNG_CONST"
fukasawa e60969
  Changed some handling of unused parameters, to avoid compiler warnings.
fukasawa e60969
    "if (unused == NULL) return;" becomes "unused = unused".
fukasawa e60969
fukasawa e60969
Version 1.2.19beta2 [May 18, 2007]
fukasawa e60969
  Only use the valid bits of tRNS value in png_do_expand() (Brian Cartier)
fukasawa e60969
fukasawa e60969
Version 1.2.19beta3 [May 19, 2007]
fukasawa e60969
  Add some "png_byte" typecasts in png_check_keyword() and write new_key
fukasawa e60969
  instead of key in zTXt chunk (Kevin Ryde).
fukasawa e60969
fukasawa e60969
Version 1.2.19beta4 [May 21, 2007]
fukasawa e60969
  Add png_snprintf() function and use it in place of sprint() for improved
fukasawa e60969
    defense against buffer overflows.
fukasawa e60969
fukasawa e60969
Version 1.2.19beta5 [May 21, 2007]
fukasawa e60969
  Fixed png_handle_tRNS() to only use the valid bits of tRNS value.
fukasawa e60969
  Changed handling of more unused parameters, to avoid compiler warnings.
fukasawa e60969
  Removed some PNG_CONST in pngwutil.c to avoid compiler warnings.
fukasawa e60969
fukasawa e60969
Version 1.2.19beta6 [May 22, 2007]
fukasawa e60969
  Added some #ifdef PNG_MMX_CODE_SUPPORTED where needed in pngvcrd.c
fukasawa e60969
  Added a special "_MSC_VER" case that defines png_snprintf to _snprintf
fukasawa e60969
fukasawa e60969
Version 1.2.19beta7 [May 22, 2007]
fukasawa e60969
  Squelched png_squelch_warnings() in pnggccrd.c and added
fukasawa e60969
    an #ifdef PNG_MMX_CODE_SUPPORTED block around the declarations that caused
fukasawa e60969
    the warnings that png_squelch_warnings was squelching.
fukasawa e60969
fukasawa e60969
Version 1.2.19beta8 [May 22, 2007]
fukasawa e60969
  Removed __MMX__ from test in pngconf.h.
fukasawa e60969
fukasawa e60969
Version 1.2.19beta9 [May 23, 2007]
fukasawa e60969
  Made png_squelch_warnings() available via PNG_SQUELCH_WARNINGS macro.
fukasawa e60969
  Revised png_squelch_warnings() so it might work.
fukasawa e60969
  Updated makefile.sgcc and makefile.solaris; added makefile.solaris-x86.
fukasawa e60969
fukasawa e60969
Version 1.2.19beta10 [May 24, 2007]
fukasawa e60969
  Resquelched png_squelch_warnings(), use "__attribute__((used))" instead.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta1 [April 20, 2006]
fukasawa e60969
  Enabled iTXt support (changes png_struct, thus requires so-number change).
fukasawa e60969
  Cleaned up PNG_ASSEMBLER_CODE_SUPPORTED vs PNG_MMX_CODE_SUPPORTED
fukasawa e60969
  Eliminated PNG_1_0_X and PNG_1_2_X macros.
fukasawa e60969
  Removed deprecated functions png_read_init, png_write_init, png_info_init,
fukasawa e60969
    png_permit_empty_plte, png_set_gray_1_2_4_to_8, png_check_sig, and
fukasawa e60969
    removed the deprecated macro PNG_MAX_UINT.
fukasawa e60969
  Moved "PNG_INTERNAL" parts of png.h and pngconf.h into pngintrn.h
fukasawa e60969
  Removed many WIN32_WCE #ifdefs (Cosmin).
fukasawa e60969
  Reduced dependency on C-runtime library when on Windows (Simon-Pierre)
fukasawa e60969
  Replaced sprintf() with png_sprintf() (Simon-Pierre)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta2 [April 20, 2006]
fukasawa e60969
  Revised makefiles and configure to avoid making links to libpng.so.*
fukasawa e60969
  Moved some leftover MMX-related defines from pngconf.h to pngintrn.h
fukasawa e60969
  Updated scripts/pngos2.def, pngw32.def, and projects/wince/png32ce.def
fukasawa e60969
fukasawa e60969
Version 1.4.0beta3 [May 10, 2006]
fukasawa e60969
  Updated scripts/pngw32.def to comment out MMX functions.
fukasawa e60969
  Added PNG_NO_GET_INT_32 and PNG_NO_SAVE_INT_32 macros.
fukasawa e60969
  Scripts/libpng.pc.in contained "configure" style version info and would
fukasawa e60969
    not work with makefiles.
fukasawa e60969
  Revised pngconf.h and added pngconf.h.in, so makefiles and configure can
fukasawa e60969
    pass defines to libpng and applications.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta4 [May 11, 2006]
fukasawa e60969
  Revised configure.ac, Makefile.am, and many of the makefiles to write
fukasawa e60969
    their defines in pngconf.h.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta5 [May 15, 2006]
fukasawa e60969
  Added a missing semicolon in Makefile.am and Makefile.in
fukasawa e60969
  Deleted extraneous square brackets from configure.ac
fukasawa e60969
fukasawa e60969
Version 1.4.0beta6 [June 2, 2006]
fukasawa e60969
  Increased sprintf buffer from 50 to 52 chars in pngrutil.c to avoid
fukasawa e60969
    buffer overflow.
fukasawa e60969
  Changed sonum from 0 to 1.
fukasawa e60969
  Removed unused prototype for png_check_sig() from png.h
fukasawa e60969
fukasawa e60969
Version 1.4.0beta7 [June 16, 2006]
fukasawa e60969
  Exported png_write_sig (Cosmin).
fukasawa e60969
  Optimized buffer in png_handle_cHRM() (Cosmin).
fukasawa e60969
  Set pHYs = 2835 x 2835 pixels per meter, and added
fukasawa e60969
    sCAL = 0.352778e-3 x 0.352778e-3 meters, in pngtest.png (Cosmin).
fukasawa e60969
  Added png_set_benign_errors(), png_benign_error(), png_chunk_benign_error().
fukasawa e60969
  Added typedef for png_int_32 and png_uint_32 on 64-bit systems.
fukasawa e60969
  Added "(unsigned long)" typecast on png_uint_32 variables in printf lists.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta8 [June 22, 2006]
fukasawa e60969
  Added demonstration of user chunk support in pngtest.c, to support the
fukasawa e60969
    public sTER chunk and a private vpAg chunk.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta9 [July 3, 2006]
fukasawa e60969
  Removed ordinals from scripts/pngw32.def and removed png_info_int and
fukasawa e60969
    png_set_gray_1_2_4_to_8 entries.
fukasawa e60969
  Inline call of png_get_uint_32() in png_get_uint_31().
fukasawa e60969
  Use png_get_uint_31() to get vpAg width and height in pngtest.c
fukasawa e60969
  Removed WINCE and Netware projects.
fukasawa e60969
  Removed standalone Y2KINFO file.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta10 [July 12, 2006]
fukasawa e60969
  Eliminated automatic copy of pngconf.h to pngconf.h.in from configure and
fukasawa e60969
    some makefiles, because it was not working reliably.  Instead, distribute
fukasawa e60969
    pngconf.h.in along with pngconf.h and cause configure and some of the
fukasawa e60969
    makefiles to update pngconf.h from pngconf.h.in.
fukasawa e60969
  Added pngconf.h to DEPENDENCIES in Makefile.am
fukasawa e60969
fukasawa e60969
Version 1.4.0beta11 [August 19, 2006]
fukasawa e60969
  Removed AC_FUNC_MALLOC from configure.ac.
fukasawa e60969
  Added a warning when writing iCCP profile with mismatched profile length.
fukasawa e60969
  Patched pnggccrd.c to assemble on x86_64 platforms.
fukasawa e60969
  Moved chunk header reading into a separate function png_read_chunk_header()
fukasawa e60969
    in pngrutil.c.  The chunk header (len+sig) is now serialized in a single
fukasawa e60969
    operation (Cosmin).
fukasawa e60969
  Implemented support for I/O states. Added png_ptr member io_state, and
fukasawa e60969
    functions png_get_io_chunk_name() and png_get_io_state() in pngget.c
fukasawa e60969
    (Cosmin).
fukasawa e60969
  Added png_get_io_chunk_name and png_get_io_state to scripts/*.def (Cosmin).
fukasawa e60969
  Renamed scripts/pngw32.* to scripts/pngwin.* (Cosmin).
fukasawa e60969
  Removed the include directories and libraries from CFLAGS and LDFLAGS
fukasawa e60969
    in scripts/makefile.gcc (Cosmin).
fukasawa e60969
  Used png_save_uint_32() to set vpAg width and height in pngtest.c (Cosmin).
fukasawa e60969
  Cast to proper type when getting/setting vpAg units in pngtest.c (Cosmin).
fukasawa e60969
  Added pngintrn.h to the Visual C++ projects (Cosmin).
fukasawa e60969
  Removed scripts/list (Cosmin).
fukasawa e60969
  Updated copyright year in scripts/pngwin.def (Cosmin).
fukasawa e60969
  Removed PNG_TYPECAST_NULL and used standard NULL consistently (Cosmin).
fukasawa e60969
  Disallowed the user to redefine png_size_t, and enforced a consistent use
fukasawa e60969
    of png_size_t across libpng (Cosmin).
fukasawa e60969
  Changed the type of png_ptr->rowbytes, PNG_ROWBYTES() and friends
fukasawa e60969
    to png_size_t (Cosmin).
fukasawa e60969
  Removed png_convert_size() and replaced png_sizeof with sizeof (Cosmin).
fukasawa e60969
  Removed some unnecessary type casts (Cosmin).
fukasawa e60969
  Changed prototype of png_get_compression_buffer_size() and
fukasawa e60969
    png_set_compression_buffer_size() to work with png_size_t instead of
fukasawa e60969
    png_uint_32 (Cosmin).
fukasawa e60969
  Removed png_memcpy_check() and png_memset_check() (Cosmin).
fukasawa e60969
  Fixed a typo (png_byte --> png_bytep) in libpng.3 and libpng.txt (Cosmin).
fukasawa e60969
  Clarified that png_zalloc() does not clear the allocated memory,
fukasawa e60969
    and png_zalloc() and png_zfree() cannot be PNGAPI (Cosmin).
fukasawa e60969
  Renamed png_mem_size_t to png_alloc_size_t, fixed its definition in
fukasawa e60969
    pngconf.h, and used it in all memory allocation functions (Cosmin).
fukasawa e60969
  Renamed pngintrn.h to pngpriv.h, added a comment at the top of the file
fukasawa e60969
    mentioning that the symbols declared in that file are private, and
fukasawa e60969
    updated the scripts and the Visual C++ projects accordingly (Cosmin).
fukasawa e60969
  Removed circular references between pngconf.h and pngconf.h.in in
fukasawa e60969
    scripts/makefile.vc*win32 (Cosmin).
fukasawa e60969
  Removing trailing '.' from the warning and error messages (Cosmin).
fukasawa e60969
  Added pngdefs.h that is built by makefile or configure, instead of
fukasawa e60969
    pngconf.h.in (Glenn).
fukasawa e60969
  Detect and fix attempt to write wrong iCCP profile length.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta12 [October 19, 2006]
fukasawa e60969
  Changed "logical" to "bitwise" in the documentation.
fukasawa e60969
  Work around Intel-Mac compiler bug by setting PNG_NO_MMX_CODE in pngconf.h
fukasawa e60969
  Add a typecast to stifle compiler warning in pngrutil.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta13 [November 10, 2006]
fukasawa e60969
  Fix potential buffer overflow in sPLT chunk handler.
fukasawa e60969
  Fix Makefile.am to not try to link to noexistent files.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta14 [November 15, 2006]
fukasawa e60969
  Check all exported functions for NULL png_ptr.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta15 [November 17, 2006]
fukasawa e60969
  Relocated two misplaced tests for NULL png_ptr.
fukasawa e60969
  Built Makefile.in with automake-1.9.6 instead of 1.9.2.
fukasawa e60969
  Build configure with autoconf-2.60 instead of 2.59
fukasawa e60969
  Add "install: all" in Makefile.am so "configure; make install" will work.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta16 [November 17, 2006]
fukasawa e60969
  Added a typecast in png_zalloc().
fukasawa e60969
fukasawa e60969
Version 1.4.0beta17 [December 4, 2006]
fukasawa e60969
  Changed "new_key[79] = '\0';" to "(*new_key)[79] = '\0';" in pngwutil.c
fukasawa e60969
  Add "png_bytep" typecast to profile while calculating length in pngwutil.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta18 [December 7, 2006]
fukasawa e60969
  Added scripts/CMakeLists.txt
fukasawa e60969
fukasawa e60969
Version 1.4.0beta19 [May 16, 2007]
fukasawa e60969
  Revised scripts/CMakeLists.txt
fukasawa e60969
  Rebuilt configure and Makefile.in with newer tools.
fukasawa e60969
  Added conditional #undef jmpbuf in pngtest.c to undo #define in AIX headers.
fukasawa e60969
  Added scripts/makefile.nommx
fukasawa e60969
fukasawa e60969
Version 1.4.0beta20 [July 9, 2008]
fukasawa e60969
  Moved several PNG_HAVE_* macros from pngpriv.h to png.h because applications
fukasawa e60969
    calling set_unknown_chunk_location() need them.
fukasawa e60969
  Moved several macro definitions from pngpriv.h to pngconf.h
fukasawa e60969
  Merge with changes to the 1.2.X branch, as of 1.2.30beta04.
fukasawa e60969
  Deleted all use of the MMX assembler code and Intel-licensed optimizations.
fukasawa e60969
  Revised makefile.mingw
fukasawa e60969
fukasawa e60969
Version 1.4.0beta21 [July 21, 2008]
fukasawa e60969
  Moved local array "chunkdata" from pngrutil.c to the png_struct, so
fukasawa e60969
    it will be freed by png_read_destroy() in case of a read error (Kurt
fukasawa e60969
    Christensen).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta22 [July 21, 2008]
fukasawa e60969
  Change "purpose" and "buffer" to png_ptr->chunkdata to avoid memory leaking.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta23 [July 22, 2008]
fukasawa e60969
  Change "chunkdata = NULL" to "png_ptr->chunkdata = NULL" several places in
fukasawa e60969
    png_decompress_chunk().
fukasawa e60969
fukasawa e60969
Version 1.4.0beta24 [July 25, 2008]
fukasawa e60969
  Change all remaining "chunkdata" to "png_ptr->chunkdata" in
fukasawa e60969
    png_decompress_chunk(), and remove "chunkdata" from parameter list.
fukasawa e60969
  Put a call to png_check_chunk_name() in png_read_chunk_header().
fukasawa e60969
  Revised png_check_chunk_name() to reject a name with a lowercase 3rd byte.
fukasawa e60969
  Removed two calls to png_check_chunk_name() occurring later in the process.
fukasawa e60969
  Define PNG_NO_ERROR_NUMBERS by default in pngconf.h
fukasawa e60969
fukasawa e60969
Version 1.4.0beta25 [July 30, 2008]
fukasawa e60969
  Added a call to png_check_chunk_name() in pngpread.c
fukasawa e60969
  Reverted png_check_chunk_name() to accept a name with a lowercase 3rd byte.
fukasawa e60969
  Added png_push_have_buffer() function to pngpread.c
fukasawa e60969
  Eliminated PNG_BIG_ENDIAN_SUPPORTED and associated png_get_* macros.
fukasawa e60969
  Made inline expansion of png_get_*() optional with PNG_USE_READ_MACROS.
fukasawa e60969
  Eliminated all PNG_USELESS_TESTS and PNG_CORRECT_PALETTE_SUPPORTED code.
fukasawa e60969
  Synced contrib directory and configure files with libpng-1.2.30beta06.
fukasawa e60969
  Eliminated no-longer-used pngdefs.h (but it's still built in the makefiles)
fukasawa e60969
  Relocated a misplaced "#endif /* PNG_NO_WRITE_FILTER */" in pngwutil.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta26 [August 4, 2008]
fukasawa e60969
  Removed png_push_have_buffer() function in pngpread.c.  It increased the
fukasawa e60969
    compiled library size slightly.
fukasawa e60969
  Changed "-Wall" to "-W -Wall" in the CFLAGS in all makefiles (Cosmin Truta)
fukasawa e60969
  Declared png_ptr "volatile" in pngread.c and pngwrite.c to avoid warnings.
fukasawa e60969
  Updated contrib/visupng/cexcept.h to version 2.0.1
fukasawa e60969
  Added PNG_LITERAL_CHARACTER macros for #, [, and ].
fukasawa e60969
fukasawa e60969
Version 1.4.0beta27 [August 5, 2008]
fukasawa e60969
  Revised usage of PNG_LITERAL_SHARP in pngerror.c.
fukasawa e60969
  Moved newline character from individual png_debug messages into the
fukasawa e60969
    png_debug macros.
fukasawa e60969
  Allow user to #define their own png_debug, png_debug1, and png_debug2.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta28 [August 5, 2008]
fukasawa e60969
  Revised usage of PNG_LITERAL_SHARP in pngerror.c.
fukasawa e60969
  Added PNG_STRING_NEWLINE macro
fukasawa e60969
fukasawa e60969
Version 1.4.0beta29 [August 9, 2008]
fukasawa e60969
  Revised usage of PNG_STRING_NEWLINE to work on non-ISO compilers.
fukasawa e60969
  Added PNG_STRING_COPYRIGHT macro.
fukasawa e60969
  Added non-ISO versions of png_debug macros.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta30 [August 14, 2008]
fukasawa e60969
  Added premultiplied alpha feature (Volker Wiendl).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta31 [August 18, 2008]
fukasawa e60969
  Moved png_set_premultiply_alpha from pngtrans.c to pngrtran.c
fukasawa e60969
  Removed extra crc check at the end of png_handle_cHRM().  Bug introduced
fukasawa e60969
    in libpng-1.4.0beta20.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta32 [August 19, 2008]
fukasawa e60969
  Added PNG_WRITE_FLUSH_SUPPORTED block around new png_flush() call.
fukasawa e60969
  Revised PNG_NO_STDIO version of png_write_flush()
fukasawa e60969
fukasawa e60969
Version 1.4.0beta33 [August 20, 2008]
fukasawa e60969
  Added png_get|set_chunk_cache_max() to limit the total number of sPLT,
fukasawa e60969
    text, and unknown chunks that can be stored.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta34 [September 6, 2008]
fukasawa e60969
  Shortened tIME_string to 29 bytes in pngtest.c
fukasawa e60969
  Fixed off-by-one error introduced in png_push_read_zTXt() function in
fukasawa e60969
    libpng-1.2.30beta04/pngpread.c (Harald van Dijk)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta35 [October 6, 2008]
fukasawa e60969
  Changed "trans_values" to "trans_color".
fukasawa e60969
  Changed so-number from 0 to 14.  Some OS do not like 0.
fukasawa e60969
  Revised makefile.darwin to fix shared library numbering.
fukasawa e60969
  Change png_set_gray_1_2_4_to_8() to png_set_expand_gray_1_2_4_to_8()
fukasawa e60969
    in example.c (debian bug report)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta36 [October 25, 2008]
fukasawa e60969
  Sync with tEXt vulnerability fix in libpng-1.2.33rc02.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta37 [November 13, 2008]
fukasawa e60969
  Added png_check_cHRM in png.c and moved checking from pngget.c, pngrutil.c,
fukasawa e60969
    and pngwrite.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta38 [November 22, 2008]
fukasawa e60969
  Added check for zero-area RGB cHRM triangle in png_check_cHRM() and
fukasawa e60969
    png_check_cHRM_fixed().
fukasawa e60969
fukasawa e60969
Version 1.4.0beta39 [November 23, 2008]
fukasawa e60969
  Revised png_warning() to write its message on standard output by default
fukasawa e60969
    when warning_fn is NULL.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta40 [November 24, 2008]
fukasawa e60969
  Eliminated png_check_cHRM().  Instead, always use png_check_cHRM_fixed().
fukasawa e60969
  In png_check_cHRM_fixed(), ensure white_y is > 0, and removed redundant
fukasawa e60969
    check for all-zero coordinates that is detected by the triangle check.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta41 [November 26, 2008]
fukasawa e60969
  Fixed string vs pointer-to-string error in png_check_keyword().
fukasawa e60969
  Rearranged test expressions in png_check_cHRM_fixed() to avoid internal
fukasawa e60969
    overflows.
fukasawa e60969
  Added PNG_NO_CHECK_cHRM conditional.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta42, 43 [December 1, 2008]
fukasawa e60969
  Merge png_debug with version 1.2.34beta04.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta44 [December 6, 2008]
fukasawa e60969
  Removed redundant check for key==NULL before calling png_check_keyword()
fukasawa e60969
    to ensure that new_key gets initialized and removed extra warning
fukasawa e60969
    (Merge with version 1.2.34beta05 -- Arvan Pritchard).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta45 [December 9, 2008]
fukasawa e60969
  In png_write_png(), respect the placement of the filler bytes in an earlier
fukasawa e60969
    call to png_set_filler() (Jim Barry).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta46 [December 10, 2008]
fukasawa e60969
  Undid previous change and added PNG_TRANSFORM_STRIP_FILLER_BEFORE and
fukasawa e60969
    PNG_TRANSFORM_STRIP_FILLER_AFTER conditionals and deprecated
fukasawa e60969
    PNG_TRANSFORM_STRIP_FILLER (Jim Barry).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta47 [December 15, 2008]
fukasawa e60969
  Support for dithering was disabled by default, because it has never
fukasawa e60969
    been well tested and doesn't work very well.  The code has not
fukasawa e60969
    been removed, however, and can be enabled by building libpng with
fukasawa e60969
    PNG_READ_DITHER_SUPPORTED defined.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta48 [February 14, 2009]
fukasawa e60969
  Added new exported function png_calloc().
fukasawa e60969
  Combined several instances of png_malloc(); png_memset() into png_calloc().
fukasawa e60969
  Removed prototype for png_freeptr() that was added in libpng-1.4.0beta24
fukasawa e60969
    but was never defined.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta49 [February 28, 2009]
fukasawa e60969
  Added png_fileno() macro to pngconf.h, used in pngwio.c
fukasawa e60969
  Corrected order of #ifdef's in png_debug definition in png.h
fukasawa e60969
  Fixed bug introduced in libpng-1.4.0beta48 with the memset arguments
fukasawa e60969
    for pcal_params.
fukasawa e60969
  Fixed order of #ifdef directives in the png_debug defines in png.h
fukasawa e60969
    (bug introduced in libpng-1.2.34/1.4.0beta29).
fukasawa e60969
  Revised comments in png_set_read_fn() and png_set_write_fn().
fukasawa e60969
fukasawa e60969
Version 1.4.0beta50 [March 18, 2009]
fukasawa e60969
  Use png_calloc() instead of png_malloc() to allocate big_row_buf when
fukasawa e60969
    reading an interlaced file, to avoid a possible UMR.
fukasawa e60969
  Undid revision of PNG_NO_STDIO version of png_write_flush().  Users
fukasawa e60969
    having trouble with fflush() can build with PNG_NO_WRITE_FLUSH defined
fukasawa e60969
    or supply their own flush_fn() replacement.
fukasawa e60969
  Revised libpng*.txt and png.h documentation about use of png_write_flush()
fukasawa e60969
    and png_set_write_fn().
fukasawa e60969
  Removed fflush() from pngtest.c.
fukasawa e60969
  Added "#define PNG_NO_WRITE_FLUSH" to contrib/pngminim/encoder/pngusr.h
fukasawa e60969
fukasawa e60969
Version 1.4.0beta51 [March 21, 2009]
fukasawa e60969
  Removed new png_fileno() macro from pngconf.h .
fukasawa e60969
fukasawa e60969
Version 1.4.0beta52 [March 27, 2009]
fukasawa e60969
  Relocated png_do_chop() ahead of building gamma tables in pngrtran.c
fukasawa e60969
    This avoids building 16-bit gamma tables unnecessarily.
fukasawa e60969
  Removed fflush() from pngtest.c.
fukasawa e60969
  Added "#define PNG_NO_WRITE_FLUSH" to contrib/pngminim/encoder/pngusr.h
fukasawa e60969
  Added a section on differences between 1.0.x and 1.2.x to libpng.3/libpng.txt
fukasawa e60969
fukasawa e60969
Version 1.4.0beta53 [April 1, 2009]
fukasawa e60969
  Removed some remaining MMX macros from pngpriv.h
fukasawa e60969
  Fixed potential memory leak of "new_name" in png_write_iCCP() (Ralph Giles)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta54 [April 13, 2009]
fukasawa e60969
  Added "ifndef PNG_SKIP_SETJMP_CHECK" block in pngconf.h to allow
fukasawa e60969
    application code writers to bypass the check for multiple inclusion
fukasawa e60969
    of setjmp.h when they know that it is safe to ignore the situation.
fukasawa e60969
  Eliminated internal use of setjmp() in pngread.c and pngwrite.c
fukasawa e60969
  Reordered ancillary chunks in pngtest.png to be the same as what
fukasawa e60969
    pngtest now produces, and made some cosmetic changes to pngtest output.
fukasawa e60969
  Eliminated deprecated png_read_init_3() and png_write_init_3() functions.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta55 [April 15, 2009]
fukasawa e60969
  Simplified error handling in pngread.c and pngwrite.c by putting
fukasawa e60969
    the new png_read_cleanup() and png_write_cleanup() functions inline.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta56 [April 25, 2009]
fukasawa e60969
  Renamed "user_chunk_data" to "my_user_chunk_data" in pngtest.c to suppress
fukasawa e60969
    "shadowed declaration" warning from gcc-4.3.3.
fukasawa e60969
  Renamed "gamma" to "png_gamma" in pngset.c to avoid "shadowed declaration"
fukasawa e60969
    warning about a global "gamma" variable in math.h on some platforms.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta57 [May 2, 2009]
fukasawa e60969
  Removed prototype for png_freeptr() that was added in libpng-1.4.0beta24
fukasawa e60969
    but was never defined (again).
fukasawa e60969
  Rebuilt configure scripts with autoconf-2.63 instead of 2.62
fukasawa e60969
  Removed pngprefs.h and MMX from makefiles
fukasawa e60969
fukasawa e60969
Version 1.4.0beta58 [May 14, 2009]
fukasawa e60969
  Changed pngw32.def to pngwin.def in makefile.mingw (typo was introduced
fukasawa e60969
    in beta57).
fukasawa e60969
  Clarified usage of sig_bit versus sig_bit_p in example.c (Vincent Torri)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta59 [May 15, 2009]
fukasawa e60969
  Reformated sources in libpng style (3-space intentation, comment format)
fukasawa e60969
  Fixed typo in libpng docs (PNG_FILTER_AVE should be PNG_FILTER_AVG)
fukasawa e60969
  Added sections about the git repository and our coding style to the
fukasawa e60969
    documentation
fukasawa e60969
  Relocated misplaced #endif in pngwrite.c, sCAL chunk handler.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta60 [May 19, 2009]
fukasawa e60969
  Conditionally compile png_read_finish_row() which is not used by
fukasawa e60969
    progressive readers.
fukasawa e60969
  Added contrib/pngminim/preader to demonstrate building minimal progressive
fukasawa e60969
    decoder, based on contrib/gregbook with embedded libpng and zlib.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta61 [May 20, 2009]
fukasawa e60969
  In contrib/pngminim/*, renamed "makefile.std" to "makefile", since there
fukasawa e60969
    is only one makefile in those directories, and revised the README files
fukasawa e60969
    accordingly.
fukasawa e60969
  More reformatting of comments, mostly to capitalize sentences.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta62 [June 2, 2009]
fukasawa e60969
  Added "#define PNG_NO_WRITE_SWAP" to contrib/pngminim/encoder/pngusr.h
fukasawa e60969
    and "define PNG_NO_READ_SWAP" to decoder/pngusr.h and preader/pngusr.h
fukasawa e60969
  Reformatted several remaining "else statement" into two lines.
fukasawa e60969
  Added a section to the libpng documentation about using png_get_io_ptr()
fukasawa e60969
    in configure scripts to detect the presence of libpng.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta63 [June 15, 2009]
fukasawa e60969
  Revised libpng*.txt and libpng.3 to mention calling png_set_IHDR()
fukasawa e60969
    multiple times and to specify the sample order in the tRNS chunk,
fukasawa e60969
    because the ISO PNG specification has a typo in the tRNS table.
fukasawa e60969
  Changed several PNG_UNKNOWN_CHUNK_SUPPORTED to
fukasawa e60969
    PNG_HANDLE_AS_UNKNOWN_SUPPORTED, to make the png_set_keep mechanism
fukasawa e60969
    available for ignoring known chunks even when not saving unknown chunks.
fukasawa e60969
  Adopted preference for consistent use of "#ifdef" and "#ifndef" versus
fukasawa e60969
    "#if defined()" and "if !defined()" where possible.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta64 [June 24, 2009]
fukasawa e60969
  Eliminated PNG_LEGACY_SUPPORTED code.
fukasawa e60969
  Moved the various unknown chunk macro definitions outside of the
fukasawa e60969
    PNG_READ|WRITE_ANCILLARY_CHUNK_SUPPORTED blocks.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta65 [June 26, 2009]
fukasawa e60969
  Added a reference to the libpng license in each file.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta66 [June 27, 2009]
fukasawa e60969
  Refer to the libpng license instead of the libpng license in each file.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta67 [July 6, 2009]
fukasawa e60969
  Relocated INVERT_ALPHA within png_read_png() and png_write_png().
fukasawa e60969
  Added high-level API transform PNG_TRANSFORM_GRAY_TO_RGB.
fukasawa e60969
  Added an "xcode" project to the projects directory (Alam Arias).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta68 [July 19, 2009]
fukasawa e60969
  Avoid some tests in filter selection in pngwutil.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta69 [July 25, 2009]
fukasawa e60969
  Simplified the new filter-selection test.  This runs faster in the
fukasawa e60969
    common "PNG_ALL_FILTERS" and PNG_FILTER_NONE cases.
fukasawa e60969
  Removed extraneous declaration from the new call to png_read_gray_to_rgb()
fukasawa e60969
    (bug introduced in libpng-1.4.0beta67).
fukasawa e60969
  Fixed up xcode project (Alam Arias)
fukasawa e60969
  Added a prototype for png_64bit_product() in png.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta70 [July 27, 2009]
fukasawa e60969
  Avoid a possible NULL dereference in debug build, in png_set_text_2().
fukasawa e60969
    (bug introduced in libpng-0.95, discovered by Evan Rouault)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta71 [July 29, 2009]
fukasawa e60969
  Rebuilt configure scripts with autoconf-2.64.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta72 [August 1, 2009]
fukasawa e60969
  Replaced *.tar.lzma with *.tar.xz in distribution.  Get the xz codec
fukasawa e60969
    from <http://tukaani.org/xz>.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta73 [August 1, 2009]
fukasawa e60969
  Reject attempt to write iCCP chunk with negative embedded profile length
fukasawa e60969
    (JD Chen) (CVE-2009-5063).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta74 [August 8, 2009]
fukasawa e60969
  Changed png_ptr and info_ptr member "trans" to "trans_alpha".
fukasawa e60969
fukasawa e60969
Version 1.4.0beta75 [August 21, 2009]
fukasawa e60969
  Removed an extra png_debug() recently added to png_write_find_filter().
fukasawa e60969
  Fixed incorrect #ifdef in pngset.c regarding unknown chunk support.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta76 [August 22, 2009]
fukasawa e60969
  Moved an incorrectly located test in png_read_row() in pngread.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta77 [August 27, 2009]
fukasawa e60969
  Removed lpXYZ.tar.bz2 (with CRLF), KNOWNBUG, libpng-x.y.z-KNOWNBUG.txt,
fukasawa e60969
    and the "noconfig" files from the distribution.
fukasawa e60969
  Moved CMakeLists.txt from scripts into the main libpng directory.
fukasawa e60969
  Various bugfixes and improvements to CMakeLists.txt (Philip Lowman)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta78 [August 31, 2009]
fukasawa e60969
  Converted all PNG_NO_* tests to PNG_*_SUPPORTED everywhere except pngconf.h
fukasawa e60969
  Eliminated PNG_NO_FREE_ME and PNG_FREE_ME_SUPPORTED macros.
fukasawa e60969
  Use png_malloc plus a loop instead of png_calloc() to initialize
fukasawa e60969
    row_pointers in png_read_png().
fukasawa e60969
fukasawa e60969
Version 1.4.0beta79 [September 1, 2009]
fukasawa e60969
  Eliminated PNG_GLOBAL_ARRAYS and PNG_LOCAL_ARRAYS; always use local arrays.
fukasawa e60969
  Eliminated PNG_CALLOC_SUPPORTED macro and always provide png_calloc().
fukasawa e60969
fukasawa e60969
Version 1.4.0beta80 [September 17, 2009]
fukasawa e60969
  Removed scripts/libpng.icc
fukasawa e60969
  Changed typecast of filler from png_byte to png_uint_16 in png_set_filler().
fukasawa e60969
    (Dennis Gustafsson)
fukasawa e60969
  Fixed typo introduced in beta78 in pngtest.c ("#if def " should be "#ifdef ")
fukasawa e60969
fukasawa e60969
Version 1.4.0beta81 [September 23, 2009]
fukasawa e60969
  Eliminated unused PNG_FLAG_FREE_* defines from pngpriv.h
fukasawa e60969
  Expanded TAB characters in pngrtran.c
fukasawa e60969
  Removed PNG_CONST from all "PNG_CONST PNG_CHNK" declarations to avoid
fukasawa e60969
    compiler complaints about doubly declaring things "const".
fukasawa e60969
  Changed all "#if [!]defined(X)" to "if[n]def X" where possible.
fukasawa e60969
  Eliminated unused png_ptr->row_buf_size
fukasawa e60969
fukasawa e60969
Version 1.4.0beta82 [September 25, 2009]
fukasawa e60969
  Moved redundant IHDR checking into new png_check_IHDR() in png.c
fukasawa e60969
    and report all errors found in the IHDR data.
fukasawa e60969
  Eliminated useless call to png_check_cHRM() from pngset.c
fukasawa e60969
fukasawa e60969
Version 1.4.0beta83 [September 25, 2009]
fukasawa e60969
  Revised png_check_IHDR() to eliminate bogus complaint about filter_type.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta84 [September 30, 2009]
fukasawa e60969
  Fixed some inconsistent indentation in pngconf.h
fukasawa e60969
  Revised png_check_IHDR() to add a test for width variable less than 32-bit.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta85 [October 1, 2009]
fukasawa e60969
  Revised png_check_IHDR() again, to check info_ptr members instead of
fukasawa e60969
    the contents of the returned parameters.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta86 [October 9, 2009]
fukasawa e60969
  Updated the "xcode" project (Alam Arias).
fukasawa e60969
  Eliminated a shadowed declaration of "pp" in png_handle_sPLT().
fukasawa e60969
fukasawa e60969
Version 1.4.0rc01 [October 19, 2009]
fukasawa e60969
  Trivial cosmetic changes.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta87 [October 30, 2009]
fukasawa e60969
  Moved version 1.4.0 back into beta.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta88 [October 30, 2009]
fukasawa e60969
  Revised libpng*.txt section about differences between 1.2.x and 1.4.0
fukasawa e60969
    because most of the new features have now been ported back to 1.2.41
fukasawa e60969
fukasawa e60969
Version 1.4.0beta89 [November 1, 2009]
fukasawa e60969
  More bugfixes and improvements to CMakeLists.txt (Philip Lowman)
fukasawa e60969
  Removed a harmless extra png_set_invert_alpha() from pngwrite.c
fukasawa e60969
  Apply png_user_chunk_cache_max within png_decompress_chunk().
fukasawa e60969
  Merged libpng-1.2.41.txt with libpng-1.4.0.txt where appropriate.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta90 [November 2, 2009]
fukasawa e60969
  Removed all remaining WIN32_WCE #ifdefs except those involving the
fukasawa e60969
    time.h "tm" structure
fukasawa e60969
fukasawa e60969
Version 1.4.0beta91 [November 3, 2009]
fukasawa e60969
  Updated scripts/pngw32.def and projects/wince/png32ce.def
fukasawa e60969
  Copied projects/wince/png32ce.def to the scripts directory.
fukasawa e60969
  Added scripts/makefile.wce
fukasawa e60969
  Patched ltmain.sh for wince support.
fukasawa e60969
  Added PNG_CONVERT_tIME_SUPPORTED macro.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta92 [November 4, 2009]
fukasawa e60969
  Make inclusion of time.h in pngconf.h depend on PNG_CONVERT_tIME_SUPPORTED
fukasawa e60969
  Make #define PNG_CONVERT_tIME_SUPPORTED depend on PNG_WRITE_tIME_SUPPORTED
fukasawa e60969
  Revised libpng*.txt to describe differences from 1.2.40 to 1.4.0 (instead
fukasawa e60969
    of differences from 1.2.41 to 1.4.0)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta93 [November 7, 2009]
fukasawa e60969
  Added PNG_DEPSTRUCT, PNG_DEPRECATED, PNG_USE_RESULT, PNG_NORETURN, and
fukasawa e60969
    PNG_ALLOCATED macros to detect deprecated direct access to the
fukasawa e60969
    png_struct or info_struct members and other deprecated usage in
fukasawa e60969
    applications (John Bowler).
fukasawa e60969
  Updated scripts/makefile* to add "-DPNG_CONFIGURE_LIBPNG" to CFLAGS,
fukasawa e60969
    to prevent warnings about direct access to png structs by libpng
fukasawa e60969
    functions while building libpng.  They need to be tested, especially
fukasawa e60969
    those using compilers other than gcc.
fukasawa e60969
  Updated projects/visualc6 and visualc71 with "/d PNG_CONFIGURE_LIBPNG".
fukasawa e60969
    They should work but still need to be updated to remove
fukasawa e60969
    references to pnggccrd.c or pngvcrd.c and ASM building.
fukasawa e60969
  Added README.txt to the beos, cbuilder5, netware, and xcode projects warning
fukasawa e60969
    that they need to be updated, to remove references to pnggccrd.c and
fukasawa e60969
    pngvcrd.c and to depend on pngpriv.h
fukasawa e60969
  Removed three direct references to read_info_ptr members in pngtest.c
fukasawa e60969
    that were detected by the new PNG_DEPSTRUCT macro.
fukasawa e60969
  Moved the png_debug macro definitions and the png_read_destroy(),
fukasawa e60969
    png_write_destroy() and png_far_to_near() prototypes from png.h
fukasawa e60969
    to pngpriv.h (John Bowler)
fukasawa e60969
  Moved the synopsis lines for png_read_destroy(), png_write_destroy()
fukasawa e60969
    png_debug(), png_debug1(), and png_debug2() from libpng.3 to libpngpf.3.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta94 [November 9, 2009]
fukasawa e60969
  Removed the obsolete, unused pnggccrd.c and pngvcrd.c files.
fukasawa e60969
  Updated CMakeLists.txt to add "-DPNG_CONFIGURE_LIBPNG" to the definitions.
fukasawa e60969
  Removed dependency of pngtest.o on pngpriv.h in the makefiles.
fukasawa e60969
  Only #define PNG_DEPSTRUCT, etc. in pngconf.h if not already defined.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta95 [November 10, 2009]
fukasawa e60969
  Changed png_check_sig() to !png_sig_cmp() in contrib programs.
fukasawa e60969
  Added -DPNG_CONFIGURE_LIBPNG to contrib/pngminm/*/makefile
fukasawa e60969
  Changed png_check_sig() to !png_sig_cmp() in contrib programs.
fukasawa e60969
  Corrected the png_get_IHDR() call in contrib/gregbook/readpng2.c
fukasawa e60969
  Changed pngminim/*/gather.sh to stop trying to remove pnggccrd.c and pngvcrd.c
fukasawa e60969
  Added dependency on pngpriv.h in contrib/pngminim/*/makefile
fukasawa e60969
fukasawa e60969
Version 1.4.0beta96 [November 12, 2009]
fukasawa e60969
  Renamed scripts/makefile.wce to scripts/makefile.cegcc
fukasawa e60969
  Revised Makefile.am to use libpng.sys while building libpng.so
fukasawa e60969
    so that only PNG_EXPORT functions are exported.
fukasawa e60969
  Removed the deprecated png_check_sig() function/macro.
fukasawa e60969
  Removed recently removed function names from scripts/*.def
fukasawa e60969
  Revised pngtest.png to put chunks in the same order written by pngtest
fukasawa e60969
    (evidently the same change made in libpng-1.0beta54 was lost).
fukasawa e60969
  Added PNG_PRIVATE macro definition in pngconf.h for possible future use.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta97 [November 13, 2009]
fukasawa e60969
  Restored pngtest.png to the libpng-1.4.0beta7 version.
fukasawa e60969
  Removed projects/beos and netware.txt; no one seems to be supporting them.
fukasawa e60969
  Revised Makefile.in
fukasawa e60969
fukasawa e60969
Version 1.4.0beta98 [November 13, 2009]
fukasawa e60969
  Added the "xcode" project to zip distributions,
fukasawa e60969
  Fixed a typo in scripts/pngwin.def introduced in beta97.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta99 [November 14, 2009]
fukasawa e60969
  Moved libpng-config.in and libpng.pc-configure.in out of the scripts
fukasawa e60969
    directory, to libpng-config.in and libpng-pc.in, respectively, and
fukasawa e60969
    modified Makefile.am and configure.ac accordingly.  Now "configure"
fukasawa e60969
    needs nothing from the "scripts" directory.
fukasawa e60969
  Avoid redefining PNG_CONST in pngconf.h
fukasawa e60969
fukasawa e60969
Version 1.4.0beta100 [November 14, 2009]
fukasawa e60969
  Removed ASM builds from projects/visualc6 and projects/visualc71
fukasawa e60969
  Removed scripts/makefile.nommx and makefile.vcawin32
fukasawa e60969
  Revised CMakeLists.txt to account for new location of libpng-config.in
fukasawa e60969
    and libpng-pc.in
fukasawa e60969
  Updated INSTALL to reflect removal and relocation of files.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta101 [November 14, 2009]
fukasawa e60969
  Restored the binary files (*.jpg, *.png, some project files) that were
fukasawa e60969
    accidentally deleted from the zip and 7z distributions when the xcode
fukasawa e60969
    project was added.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta102 [November 18, 2009]
fukasawa e60969
  Added libpng-config.in and libpng-pc.in to the zip and 7z distributions.
fukasawa e60969
  Fixed a typo in projects/visualc6/pngtest.dsp, introduced in beta100.
fukasawa e60969
  Moved descriptions of makefiles and other scripts out of INSTALL into
fukasawa e60969
    scripts/README.txt
fukasawa e60969
  Updated the copyright year in scripts/pngwin.rc from 2006 to 2009.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta103 [November 21, 2009]
fukasawa e60969
  Removed obsolete comments about ASM from projects/visualc71/README_zlib.txt
fukasawa e60969
  Align row_buf on 16-byte boundary in memory.
fukasawa e60969
  Restored the PNG_WRITE_FLUSH_AFTER_IEND_SUPPORTED guard around the call
fukasawa e60969
    to png_flush() after png_write_IEND().  See 1.4.0beta32, 1.4.0beta50
fukasawa e60969
    changes above and 1.2.30, 1.2.30rc01 and rc03 in 1.2.41 CHANGES.  Someone
fukasawa e60969
    needs this feature.
fukasawa e60969
  Make the 'png_jmpbuf' macro expand to a call that records the correct
fukasawa e60969
    longjmp function as well as returning a pointer to the setjmp
fukasawa e60969
    jmp_buf buffer, and marked direct access to jmpbuf 'deprecated'.
fukasawa e60969
    (John Bowler)
fukasawa e60969
fukasawa e60969
Version 1.4.0beta104 [November 22, 2009]
fukasawa e60969
  Removed png_longjmp_ptr from scripts/*.def and libpng.3
fukasawa e60969
  Rebuilt configure scripts with autoconf-2.65
fukasawa e60969
fukasawa e60969
Version 1.4.0beta105 [November 25, 2009]
fukasawa e60969
  Use fast integer PNG_DIVIDE_BY_255() or PNG_DIVIDE_BY_65535()
fukasawa e60969
    to accomplish alpha premultiplication when
fukasawa e60969
    PNG_READ_COMPOSITE_NODIV_SUPPORTED is defined.
fukasawa e60969
  Changed "/255" to "/255.0" in background calculations to make it clear
fukasawa e60969
    that the 255 is used as a double.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta106 [November 27, 2009]
fukasawa e60969
  Removed premultiplied alpha feature.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta107 [December 4, 2009]
fukasawa e60969
  Updated README
fukasawa e60969
  Added "#define PNG_NO_PEDANTIC_WARNINGS" in the libpng source files.
fukasawa e60969
  Removed "-DPNG_CONFIGURE_LIBPNG" from the makefiles and projects.
fukasawa e60969
  Revised scripts/makefile.netbsd, makefile.openbsd, and makefile.sco
fukasawa e60969
    to put png.h and pngconf.h in $prefix/include, like the other scripts,
fukasawa e60969
    instead of in $prefix/include/libpng.  Also revised makefile.sco
fukasawa e60969
    to put them in $prefix/include/libpng15 instead of in
fukasawa e60969
    $prefix/include/libpng/libpng15.
fukasawa e60969
fukasawa e60969
Version 1.4.0beta108 [December 11, 2009]
fukasawa e60969
  Removed leftover "-DPNG_CONFIGURE_LIBPNG" from contrib/pngminim/*/makefile
fukasawa e60969
  Relocated png_do_chop() to its original position in pngrtran.c; the
fukasawa e60969
    change in version 1.2.41beta08 caused transparency to be handled wrong
fukasawa e60969
    in some 16-bit datastreams (Yusaku Sugai).
fukasawa e60969
fukasawa e60969
Version 1.4.0beta109 [December 13, 2009]
fukasawa e60969
  Added "bit_depth" parameter to the private png_build_gamma_table() function.
fukasawa e60969
  Pass bit_depth=8 to png_build_gamma_table() when bit_depth is 16 but the
fukasawa e60969
    PNG_16_TO_8 transform has been set, to avoid unnecessary build of 16-bit
fukasawa e60969
    tables.
fukasawa e60969
fukasawa e60969
Version 1.4.0rc02 [December 20, 2009]
fukasawa e60969
  Declared png_cleanup_needed "volatile" in pngread.c and pngwrite.c
fukasawa e60969
fukasawa e60969
Version 1.4.0rc03 [December 22, 2009]
fukasawa e60969
  Renamed libpng-pc.in back to libpng.pc.in and revised CMakeLists.txt
fukasawa e60969
    (revising the change in 1.4.0beta99)
fukasawa e60969
fukasawa e60969
Version 1.4.0rc04 [December 25, 2009]
fukasawa e60969
  Swapped PNG_UNKNOWN_CHUNKS_SUPPORTED and PNG_HANDLE_AS_UNKNOWN_SUPPORTED
fukasawa e60969
    in pngset.c to be consistent with other changes in version 1.2.38.
fukasawa e60969
fukasawa e60969
Version 1.4.0rc05 [December 25, 2009]
fukasawa e60969
  Changed "libpng-pc.in" to "libpng.pc.in" in configure.ac, configure, and
fukasawa e60969
    Makefile.in to be consistent with changes in libpng-1.4.0rc03
fukasawa e60969
fukasawa e60969
Version 1.4.0rc06 [December 29, 2009]
fukasawa e60969
  Reverted the gamma_table changes from libpng-1.4.0beta109.
fukasawa e60969
  Fixed some indentation errors.
fukasawa e60969
fukasawa e60969
Version 1.4.0rc07 [January 1, 2010]
fukasawa e60969
  Revised libpng*.txt and libpng.3 about 1.2.x->1.4.x differences.
fukasawa e60969
  Use png_calloc() instead of png_malloc(); png_memset() in pngrutil.c
fukasawa e60969
  Update copyright year to 2010.
fukasawa e60969
fukasawa e60969
Version 1.4.0rc08 [January 2, 2010]
fukasawa e60969
  Avoid deprecated references to png_ptr-io_ptr and png_ptr->error_ptr
fukasawa e60969
    in pngtest.c
fukasawa e60969
fukasawa e60969
Version 1.4.0 [January 3, 2010]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta01 [January 8, 2010]
fukasawa e60969
  Updated CMakeLists.txt for consistent indentation and to avoid an
fukasawa e60969
    unclosed if-statement warning (Philip Lowman).
fukasawa e60969
  Revised Makefile.am and Makefile.in to remove references to Y2KINFO,
fukasawa e60969
    KNOWNBUG, and libpng.la (Robert Schwebel).
fukasawa e60969
  Revised the makefiles to install the same files and symbolic
fukasawa e60969
    links as configure, except for libpng.la and libpng14.la.
fukasawa e60969
  Make png_set|get_compression_buffer_size() available even when
fukasawa e60969
    PNG_WRITE_SUPPORTED is not enabled.
fukasawa e60969
  Revised Makefile.am and Makefile.in to simplify their maintenance.
fukasawa e60969
  Revised scripts/makefile.linux to install a link to libpng14.so.14.1
fukasawa e60969
fukasawa e60969
Version 1.4.1beta02 [January 9, 2010]
fukasawa e60969
  Revised the rest of the makefiles to install a link to libpng14.so.14.1
fukasawa e60969
fukasawa e60969
Version 1.4.1beta03 [January 10, 2010]
fukasawa e60969
  Removed png_set_premultiply_alpha() from scripts/*.def
fukasawa e60969
fukasawa e60969
Version 1.4.1rc01 [January 16, 2010]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta04 [January 23, 2010]
fukasawa e60969
  Revised png_decompress_chunk() to improve speed and memory usage when
fukasawa e60969
    decoding large chunks.
fukasawa e60969
  Added png_set|get_chunk_malloc_max() functions.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta05 [January 26, 2010]
fukasawa e60969
  Relocated "int k" declaration in pngtest.c to minimize its scope.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta06 [January 28, 2010]
fukasawa e60969
  Revised png_decompress_chunk() to use a two-pass method suggested by
fukasawa e60969
    John Bowler.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta07 [February 6, 2010]
fukasawa e60969
  Folded some long lines in the source files.
fukasawa e60969
  Added defineable PNG_USER_CHUNK_CACHE_MAX, PNG_USER_CHUNK_MALLOC_MAX,
fukasawa e60969
    and a PNG_USER_LIMITS_SUPPORTED flag.
fukasawa e60969
  Eliminated use of png_ptr->irowbytes and reused the slot in png_ptr as
fukasawa e60969
    png_ptr->png_user_chunk_malloc_max.
fukasawa e60969
  Revised png_push_save_buffer() to do fewer but larger png_malloc() calls.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta08 [February 6, 2010]
fukasawa e60969
  Minor cleanup and updating of dates and copyright year.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta01 [February 7, 2010]
fukasawa e60969
  Moved declaration of png_struct into private pngstruct.h and png_info
fukasawa e60969
    into pnginfo.h
fukasawa e60969
fukasawa e60969
Version 1.4.1beta09 and 1.5.0beta02 [February 7, 2010]
fukasawa e60969
  Reverted to original png_push_save_buffer() code.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta10 and 1.5.0beta03 [February 8, 2010]
fukasawa e60969
  Return allocated "old_buffer" in png_push_save_buffer() before
fukasawa e60969
    calling png_error(), to avoid a potential memory leak.
fukasawa e60969
  Updated configure script to use SO number 15.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta04 [February 9, 2010]
fukasawa e60969
  Removed malformed "incomplete struct declaration" of png_info from png.h
fukasawa e60969
fukasawa e60969
Version 1.5.0beta05 [February 12, 2010]
fukasawa e60969
  Removed PNG_DEPSTRUCT markup in pngstruct.h and pnginfo.h, and undid the
fukasawa e60969
    linewrapping that it entailed.
fukasawa e60969
  Revised comments in pngstruct.h and pnginfo.h and added pointers to
fukasawa e60969
    the libpng license.
fukasawa e60969
  Changed PNG_INTERNAL to PNG_EXPOSE_INTERNAL_STRUCTURES
fukasawa e60969
  Removed the cbuilder5 project, which has not been updated to 1.4.0.
fukasawa e60969
fukasawa e60969
Version 1.4.1beta12 and 1.5.0beta06 [February 14, 2010]
fukasawa e60969
  Fixed type declaration of png_get_chunk_malloc_max() in pngget.c (Daisuke
fukasawa e60969
    Nishikawa)
fukasawa e60969
fukasawa e60969
Version 1.5.0beta07 [omitted]
fukasawa e60969
fukasawa e60969
Version 1.5.0beta08 [February 19, 2010]
fukasawa e60969
  Changed #ifdef PNG_NO_STDIO_SUPPORTED to #ifdef PNG_NO_CONSOLE_IO_SUPPORTED
fukasawa e60969
    wherever png_snprintf() is used to construct error and warning messages.
fukasawa e60969
  Noted in scripts/makefile.mingw that it expects to be run under MSYS.
fukasawa e60969
  Removed obsolete unused MMX-querying support from contrib/gregbook
fukasawa e60969
  Added exported png_longjmp() function.
fukasawa e60969
  Removed the AIX redefinition of jmpbuf in png.h
fukasawa e60969
  Added -D_ALLSOURCE in configure.ac, makefile.aix, and CMakeLists.txt
fukasawa e60969
    when building on AIX.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta09 [February 19, 2010]
fukasawa e60969
  Removed -D_ALLSOURCE from configure.ac, makefile.aix, and CMakeLists.txt.
fukasawa e60969
  Changed the name of png_ptr->jmpbuf to png_ptr->png_jmpbuf in pngstruct.h
fukasawa e60969
fukasawa e60969
Version 1.5.0beta10 [February 25, 2010]
fukasawa e60969
  Removed unused gzio.c from contrib/pngminim gather and makefile scripts
fukasawa e60969
  Removed replacement error handlers from contrib/gregbook.  Because of
fukasawa e60969
    the new png_longjmp() function they are no longer needed.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta11 [March 6, 2010]
fukasawa e60969
  Removed checking for already-included setjmp.h from pngconf.h
fukasawa e60969
  Fixed inconsistent indentations and made numerous cosmetic changes.
fukasawa e60969
  Revised the "SEE ALSO" style of libpng.3, libpngpf.3, and png.5
fukasawa e60969
fukasawa e60969
Version 1.5.0beta12 [March 9, 2010]
fukasawa e60969
  Moved "#include png.h" inside pngpriv.h and removed "#include png.h" from
fukasawa e60969
    the source files, along with "#define PNG_EXPOSE_INTERNAL_STRUCTURES"
fukasawa e60969
    and "#define PNG_NO_PEDANTIC_WARNINGS" (John Bowler).
fukasawa e60969
  Created new pngdebug.h and moved debug definitions there.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta13 [March 10, 2010]
fukasawa e60969
  Protect pngstruct.h, pnginfo.h, and pngdebug.h from being included twice.
fukasawa e60969
  Revise the "#ifdef" blocks in png_inflate() so it will compile when neither
fukasawa e60969
    PNG_USER_CHUNK_MALLOC_MAX nor PNG_SET_CHUNK_MALLOC_LIMIT_SUPPORTED
fukasawa e60969
    is defined.
fukasawa e60969
  Removed unused png_measure_compressed_chunk() from pngpriv.h and libpngpf.3
fukasawa e60969
  Moved the 'config.h' support from pngconf.h to pngpriv.h
fukasawa e60969
  Removed PNGAPI from the png_longjmp_ptr typedef.
fukasawa e60969
  Eliminated dependence of pngtest.c on the private pngdebug.h file.
fukasawa e60969
  Make all png_debug macros into *unterminated* statements or
fukasawa e60969
    expressions (i.e. a trailing ';' must always be added) and correct
fukasawa e60969
    the format statements in various png_debug messages.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta14 [March 14, 2010]
fukasawa e60969
  Removed direct access to png_ptr->io_ptr from the Windows code in pngtest.c
fukasawa e60969
  Revised Makefile.am to account for recent additions and replacements.
fukasawa e60969
  Corrected CE and OS/2 DEF files (scripts/png*def) for symbols removed and
fukasawa e60969
    added ordinal numbers to the Windows DEF file and corrected the duplicated
fukasawa e60969
    ordinal numbers on CE symbols that are commented out.
fukasawa e60969
  Added back in export symbols that can be present in the Windows build but
fukasawa e60969
    are disabled by default.
fukasawa e60969
  PNG_EXPORT changed to include an 'ordinal' field for DEF file generation.
fukasawa e60969
    PNG_CALLBACK added to make callback definitions uniform.  PNGAPI split
fukasawa e60969
    into PNGCAPI (base C form), PNGAPI (exports) and PNGCBAPI (callbacks),
fukasawa e60969
    and appropriate changes made to all files.  Cygwin builds re-hinged to
fukasawa e60969
    allow procedure call standard changes and to remove the need for the DEF
fukasawa e60969
    file (fixes build on Cygwin).
fukasawa e60969
  Enabled 'attribute' warnings that are relevant to library APIs and callbacks.
fukasawa e60969
  Changed rules for generation of the various symbol files and added a new
fukasawa e60969
    rule for a DEF file (which is also added to the distribution).
fukasawa e60969
  Updated the symbol file generation to stop it adding spurious spaces
fukasawa e60969
    to EOL (coming from preprocessor macro expansion).  Added a facility
fukasawa e60969
    to join tokens in the output and rewrite *.dfn to use this.
fukasawa e60969
  Eliminated scripts/*.def in favor of libpng.def; updated projects/visualc71
fukasawa e60969
    and removed scripts/makefile.cygwin.
fukasawa e60969
  Made PNG_BUILD_DLL safe: it can be set whenever a DLL is being built.
fukasawa e60969
  Removed the include of sys/types.h - apparently unnecessary now on the
fukasawa e60969
    platforms on which it happened (all but Mac OS and RISC OS).
fukasawa e60969
  Moved the Mac OS test into pngpriv.h (the only place it is used.)
fukasawa e60969
fukasawa e60969
Version 1.5.0beta15 [March 17, 2010]
fukasawa e60969
  Added symbols.chk target to Makefile.am to validate the symbols in png.h
fukasawa e60969
    against the new DEF file scripts/symbols.def.
fukasawa e60969
  Changed the default DEF file back to pngwin.def.
fukasawa e60969
  Removed makefile.mingw.
fukasawa e60969
  Eliminated PNG_NO_EXTERN and PNG_ALL_EXTERN
fukasawa e60969
fukasawa e60969
Version 1.5.0beta16 [April 1, 2010]
fukasawa e60969
  Make png_text_struct independent of PNG_iTXt_SUPPORTED, so that
fukasawa e60969
    fields are initialized in all configurations.  The READ/WRITE
fukasawa e60969
    macros (PNG_(READ|WRITE)_iTXt_SUPPORTED) still function as
fukasawa e60969
    before to disable code to actually read or write iTXt chunks
fukasawa e60969
    and iTXt_SUPPORTED can be used to detect presence of either
fukasawa e60969
    read or write support (but it is probably better to check for
fukasawa e60969
    the one actually required - read or write.)
fukasawa e60969
  Combined multiple png_warning() calls for a single error.
fukasawa e60969
  Restored the macro definition of png_check_sig().
fukasawa e60969
fukasawa e60969
Version 1.5.0beta17 [April 17, 2010]
fukasawa e60969
  Added some "(long)" typecasts to printf calls in png_handle_cHRM().
fukasawa e60969
  Documented the fact that png_set_dither() was disabled since libpng-1.4.0.
fukasawa e60969
  Reenabled png_set_dither() but renamed it to png_set_quantize() to reflect
fukasawa e60969
    more accurately what it actually does.  At the same time, renamed
fukasawa e60969
    the PNG_DITHER_[RED,GREEN_BLUE]_BITS macros to
fukasawa e60969
    PNG_QUANTIZE_[RED,GREEN,BLUE]_BITS.
fukasawa e60969
  Added some "(long)" typecasts to printf calls in png_handle_cHRM().
fukasawa e60969
  Freeze build-time only configuration in the build.
fukasawa e60969
    In all prior versions of libpng most configuration options
fukasawa e60969
    controlled by compiler #defines had to be repeated by the
fukasawa e60969
    application code that used libpng.  This patch changes this
fukasawa e60969
    so that compilation options that can only be changed at build
fukasawa e60969
    time are frozen in the build.  Options that are compiler
fukasawa e60969
    dependent (and those that are system dependent) are evaluated
fukasawa e60969
    each time - pngconf.h holds these.  Options that can be changed
fukasawa e60969
    per-file in the application are in png.h.  Frozen options are
fukasawa e60969
    in the new installed header file pnglibconf.h (John Bowler)
fukasawa e60969
  Removed the xcode project because it has not been updated to work
fukasawa e60969
    with libpng-1.5.0.
fukasawa e60969
  Removed the ability to include optional pngusr.h
fukasawa e60969
fukasawa e60969
Version 1.5.0beta18 [April 17, 2010]
fukasawa e60969
  Restored the ability to include optional pngusr.h
fukasawa e60969
  Moved replacements for png_error() and png_warning() from the
fukasawa e60969
    contrib/pngminim project to pngerror.c, for use when warnings or
fukasawa e60969
    errors are disabled via PNG_NO_WARN or PNG_NO_ERROR_TEXT, to avoid
fukasawa e60969
    storing unneeded error/warning text.
fukasawa e60969
  Updated contrib/pngminim project to work with the new pnglibconf.h
fukasawa e60969
  Added some PNG_NO_* defines to contrib/pngminim/*/pngusr.h to save space.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta19 [April 24, 2010]
fukasawa e60969
  Added PNG_{READ,WRITE}_INT_FUNCTIONS_SUPPORTED.  This allows the functions
fukasawa e60969
    to read and write ints to be disabled independently of PNG_USE_READ_MACROS,
fukasawa e60969
    which allows libpng to be built with the functions even though the default
fukasawa e60969
    is to use the macros - this allows applications to choose at app build
fukasawa e60969
    time whether or not to use macros (previously impossible because the
fukasawa e60969
    functions weren't in the default build.)
fukasawa e60969
  Changed Windows calling convention back to __cdecl for API functions.
fukasawa e60969
    For Windows/x86 platforms only:
fukasawa e60969
      __stdcall is no longer needed for Visual Basic, so libpng-1.5.0 uses
fukasawa e60969
      __cdecl throughout (both API functions and callbacks) on Windows/x86
fukasawa e60969
      platforms.
fukasawa e60969
  Replaced visualc6 and visualc71 projects with new vstudio project
fukasawa e60969
  Relaxed the overly-restrictive permissions of some files.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta20 [April 24, 2010]
fukasawa e60969
  Relaxed more overly-restrictive permissions of some files.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta21 [April 27, 2010]
fukasawa e60969
  Removed some unwanted binary bytes and changed CRLF to NEWLINE in the new
fukasawa e60969
    vstudio project files, and some trivial editing of some files in the
fukasawa e60969
    scripts directory.
fukasawa e60969
  Set PNG_NO_READ_BGR, PNG_NO_IO_STATE, and PNG_NO_TIME_RFC1123 in
fukasawa e60969
    contrib/pngminim/decoder/pngusr.h to make a smaller decoder application.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta22 [April 28, 2010]
fukasawa e60969
  Fixed dependencies of GET_INT_32 - it does not require READ_INT_FUNCTIONS
fukasawa e60969
    because it has a macro equivalent.
fukasawa e60969
  Improved the options.awk script; added an "everything off" option.
fukasawa e60969
  Revised contrib/pngminim to use the "everything off" option in pngusr.dfa.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta23 [April 29, 2010]
fukasawa e60969
  Corrected PNG_REMOVED macro to take five arguments.
fukasawa e60969
    The macro was documented with two arguments (name,ordinal), however
fukasawa e60969
    the symbol checking .dfn files assumed five arguments.  The five
fukasawa e60969
    argument form seems more useful so it is changed to that.
fukasawa e60969
  Corrected PNG_UNKNOWN_CHUNKS_SUPPORTED to PNG_HANDLE_AS_UNKNOWN_SUPPORTED
fukasawa e60969
    in gregbook/readpng2.c
fukasawa e60969
  Corrected protection of png_get_user_transform_ptr. The API declaration in
fukasawa e60969
    png.h is removed if both READ and WRITE USER_TRANSFORM are turned off
fukasawa e60969
    but was left defined in pngtrans.c
fukasawa e60969
  Added logunsupported=1 to cause pnglibconf.h to document disabled options.
fukasawa e60969
    This makes the installed pnglibconf.h more readable but causes no
fukasawa e60969
    other change.  The intention is that users of libpng will find it
fukasawa e60969
    easier to understand if an API they need is missing.
fukasawa e60969
  Include png_reset_zstream() in png.c only when PNG_READ_SUPPORTED is defined.
fukasawa e60969
  Removed dummy_inflate.c from contrib/pngminim/encoder
fukasawa e60969
  Removed contrib/pngminim/*/gather.sh; gathering is now done in the makefile.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta24 [May 7, 2010]
fukasawa e60969
  Use bitwise "&" instead of arithmetic mod in pngrutil.c calculation of the
fukasawa e60969
    offset of the png_ptr->rowbuf pointer into png_ptr->big_row_buf.
fukasawa e60969
  Added more blank lines for readability.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta25 [June 18, 2010]
fukasawa e60969
  In pngpread.c: png_push_have_row() add check for new_row > height
fukasawa e60969
  Removed the now-redundant check for out-of-bounds new_row from example.c
fukasawa e60969
fukasawa e60969
Version 1.5.0beta26 [June 18, 2010]
fukasawa e60969
  In pngpread.c: png_push_process_row() add check for too many rows.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta27 [June 18, 2010]
fukasawa e60969
  Removed the check added in beta25 as it is now redundant.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta28 [June 20, 2010]
fukasawa e60969
  Rewrote png_process_IDAT_data to consistently treat extra data as warnings
fukasawa e60969
    and handle end conditions more cleanly.
fukasawa e60969
  Removed the new (beta26) check in png_push_process_row().
fukasawa e60969
fukasawa e60969
Version 1.5.0beta29 [June 21, 2010]
fukasawa e60969
  Revised scripts/options.awk to work on Sunos (but still doesn't work)
fukasawa e60969
  Added comment to options.awk and contrib/pngminim/*/makefile to try nawk.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta30 [June 22, 2010]
fukasawa e60969
  Stop memory leak when reading a malformed sCAL chunk.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta31 [June 26, 2010]
fukasawa e60969
  Revised pngpread.c patch of beta28 to avoid an endless loop.
fukasawa e60969
  Removed some trailing blanks.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta32 [June 26, 2010]
fukasawa e60969
  Removed leftover scripts/options.patch and scripts/options.rej
fukasawa e60969
fukasawa e60969
Version 1.5.0beta33 [July 6, 3010]
fukasawa e60969
  Made FIXED and FLOATING options consistent in the APIs they enable and
fukasawa e60969
    disable.  Corrected scripts/options.awk to handle both command line
fukasawa e60969
    options and options specified in the .dfa files.
fukasawa e60969
  Changed char *msg to PNG_CONST char *msg in pngrutil.c
fukasawa e60969
  Make png_set_sRGB_gAMA_and_cHRM set values using either the fixed or
fukasawa e60969
    floating point APIs, but not both.
fukasawa e60969
  Reversed patch to remove error handler when the jmp_buf is stored in the
fukasawa e60969
    main program structure, not the png_struct.
fukasawa e60969
    The error handler is needed because the default handler in libpng will
fukasawa e60969
    always use the jmp_buf in the library control structure; this is never
fukasawa e60969
    set.  The gregbook code is a useful example because, even though it
fukasawa e60969
    uses setjmp/longjmp, it shows how error handling can be implemented
fukasawa e60969
    using control mechanisms not directly supported by libpng.  The
fukasawa e60969
    technique will work correctly with mechanisms such as Microsoft
fukasawa e60969
    Structure Exceptions or C++ exceptions (compiler willing - note that gcc
fukasawa e60969
    does not by default support interworking of C and C++ error handling.)
fukasawa e60969
  Reverted changes to call png_longjmp in contrib/gregbook where it is not
fukasawa e60969
    appropriate.  If mainprog->jmpbuf is used by setjmp, then png_longjmp
fukasawa e60969
    cannot be used.
fukasawa e60969
  Changed "extern PNG_EXPORT" to "PNG_EXPORT" in png.h (Jan Nijtmans)
fukasawa e60969
  Changed "extern" to "PNG_EXTERN" in pngpriv.h (except for the 'extern "C" {')
fukasawa e60969
fukasawa e60969
Version 1.5.0beta34 [July 12, 2010]
fukasawa e60969
  Put #ifndef PNG_EXTERN, #endif around the define PNG_EXTERN in pngpriv.h
fukasawa e60969
fukasawa e60969
Version 1.5.0beta35 [July 24, 2010]
fukasawa e60969
  Removed some newly-added TAB characters.
fukasawa e60969
  Added -DNO_PNG_SNPRINTF to CFLAGS in scripts/makefile.dj2
fukasawa e60969
  Moved the definition of png_snprintf() outside of the enclosing
fukasawa e60969
    #ifdef blocks in pngconf.h
fukasawa e60969
fukasawa e60969
Version 1.5.0beta36 [July 29, 2010]
fukasawa e60969
  Patches by John Bowler:
fukasawa e60969
  Fixed point APIs are now supported throughout (no missing APIs).
fukasawa e60969
  Internal fixed point arithmetic support exists for all internal floating
fukasawa e60969
    point operations.
fukasawa e60969
  sCAL validates the floating point strings it is passed.
fukasawa e60969
  Safe, albeit rudimentary, Watcom support is provided by PNG_API_RULE==2
fukasawa e60969
  Two new APIs exist to get the number of passes without turning on the
fukasawa e60969
    PNG_INTERLACE transform and to get the number of rows in the current
fukasawa e60969
    pass.
fukasawa e60969
  A new test program, pngvalid.c, validates the gamma code.
fukasawa e60969
  Errors in the 16-bit gamma correction (overflows) have been corrected.
fukasawa e60969
  cHRM chunk testing is done consistently (previously the floating point
fukasawa e60969
    API bypassed it, because the test really didn't work on FP, now the test
fukasawa e60969
    is performed on the actual values to be stored in the PNG file so it
fukasawa e60969
    works in the FP case too.)
fukasawa e60969
  Most floating point APIs now simply call the fixed point APIs after
fukasawa e60969
    converting the values to the fixed point form used in the PNG file.
fukasawa e60969
  The standard headers no longer include zlib.h, which is currently only
fukasawa e60969
    required for pngstruct.h and can therefore be internal.
fukasawa e60969
  Revised png_get_int_32 to undo the PNG two's complement representation of
fukasawa e60969
    negative numbers.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta37 [July 30, 2010]
fukasawa e60969
  Added a typecast in png_get_int_32() in png.h and pngrutil.h to avoid
fukasawa e60969
    a compiler warning.
fukasawa e60969
  Replaced oFFs 0,0 with oFFs -10,20 in pngtest.png
fukasawa e60969
fukasawa e60969
Version 1.5.0beta38 [July 31, 2010]
fukasawa e60969
  Implemented remaining "_fixed" functions.
fukasawa e60969
  Corrected a number of recently introduced warnings mostly resulting from
fukasawa e60969
    safe but uncast assignments to shorter integers.  Also added a zlib
fukasawa e60969
    VStudio release library project because the latest zlib Official Windows
fukasawa e60969
    build does not include such a thing.
fukasawa e60969
  Revised png_get_int_16() to be similar to png_get_int_32().
fukasawa e60969
  Restored projects/visualc71.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta39 [August 2, 2010]
fukasawa e60969
  VisualC/GCC warning fixes, VisualC build fixes
fukasawa e60969
  The changes include support for function attributes in VC in addition to
fukasawa e60969
    those already present in GCC - necessary because without these some
fukasawa e60969
    warnings are unavoidable.  Fixes include signed/unsigned fixes in
fukasawa e60969
    pngvalid and checks with gcc -Wall -Wextra -Wunused.
fukasawa e60969
  VC requires function attributes on function definitions as well as
fukasawa e60969
    declarations, PNG_FUNCTION has been added to enable this and the
fukasawa e60969
    relevant function definitions changed.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta40 [August 6, 2010]
fukasawa e60969
  Correct use of _WINDOWS_ in pngconf.h
fukasawa e60969
  Removed png_mem_ #defines; they are no longer used.
fukasawa e60969
  Added the sRGB chunk to pngtest.png
fukasawa e60969
fukasawa e60969
Version 1.5.0beta41 [August 11, 2010]
fukasawa e60969
  Added the cHRM chunk to pngtest.png
fukasawa e60969
  Don't try to use version-script with cygwin/mingw.
fukasawa e60969
  Revised contrib/gregbook to work under cygwin/mingw.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta42 [August 18, 2010]
fukasawa e60969
  Add .dll.a to the list of extensions to be symlinked by Makefile.am (Yaakov)
fukasawa e60969
  Made all API functions that have const arguments and constant string
fukasawa e60969
    literal pointers declare them (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.5.0beta43 [August 20, 2010]
fukasawa e60969
  Removed spurious tabs, shorten long lines (no source change)
fukasawa e60969
    Also added scripts/chkfmt to validate the format of all the files that can
fukasawa e60969
    reasonably be validated (it is suggested to run "make distclean" before
fukasawa e60969
    checking, because some machine generated files have long lines.)
fukasawa e60969
  Reformatted the CHANGES file to be more consistent throughout.
fukasawa e60969
  Made changes to address various issues identified by GCC, mostly
fukasawa e60969
    signed/unsigned and shortening problems on assignment but also a few
fukasawa e60969
    difficult to optimize (for GCC) loops.
fukasawa e60969
  Fixed non-GCC fixed point builds.  In png.c a declaration was misplaced
fukasawa e60969
    in an earlier update.  Fixed to declare the auto variables at the head.
fukasawa e60969
  Use cexcept.h in pngvalid.c.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta44 [August 24, 2010]
fukasawa e60969
  Updated CMakeLists.txt to use CMAKE_INSTALL_LIBDIR variable; useful for
fukasawa e60969
    installing libpng in /usr/lib64 (Funda Wang).
fukasawa e60969
  Revised CMakeLists.txt to put the man pages in share/man/man* not man/man*
fukasawa e60969
  Revised CMakeLists.txt to make symlinks instead of copies when installing.
fukasawa e60969
  Changed PNG_LIB_NAME from pngNN to libpngNN in CMakeLists.txt (Philip Lowman)
fukasawa e60969
  Implemented memory checks within pngvalid
fukasawa e60969
  Reformatted/rearranged pngvalid.c to assist use of progressive reader.
fukasawa e60969
  Check interlaced images in pngvalid
fukasawa e60969
  Clarified pngusr.h comments in pnglibconf.dfa
fukasawa e60969
  Simplified the pngvalid error-handling code now that cexcept.h is in place.
fukasawa e60969
  Implemented progressive reader in pngvalid.c for standard tests
fukasawa e60969
  Implemented progressive read in pngvalid.c gamma tests
fukasawa e60969
  Turn on progressive reader in pngvalid.c by default and tidy code.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta45 [August 26, 2010]
fukasawa e60969
  Added an explicit make step to projects/vstudio for pnglibconf.h
fukasawa e60969
    Also corrected zlib.vcxproj into which Visual Studio had introduced
fukasawa e60969
    what it calls an "authoring error".  The change to make pnglibconf.h
fukasawa e60969
    simply copies the file; in the future it may actually generate the
fukasawa e60969
    file from scripts/pnglibconf.dfa as the other build systems do.
fukasawa e60969
  Changed pngvalid to work when floating point APIs are disabled
fukasawa e60969
  Renamed the prebuilt scripts/pnglibconf.h to scripts/pnglibconf.h.prebuilt
fukasawa e60969
  Supply default values for PNG_USER_PRIVATEBUILD and PNG_USER_DLLFNAME_POSTFIX
fukasawa e60969
    in pngpriv.h in case the user neglected to define them in their pngusr.h
fukasawa e60969
fukasawa e60969
Version 1.5.0beta46 [August 28, 2010]
fukasawa e60969
  Added new private header files to libpng_sources in CMakeLists.txt
fukasawa e60969
  Added PNG_READ_16BIT, PNG_WRITE_16BIT, and PNG_16BIT options.
fukasawa e60969
  Added reference to scripts/pnglibconf.h.prebuilt in the visualc71 project.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta47 [September 11, 2010]
fukasawa e60969
  Fixed a number of problems with 64-bit compilation reported by Visual
fukasawa e60969
    Studio 2010 (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.5.0beta48 [October 4, 2010]
fukasawa e60969
  Updated CMakeLists.txt (Philip Lowman).
fukasawa e60969
  Revised autogen.sh to recognize and use $AUTOCONF, $AUTOMAKE, $AUTOHEADER,
fukasawa e60969
    $AUTOPOINT, $ACLOCAL and $LIBTOOLIZE
fukasawa e60969
  Fixed problem with symbols creation in Makefile.am which was assuming that
fukasawa e60969
    all versions of ccp write to standard output by default (Martin Banky). The
fukasawa e60969
    bug was introduced in libpng-1.2.9beta5.
fukasawa e60969
  Removed unused mkinstalldirs.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta49 [October 8, 2010]
fukasawa e60969
  Undid Makefile.am revision of 1.5.0beta48.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta50 [October 14, 2010]
fukasawa e60969
  Revised Makefile.in to account for mkinstalldirs being removed.
fukasawa e60969
  Added some "(unsigned long)" typecasts in printf statements in pngvalid.c.
fukasawa e60969
  Suppressed a compiler warning in png_handle_sPLT().
fukasawa e60969
  Check for out-of-range text compression mode in png_set_text().
fukasawa e60969
fukasawa e60969
Version 1.5.0beta51 [October 15, 2010]
fukasawa e60969
  Changed embedded dates to "(PENDING RELEASE) in beta releases (and future
fukasawa e60969
    rc releases) to minimize the difference between releases.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta52 [October 16, 2010]
fukasawa e60969
  Restored some of the embedded dates (in png.h, png.c, documentation, etc.)
fukasawa e60969
fukasawa e60969
Version 1.5.0beta53 [October 18, 2010]
fukasawa e60969
  Updated INSTALL to mention using "make maintainer-clean" and to remove
fukasawa e60969
    obsolete statement about a custom ltmain.sh
fukasawa e60969
  Disabled "color-tests" by default in Makefile.am so it will work with
fukasawa e60969
    automake versions earlier than 1.11.1
fukasawa e60969
  Use document name "libpng-manual.txt" instead of "libpng-<version>.txt"
fukasawa e60969
    to simplify version differences.
fukasawa e60969
  Removed obsolete remarks about setjmp handling from INSTALL.
fukasawa e60969
  Revised and renamed the typedef in png.h and png.c that was designed
fukasawa e60969
    to catch library and header mismatch.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta54 [November 10, 2010]
fukasawa e60969
  Require 48 bytes, not 64 bytes, for big_row_buf in overflow checks.
fukasawa e60969
  Used a consistent structure for the pngget.c functions.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta55 [November 21, 2010]
fukasawa e60969
  Revised png_get_uint_32, png_get_int_32, png_get_uint_16 (Cosmin)
fukasawa e60969
  Moved reading of file signature into png_read_sig (Cosmin)
fukasawa e60969
  Fixed atomicity of chunk header serialization (Cosmin)
fukasawa e60969
  Added test for io_state in pngtest.c (Cosmin)
fukasawa e60969
  Added "#!/bin/sh" at the top of contrib/pngminim/*/gather.sh scripts.
fukasawa e60969
  Changes to remove gcc warnings (John Bowler)
fukasawa e60969
    Certain optional gcc warning flags resulted in warnings in libpng code.
fukasawa e60969
    With these changes only -Wconversion and -Wcast-qual cannot be turned on.
fukasawa e60969
    Changes are trivial rearrangements of code.  -Wconversion is not possible
fukasawa e60969
    for pngrutil.c (because of the widespread use of += et al on variables
fukasawa e60969
    smaller than (int) or (unsigned int)) and -Wcast-qual is not possible
fukasawa e60969
    with pngwio.c and pngwutil.c because the 'write' callback and zlib
fukasawa e60969
    compression both fail to declare their input buffers with 'const'.
fukasawa e60969
fukasawa e60969
Version 1.5.0beta56 [December 7, 2010]
fukasawa e60969
  Added the private PNG_UNUSED() macro definition in pngpriv.h.
fukasawa e60969
  Added some commentary about PNG_EXPORT in png.h and pngconf.h
fukasawa e60969
  Revised PNG_EXPORT() macro and added PNG_EXPORTA() macro, with the
fukasawa e60969
    objective of simplifying and improving the cosmetic appearance of png.h.
fukasawa e60969
  Fixed some incorrect "=" macro names in pnglibconf.dfa
fukasawa e60969
  Included documentation of changes in 1.5.0 from 1.4.x in libpng-manual.txt
fukasawa e60969
fukasawa e60969
Version 1.5.0beta57 [December 9, 2010]
fukasawa e60969
  Documented the pngvalid gamma error summary with additional comments and
fukasawa e60969
    print statements.
fukasawa e60969
  Improved missing symbol handling in checksym.awk; symbols missing in both
fukasawa e60969
    the old and new files can now be optionally ignored, treated as errors
fukasawa e60969
    or warnings.
fukasawa e60969
  Removed references to pngvcrd.c and pnggccrd.c from the vstudio project.
fukasawa e60969
  Updated "libpng14" to "libpng15" in the visualc71 project.
fukasawa e60969
  Enabled the strip16 tests in pngvalid.`
fukasawa e60969
  Don't display test results (except PASS/FAIL) when running "make test".
fukasawa e60969
    Instead put them in pngtest-log.txt
fukasawa e60969
  Added "--with-zprefix=<string>" to configure.ac
fukasawa e60969
  Updated the prebuilt configuration files to autoconf version 2.68
fukasawa e60969
fukasawa e60969
Version 1.5.0beta58 [December 19, 2010]
fukasawa e60969
  Fixed interlace image handling and add test cases (John Bowler)
fukasawa e60969
  Fixed the clean rule in Makefile.am to remove pngtest-log.txt
fukasawa e60969
  Made minor changes to work around warnings in gcc 3.4
fukasawa e60969
fukasawa e60969
Version 1.5.0rc01 [December 27, 2010]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.0rc02 [December 27, 2010]
fukasawa e60969
  Eliminated references to the scripts/*.def files in project/visualc71.
fukasawa e60969
fukasawa e60969
Version 1.5.0rc03 [December 28, 2010]
fukasawa e60969
  Eliminated scripts/*.def and revised Makefile.am accordingly
fukasawa e60969
fukasawa e60969
Version 1.5.0rc04 [December 29, 2010]
fukasawa e60969
  Fixed bug in background transformation handling in pngrtran.c (it was
fukasawa e60969
    looking for the flag in png_ptr->transformations instead of in
fukasawa e60969
    png_ptr->flags) (David Raymond).
fukasawa e60969
fukasawa e60969
Version 1.5.0rc05 [December 31, 2010]
fukasawa e60969
  Fixed typo in a comment in CMakeLists.txt (libpng14 => libpng15) (Cosmin)
fukasawa e60969
fukasawa e60969
Version 1.5.0rc06 [January 4, 2011]
fukasawa e60969
  Changed the new configure option "zprefix=string" to "zlib-prefix=string"
fukasawa e60969
fukasawa e60969
Version 1.5.0rc07 [January 4, 2011]
fukasawa e60969
  Updated copyright year.
fukasawa e60969
fukasawa e60969
Version 1.5.0 [January 6, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
version 1.5.1beta01 [January 8, 2011]
fukasawa e60969
  Added description of png_set_crc_action() to the manual.
fukasawa e60969
  Added a note in the manual that the type of the iCCP profile was changed
fukasawa e60969
    from png_charpp to png_bytepp in png_get_iCCP().  This change happened
fukasawa e60969
    in version 1.5.0beta36 but is not noted in the CHANGES.  Similarly,
fukasawa e60969
    it was changed from png_charpp to png_const_bytepp in png_set_iCCP().
fukasawa e60969
  Ensure that png_rgb_to_gray ignores palette mapped images, if libpng
fukasawa e60969
    internally happens to call it with one, and fixed a failure to handle
fukasawa e60969
    palette mapped images correctly.  This fixes CVE-2690.
fukasawa e60969
fukasawa e60969
Version 1.5.1beta02 [January 14, 2011]
fukasawa e60969
  Fixed a bug in handling of interlaced images (bero at arklinux.org).
fukasawa e60969
  Updated CMakeLists.txt (Clifford Yapp)
fukasawa e60969
fukasawa e60969
Version 1.5.1beta03 [January 14, 2011]
fukasawa e60969
  Fixed typecasting of some png_debug() statements (Cosmin)
fukasawa e60969
fukasawa e60969
Version 1.5.1beta04 [January 16, 2011]
fukasawa e60969
  Updated documentation of png_set|get_tRNS() (Thomas Klausner).
fukasawa e60969
  Mentioned in the documentation that applications must #include "zlib.h"
fukasawa e60969
    if they need access to anything in zlib.h, and that a number of
fukasawa e60969
    macros such as png_memset() are no longer accessible by applications.
fukasawa e60969
  Corrected pngvalid gamma test "sample" function to access all of the color
fukasawa e60969
    samples of each pixel, instead of sampling the red channel three times.
fukasawa e60969
  Prefixed variable names index, div, exp, gamma with "png_" to avoid "shadow"
fukasawa e60969
    warnings, and (mistakenly) changed png_exp() to exp().
fukasawa e60969
fukasawa e60969
Version 1.5.1beta05 [January 16, 2011]
fukasawa e60969
  Changed variable names png_index, png_div, png_exp, and png_gamma to
fukasawa e60969
    char_index, divisor, exp_b10, and gamma_val, respectively, and
fukasawa e60969
    changed exp() back to png_exp().
fukasawa e60969
fukasawa e60969
Version 1.5.1beta06 [January 20, 2011]
fukasawa e60969
  Prevent png_push_crc_skip() from hanging while reading an unknown chunk
fukasawa e60969
    or an over-large compressed zTXt chunk with the progressive reader.
fukasawa e60969
  Eliminated more GCC "shadow" warnings.
fukasawa e60969
  Revised png_fixed() in png.c to avoid compiler warning about reaching the
fukasawa e60969
    end without returning anything.
fukasawa e60969
fukasawa e60969
Version 1.5.1beta07 [January 22, 2011]
fukasawa e60969
  In the manual, describe the png_get_IHDR() arguments in the correct order.
fukasawa e60969
  Added const_png_structp and const_png_infop types, and used them in
fukasawa e60969
    prototypes for most png_get_*() functions.
fukasawa e60969
fukasawa e60969
Version 1.5.1beta08 [January 23, 2011]
fukasawa e60969
  Added png_get_io_chunk_type() and deprecated png_get_io_chunk_name()
fukasawa e60969
  Added synopses for the IO_STATE functions and other missing synopses
fukasawa e60969
    to the manual. Removed the synopses from libpngpf.3 because they
fukasawa e60969
    were out of date and no longer useful.  Better information can be
fukasawa e60969
    obtained by reading the prototypes and comments in pngpriv.h
fukasawa e60969
  Attempted to fix cpp on Solaris with S. Studio 12 cc, fix build
fukasawa e60969
    Added a make macro DFNCPP that is a CPP that will accept the tokens in
fukasawa e60969
    a .dfn file and adds configure stuff to test for such a CPP.  ./configure
fukasawa e60969
    should fail if one is not available.
fukasawa e60969
  Corrected const_png_ in png.h to png_const_ to avoid polluting the namespace.
fukasawa e60969
  Added png_get_current_row_number and png_get_current_pass_number for the
fukasawa e60969
    benefit of the user transform callback.
fukasawa e60969
  Added png_process_data_pause and png_process_data_skip for the benefit of
fukasawa e60969
    progressive readers that need to stop data processing or want to optimize
fukasawa e60969
    skipping of unread data (e.g., if the reader marks a chunk to be skipped.)
fukasawa e60969
fukasawa e60969
Version 1.5.1beta09 [January 24, 2011]
fukasawa e60969
  Enhanced pngvalid, corrected an error in gray_to_rgb, corrected doc error.
fukasawa e60969
    pngvalid contains tests of transforms, which tests are currently disabled
fukasawa e60969
    because they are incompletely tested.  gray_to_rgb was failing to expand
fukasawa e60969
    the bit depth for smaller bit depth images; this seems to be a long
fukasawa e60969
    standing error and resulted, apparently, in invalid output
fukasawa e60969
    (CVE-2011-0408, CERT VU#643140).  The documentation did not accurately
fukasawa e60969
    describe what libpng really does when converting RGB to gray.
fukasawa e60969
fukasawa e60969
Version 1.5.1beta10 [January 27, 2010]
fukasawa e60969
  Fixed incorrect examples of callback prototypes in the manual, that were
fukasawa e60969
    introduced in libpng-1.0.0.
fukasawa e60969
  In addition the order of the png_get_uint macros with respect to the
fukasawa e60969
    relevant function definitions has been reversed.  This helps the
fukasawa e60969
    preprocessing of the symbol files be more robust.  Furthermore, the
fukasawa e60969
    symbol file preprocessing now uses -DPNG_NO_USE_READ_MACROS even when
fukasawa e60969
    the library may actually be built with PNG_USE_READ_MACROS; this stops
fukasawa e60969
    the read macros interfering with the symbol file format.
fukasawa e60969
  Made the manual, synopses, and function prototypes use the function
fukasawa e60969
    argument names file_gamma, int_file_gamma, and srgb_intent consistently.
fukasawa e60969
fukasawa e60969
Version 1.5.1beta11 [January 28, 2011]
fukasawa e60969
  Changed PNG_UNUSED from "param=param;" to "{if(param){}}".
fukasawa e60969
  Corrected local variable type in new API png_process_data_skip()
fukasawa e60969
    The type was self-evidently incorrect but only causes problems on 64-bit
fukasawa e60969
    architectures.
fukasawa e60969
  Added transform tests to pngvalid and simplified the arguments.
fukasawa e60969
fukasawa e60969
Version 1.5.1rc01 [January 29, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.1rc02 [January 31, 2011]
fukasawa e60969
  Added a request in the manual that applications do not use "png_" or
fukasawa e60969
    "PNG_" to begin any of their own symbols.
fukasawa e60969
  Changed PNG_UNUSED to "(void)param;" and updated the commentary in pngpriv.h
fukasawa e60969
fukasawa e60969
Version 1.5.1 [February 3, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.2beta01 [February 13, 2011]
fukasawa e60969
  More -Wshadow fixes for older gcc compilers.  Older gcc versions apparently
fukasawa e60969
    check formal parameters names in function declarations (as well as
fukasawa e60969
    definitions) to see if they match a name in the global namespace.
fukasawa e60969
  Revised PNG_EXPORTA macro to not use an empty parameter, to accommodate the
fukasawa e60969
    old VisualC++ preprocessor.
fukasawa e60969
  Turned on interlace handling in png_read_png().
fukasawa e60969
  Fixed gcc pendantic warnings.
fukasawa e60969
  Handle longjmp in Cygwin.
fukasawa e60969
  Fixed png_get_current_row_number() in the interlaced case.
fukasawa e60969
  Cleaned up ALPHA flags and transformations.
fukasawa e60969
  Implemented expansion to 16 bits.
fukasawa e60969
fukasawa e60969
Version 1.5.2beta02 [February 19, 2011]
fukasawa e60969
  Fixed mistake in the descriptions of user read_transform and write_transform
fukasawa e60969
    function prototypes in the manual.  The row_info struct is png_row_infop.
fukasawa e60969
  Reverted png_get_current_row_number() to previous (1.5.2beta01) behavior.
fukasawa e60969
  Corrected png_get_current_row_number documentation
fukasawa e60969
  Fixed the read/write row callback documentation.
fukasawa e60969
    This documents the current behavior, where the callback is called after
fukasawa e60969
    every row with information pertaining to the next row.
fukasawa e60969
fukasawa e60969
Version 1.5.2beta03 [March 3, 2011]
fukasawa e60969
  Fixed scripts/makefile.vcwin32
fukasawa e60969
  Updated contrib/pngsuite/README to add the word "modify".
fukasawa e60969
  Define PNG_ALLOCATED to blank when _MSC_VER<1300.
fukasawa e60969
fukasawa e60969
Version 1.5.2rc01 [March 19, 2011]
fukasawa e60969
  Define remaining attributes to blank when MSC_VER<1300.
fukasawa e60969
  ifdef out mask arrays in pngread.c when interlacing is not supported.
fukasawa e60969
fukasawa e60969
Version 1.5.2rc02 [March 22, 2011]
fukasawa e60969
  Added a hint to try CPP=/bin/cpp if "cpp -E" fails in scripts/pnglibconf.mak
fukasawa e60969
    and in contrib/pngminim/*/makefile, eg., on SunOS 5.10, and removed "strip"
fukasawa e60969
    from the makefiles.
fukasawa e60969
  Fixed a bug (present since libpng-1.0.7) that makes png_handle_sPLT() fail
fukasawa e60969
    to compile when PNG_NO_POINTER_INDEXING is defined (Chubanov Kirill)
fukasawa e60969
fukasawa e60969
Version 1.5.2rc03 [March 24, 2011]
fukasawa e60969
  Don't include standard header files in png.h while building the symbol table,
fukasawa e60969
    to avoid cpp failure on SunOS (introduced PNG_BUILDING_SYMBOL_TABLE macro).
fukasawa e60969
fukasawa e60969
Version 1.5.2 [March 31, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.3beta01 [April 1, 2011]
fukasawa e60969
  Re-initialize the zlib compressor before compressing non-IDAT chunks.
fukasawa e60969
  Added API functions (png_set_text_compression_level() and four others) to
fukasawa e60969
    set parameters for zlib compression of non-IDAT chunks.
fukasawa e60969
fukasawa e60969
Version 1.5.3beta02 [April 3, 2011]
fukasawa e60969
  Updated scripts/symbols.def with new API functions.
fukasawa e60969
  Only compile the new zlib re-initializing code when text or iCCP is
fukasawa e60969
    supported, using PNG_WRITE_COMPRESSED_TEXT_SUPPORTED macro.
fukasawa e60969
  Improved the optimization of the zlib CMF byte (see libpng-1.2.6beta03).
fukasawa e60969
  Optimize the zlib CMF byte in non-IDAT compressed chunks
fukasawa e60969
fukasawa e60969
Version 1.5.3beta03 [April 16, 2011]
fukasawa e60969
  Fixed gcc -ansi -pedantic compile. A strict ANSI system does not have
fukasawa e60969
    snprintf, and the "__STRICT_ANSI__" detects that condition more reliably
fukasawa e60969
    than __STDC__ (John Bowler).
fukasawa e60969
  Removed the PNG_PTR_NORETURN attribute because it too dangerous. It tells
fukasawa e60969
    the compiler that a user supplied callback (the error handler) does not
fukasawa e60969
    return, yet there is no guarantee in practice that the application code
fukasawa e60969
    will correctly implement the error handler because the compiler only
fukasawa e60969
    issues a warning if there is a mistake (John Bowler).
fukasawa e60969
  Removed the no-longer-used PNG_DEPSTRUCT macro.
fukasawa e60969
  Updated the zlib version to 1.2.5 in the VStudio project.
fukasawa e60969
  Fixed 64-bit builds where png_uint_32 is smaller than png_size_t in
fukasawa e60969
    pngwutil.c (John Bowler).
fukasawa e60969
  Fixed bug with stripping the filler or alpha channel when writing, that
fukasawa e60969
    was introduced in libpng-1.5.2beta01 (bug report by Andrew Church).
fukasawa e60969
fukasawa e60969
Version 1.5.3beta04 [April 27, 2011]
fukasawa e60969
  Updated pngtest.png with the new zlib CMF optimization.
fukasawa e60969
  Cleaned up conditional compilation code and of background/gamma handling
fukasawa e60969
    Internal changes only except a new option to avoid compiling the
fukasawa e60969
    png_build_grayscale_palette API (which is not used at all internally.)
fukasawa e60969
    The main change is to move the transform tests (READ_TRANSFORMS,
fukasawa e60969
    WRITE_TRANSFORMS) up one level to the caller of the APIs.  This avoids
fukasawa e60969
    calls to spurious functions if all transforms are disabled and slightly
fukasawa e60969
    simplifies those functions.  Pngvalid modified to handle this.
fukasawa e60969
    A minor change is to stop the strip_16 and expand_16 interfaces from
fukasawa e60969
    disabling each other; this allows the future alpha premultiplication
fukasawa e60969
    code to use 16-bit intermediate values while still producing 8-bit output.
fukasawa e60969
    png_do_background and png_do_gamma have been simplified to take a single
fukasawa e60969
    pointer to the png_struct rather than pointers to every item required
fukasawa e60969
    from the png_struct. This makes no practical difference to the internal
fukasawa e60969
    code.
fukasawa e60969
  A serious bug in the pngvalid internal routine 'standard_display_init' has
fukasawa e60969
    been fixed - this failed to initialize the red channel and accidentally
fukasawa e60969
    initialized the alpha channel twice.
fukasawa e60969
  Changed png_struct jmp_buf member name from png_jmpbuf to tmp_jmpbuf to
fukasawa e60969
    avoid a possible clash with the png_jmpbuf macro on some platforms.
fukasawa e60969
fukasawa e60969
Version 1.5.3beta05 [May 6, 2011]
fukasawa e60969
  Added the "_POSIX_SOURCE" feature test macro to ensure libpng sees the
fukasawa e60969
    correct API. _POSIX_SOURCE is defined in pngpriv.h, pngtest.c and
fukasawa e60969
    pngvalid.c to ensure that POSIX conformant systems disable non-POSIX APIs.
fukasawa e60969
  Removed png_snprintf and added formatted warning messages.  This change adds
fukasawa e60969
    internal APIs to allow png_warning messages to have parameters without
fukasawa e60969
    requiring the host OS to implement snprintf.  As a side effect the
fukasawa e60969
    dependency of the tIME-supporting RFC1132 code on stdio is removed and
fukasawa e60969
    PNG_NO_WARNINGS does actually work now.
fukasawa e60969
  Pass "" instead of '\0' to png_default_error() in png_err().  This mistake
fukasawa e60969
    was introduced in libpng-1.2.20beta01.  This fixes CVE-2011-2691.
fukasawa e60969
  Added PNG_WRITE_OPTIMIZE_CMF_SUPPORTED macro to make the zlib "CMF" byte
fukasawa e60969
    optimization configureable.
fukasawa e60969
  IDAT compression failed if preceded by a compressed text chunk (bug
fukasawa e60969
    introduced in libpng-1.5.3beta01-02).  This was because the attempt to
fukasawa e60969
    reset the zlib stream in png_write_IDAT happened after the first IDAT
fukasawa e60969
    chunk had been deflated - much too late.  In this change internal
fukasawa e60969
    functions were added to claim/release the z_stream and, hopefully, make
fukasawa e60969
    the code more robust.  Also deflateEnd checking is added - previously
fukasawa e60969
    libpng would ignore an error at the end of the stream.
fukasawa e60969
fukasawa e60969
Version 1.5.3beta06 [May 8, 2011]
fukasawa e60969
  Removed the -D_ALL_SOURCE from definitions for AIX in CMakeLists.txt
fukasawa e60969
  Implemented premultiplied alpha support: png_set_alpha_mode API
fukasawa e60969
fukasawa e60969
Version 1.5.3beta07 [May 11, 2011]
fukasawa e60969
  Added expand_16 support to the high level interface.
fukasawa e60969
  Added named value and 'flag' gamma support to png_set_gamma.  Made a minor
fukasawa e60969
    change from the previous (unreleased) ABI/API to hide the exact value used
fukasawa e60969
    for Macs - it's not a good idea to embed this in the ABI!
fukasawa e60969
  Moved macro definitions for PNG_HAVE_IHDR, PNG_HAVE_PLTE, and PNG_AFTER_IDAT
fukasawa e60969
    from pngpriv.h to png.h because they must be visible to applications
fukasawa e60969
    that call png_set_unknown_chunks().
fukasawa e60969
  Check for up->location !PNG_AFTER_IDAT when writing unknown chunks
fukasawa e60969
    before IDAT.
fukasawa e60969
fukasawa e60969
Version 1.5.3beta08 [May 16, 2011]
fukasawa e60969
  Improved "pngvalid --speed" to exclude more of pngvalid from the time.
fukasawa e60969
  Documented png_set_alpha_mode(), other changes in libpng.3/libpng-manual.txt
fukasawa e60969
  The cHRM chunk now sets the defaults for png_set_rgb_to_gray() (when negative
fukasawa e60969
    parameters are supplied by the caller), while in the absence of cHRM
fukasawa e60969
    sRGB/Rec 709 values are still used.  This introduced a divide-by-zero
fukasawa e60969
    bug in png_handle_cHRM().
fukasawa e60969
  The bKGD chunk no longer overwrites the background value set by
fukasawa e60969
    png_set_background(), allowing the latter to be used before the file
fukasawa e60969
    header is read. It never performed any useful function to override
fukasawa e60969
    the default anyway.
fukasawa e60969
  Added memory overwrite and palette image checks to pngvalid.c
fukasawa e60969
    Previously palette image code was poorly checked. Since the transformation
fukasawa e60969
    code has a special palette path in most cases this was a severe weakness.
fukasawa e60969
  Minor cleanup and some extra checking in pngrutil.c and pngrtran.c. When
fukasawa e60969
    expanding an indexed image, always expand to RGBA if transparency is
fukasawa e60969
    present.
fukasawa e60969
fukasawa e60969
Version 1.5.3beta09 [May 17, 2011]
fukasawa e60969
  Reversed earlier 1.5.3 change of transformation order; move png_expand_16
fukasawa e60969
    back where it was.  The change doesn't work because it requires 16-bit
fukasawa e60969
    gamma tables when the code only generates 8-bit ones.  This fails
fukasawa e60969
    silently; the libpng code just doesn't do any gamma correction.  Moving
fukasawa e60969
    the tests back leaves the old, inaccurate, 8-bit gamma calculations, but
fukasawa e60969
    these are clearly better than none!
fukasawa e60969
fukasawa e60969
Version 1.5.3beta10 [May 20, 2011]
fukasawa e60969
fukasawa e60969
  png_set_background() and png_expand_16() did not work together correctly.
fukasawa e60969
    This problem is present in 1.5.2; if png_set_background is called with
fukasawa e60969
    need_expand false and the matching 16 bit color libpng erroneously just
fukasawa e60969
    treats it as an 8-bit color because of where png_do_expand_16 is in the
fukasawa e60969
    transform list.  This simple fix reduces the supplied colour to 8-bits,
fukasawa e60969
    so it gets smashed, but this is better than the current behavior.
fukasawa e60969
  Added tests for expand16, more fixes for palette image tests to pngvalid.
fukasawa e60969
    Corrects the code for palette image tests and disables attempts to
fukasawa e60969
    validate palette colors.
fukasawa e60969
fukasawa e60969
Version 1.5.3rc01 [June 3, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.3rc02 [June 8, 2011]
fukasawa e60969
  Fixed uninitialized memory read in png_format_buffer() (Bug report by
fukasawa e60969
    Frank Busse, CVE-2011-2501, related to CVE-2004-0421).
fukasawa e60969
fukasawa e60969
Version 1.5.3beta11 [June 11, 2011]
fukasawa e60969
  Fixed png_handle_sCAL which is broken in 1.5. This fixes CVE 2011-2692.
fukasawa e60969
  Added sCAL to pngtest.png
fukasawa e60969
  Revised documentation about png_set_user_limits() to say that it also affects
fukasawa e60969
    png writing.
fukasawa e60969
  Revised handling of png_set_user_limits() so that it can increase the
fukasawa e60969
    limit beyond the PNG_USER_WIDTH|HEIGHT_MAX; previously it could only
fukasawa e60969
    reduce it.
fukasawa e60969
  Make the 16-to-8 scaling accurate. Dividing by 256 with no rounding is
fukasawa e60969
    wrong (high by one) 25% of the time. Dividing by 257 with rounding is
fukasawa e60969
    wrong in 128 out of 65536 cases. Getting the right answer all the time
fukasawa e60969
    without division is easy.
fukasawa e60969
  Added "_SUPPORTED" to the PNG_WRITE_CUSTOMIZE_ZTXT_COMPRESSION macro.
fukasawa e60969
  Added projects/owatcom, an IDE project for OpenWatcom to replace
fukasawa e60969
    scripts/makefile.watcom.  This project works with OpenWatcom 1.9. The
fukasawa e60969
    IDE autogenerates appropriate makefiles (libpng.mk) for batch processing.
fukasawa e60969
    The project is configurable, unlike the Visual Studio project, so long
fukasawa e60969
    as the developer has an awk.
fukasawa e60969
  Changed png_set_gAMA to limit the gamma value range so that the inverse
fukasawa e60969
    of the stored value cannot overflow the fixed point representation,
fukasawa e60969
    and changed other things OpenWatcom warns about.
fukasawa e60969
  Revised pngvalid.c to test PNG_ALPHA_MODE_SUPPORTED correctly. This allows
fukasawa e60969
    pngvalid to build when ALPHA_MODE is not supported, which is required if
fukasawa e60969
    it is to build on libpng 1.4.
fukasawa e60969
  Removed string/memory macros that are no longer used and are not
fukasawa e60969
    necessarily fully supportable, particularly png_strncpy and png_snprintf.
fukasawa e60969
  Added log option to pngvalid.c and attempted to improve gamma messages.
fukasawa e60969
fukasawa e60969
Version 1.5.3 [omitted]
fukasawa e60969
  People found the presence of a beta release following an rc release
fukasawa e60969
    to be confusing; therefore we bump the version to libpng-1.5.4beta01
fukasawa e60969
    and there will be no libpng-1.5.3 release.
fukasawa e60969
fukasawa e60969
Version 1.5.4beta01 [June 14, 2011]
fukasawa e60969
  Made it possible to undefine PNG_READ_16_TO_8_ACCURATE_SCALE_SUPPORTED
fukasawa e60969
    to get the same (inaccurate) output as libpng-1.5.2 and earlier.
fukasawa e60969
  Moved definitions of PNG_HAVE_IHDR, PNG_AFTER_IDAT, and PNG_HAVE_PLTE
fukasawa e60969
    outside of an unknown-chunk block in png.h because they are also
fukasawa e60969
    needed for other uses.
fukasawa e60969
fukasawa e60969
Version 1.5.4beta02 [June 14, 2011]
fukasawa e60969
  Fixed and clarified LEGACY 16-to-8 scaling code.
fukasawa e60969
  Added png_set_chop_16() API, to match inaccurate results from previous
fukasawa e60969
    libpng versions.
fukasawa e60969
  Removed the ACCURATE and LEGACY options (they are no longer useable)
fukasawa e60969
  Use the old scaling method for background if png_set_chop_16() was
fukasawa e60969
    called.
fukasawa e60969
  Made png_set_chop_16() API removeable by disabling PNG_CHOP_16_TO_8_SUPPORTED
fukasawa e60969
fukasawa e60969
Version 1.5.4beta03 [June 15, 2011]
fukasawa e60969
  Fixed a problem in png_do_expand_palette() exposed by optimization in
fukasawa e60969
    1.5.3beta06
fukasawa e60969
  Also removed a spurious and confusing "trans" member ("trans") from png_info.
fukasawa e60969
  The palette expand optimization prevented expansion to an intermediate RGBA
fukasawa e60969
    form if tRNS was present but alpha was marked to be stripped; this exposed
fukasawa e60969
    a check for tRNS in png_do_expand_palette() which is inconsistent with the
fukasawa e60969
    code elsewhere in libpng.
fukasawa e60969
  Correction to the expand_16 code; removed extra instance of
fukasawa e60969
    png_set_scale_16_to_8 from pngpriv.h
fukasawa e60969
fukasawa e60969
Version 1.5.4beta04 [June 16, 2011]
fukasawa e60969
  Added a missing "#ifdef PNG_READ_BACKGROUND_SUPPORTED/#endif" in pngrtran.c
fukasawa e60969
  Added PNG_TRANSFORM_CHOP_16 to the high-level read transforms.
fukasawa e60969
  Made PNG_READ_16_TO_8_ACCURATE_SCALE configurable again.  If this is
fukasawa e60969
    not enabled, png_set_strip_16() and png_do_scale_16_to_8() aren't built.
fukasawa e60969
  Revised contrib/visupng, gregbook, and pngminim to demonstrate chop_16_to_8
fukasawa e60969
fukasawa e60969
Version 1.5.4beta05 [June 16, 2011]
fukasawa e60969
  Renamed png_set_strip_16() to png_set_scale_16() and renamed
fukasawa e60969
    png_set_chop_16() to png_set_strip(16) in an attempt to minimize the
fukasawa e60969
    behavior changes between libpng14 and libpng15.
fukasawa e60969
fukasawa e60969
Version 1.5.4beta06 [June 18, 2011]
fukasawa e60969
  Fixed new bug that was causing both strip_16 and scale_16 to be applied.
fukasawa e60969
fukasawa e60969
Version 1.5.4beta07 [June 19, 2011]
fukasawa e60969
  Fixed pngvalid, simplified macros, added checking for 0 in sCAL.
fukasawa e60969
    The ACCURATE scale macro is no longer defined in 1.5 - call the
fukasawa e60969
    png_scale_16_to_8 API.  Made sure that PNG_READ_16_TO_8 is still defined
fukasawa e60969
    if the png_strip_16_to_8 API is present.  png_check_fp_number now
fukasawa e60969
    maintains some state so that positive, negative and zero values are
fukasawa e60969
    identified.  sCAL uses these to be strictly spec conformant.
fukasawa e60969
fukasawa e60969
Version 1.5.4beta08 [June 23, 2011]
fukasawa e60969
  Fixed pngvalid if ACCURATE_SCALE is defined.
fukasawa e60969
  Updated scripts/pnglibconf.h.prebuilt.
fukasawa e60969
fukasawa e60969
Version 1.5.4rc01 [June 30, 2011]
fukasawa e60969
  Define PNG_ALLOCATED to "restrict" only if MSC_VER >= 1400.
fukasawa e60969
fukasawa e60969
Version 1.5.4 [July 7, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.5beta01 [July 13, 2011]
fukasawa e60969
  Fixed some typos and made other minor changes in the manual.
fukasawa e60969
  Updated contrib/pngminus/makefile.std (Samuli Souminen)
fukasawa e60969
fukasawa e60969
Version 1.5.5beta02 [July 14, 2011]
fukasawa e60969
  Revised Makefile.am and Makefile.in to look in the right directory for
fukasawa e60969
    pnglibconf.h.prebuilt
fukasawa e60969
fukasawa e60969
Version 1.5.5beta03 [July 27, 2011]
fukasawa e60969
  Enabled compilation with g++ compiler.  This compiler does not recognize
fukasawa e60969
    the file extension, so it always compiles with C++ rules.  Made minor
fukasawa e60969
    changes to pngrutil.c to cast results where C++ expects it but C does not.
fukasawa e60969
  Minor editing of libpng.3 and libpng-manual.txt.
fukasawa e60969
fukasawa e60969
Version 1.5.5beta04 [July 29, 2011]
fukasawa e60969
  Revised CMakeLists.txt (Clifford Yapp)
fukasawa e60969
  Updated commentary about the png_rgb_to_gray() default coefficients
fukasawa e60969
    in the manual and in pngrtran.c
fukasawa e60969
fukasawa e60969
Version 1.5.5beta05 [August 17, 2011]
fukasawa e60969
  Prevent unexpected API exports from non-libpng DLLs on Windows.  The "_DLL"
fukasawa e60969
    is removed from the test of whether a DLL is being built (this erroneously
fukasawa e60969
    caused the libpng APIs to be marked as DLL exports in static builds under
fukasawa e60969
    Microsoft Visual Studio).  Almost all of the libpng building configuration
fukasawa e60969
    is moved from pngconf.h to pngpriv.h, but PNG_DLL_EXPORT remains in
fukasawa e60969
    pngconf.h, though, so that it is colocated with the import definition (it
fukasawa e60969
    is no longer used anywhere in the installed headers).  The VStudio project
fukasawa e60969
    definitions have been cleaned up: "_USRDLL" has been removed from the
fukasawa e60969
    static library builds (this was incorrect), and PNG_USE_DLL has been added
fukasawa e60969
    to pngvalid to test the functionality (pngtest does not supply it,
fukasawa e60969
    deliberately).  The spurious "_EXPORTS" has been removed from the
fukasawa e60969
    libpng build (all these errors were a result of copy/paste between project
fukasawa e60969
    configurations.)
fukasawa e60969
  Added new types and internal functions for CIE RGB end point handling to
fukasawa e60969
    pngpriv.h (functions yet to be implemented).
fukasawa e60969
fukasawa e60969
Version 1.5.5beta06 [August 26, 2011]
fukasawa e60969
  Ensure the CMAKE_LIBRARY_OUTPUT_DIRECTORY is set in CMakeLists.txt
fukasawa e60969
    (Clifford Yap)
fukasawa e60969
  Fixes to rgb_to_gray and cHRM XYZ APIs (John Bowler):
fukasawa e60969
    The rgb_to_gray code had errors when combined with gamma correction.
fukasawa e60969
    Some pixels were treated as true grey when they weren't and such pixels
fukasawa e60969
    and true grey ones were not gamma corrected (the original value of the
fukasawa e60969
    red component was used instead).  APIs to get and set cHRM using color
fukasawa e60969
    space end points have been added and the rgb_to_gray code that defaults
fukasawa e60969
    based on cHRM, and the divide-by-zero bug in png_handle_cHRM (CERT
fukasawa e60969
    VU#477046, CVE-2011-3328, introduced in 1.5.4) have been corrected.
fukasawa e60969
  A considerable number of tests has been added to pngvalid for the
fukasawa e60969
    rgb_to_gray transform.
fukasawa e60969
  Arithmetic errors in rgb_to_gray whereby the calculated gray value was
fukasawa e60969
    truncated to the bit depth rather than rounded have been fixed except in
fukasawa e60969
    the 8-bit non-gamma-corrected case (where consistency seems more important
fukasawa e60969
    than correctness.)  The code still has considerable inaccuracies in the
fukasawa e60969
    8-bit case because 8-bit linear arithmetic is used.
fukasawa e60969
fukasawa e60969
Version 1.5.5beta07 [September 7, 2011]
fukasawa e60969
  Added "$(ARCH)" option to makefile.darwin
fukasawa e60969
  Added SunOS support to configure.ac and Makefile.am
fukasawa e60969
  Changed png_chunk_benign_error() to png_warning() in png.c, in
fukasawa e60969
    png_XYZ_from_xy_checked().
fukasawa e60969
fukasawa e60969
Version 1.5.5beta08 [September 10, 2011]
fukasawa e60969
  Fixed 64-bit compilation errors (gcc). The errors fixed relate
fukasawa e60969
    to conditions where types that are 32 bits in the GCC 32-bit
fukasawa e60969
    world (uLong and png_size_t) become 64 bits in the 64-bit
fukasawa e60969
    world.  This produces potential truncation errors which the
fukasawa e60969
    compiler correctly flags.
fukasawa e60969
  Relocated new HAVE_SOLARIS_LD definition in configure.ac
fukasawa e60969
  Constant changes for 64-bit compatibility (removal of L suffixes). The
fukasawa e60969
    16-bit cases still use "L" as we don't have a 16-bit test system.
fukasawa e60969
fukasawa e60969
Version 1.5.5rc01 [September 15, 2011]
fukasawa e60969
  Removed "L" suffixes in pngpriv.h
fukasawa e60969
fukasawa e60969
Version 1.5.5 [September 22, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.6beta01 [September 22, 2011]
fukasawa e60969
  Fixed some 64-bit type conversion warnings in pngrtran.c
fukasawa e60969
  Moved row_info from png_struct to a local variable.
fukasawa e60969
  The various interlace mask arrays have been made into arrays of
fukasawa e60969
    bytes and made PNG_CONST and static (previously some arrays were
fukasawa e60969
    marked PNG_CONST and some weren't).
fukasawa e60969
  Additional checks have been added to the transform code to validate the
fukasawa e60969
    pixel depths after the transforms on both read and write.
fukasawa e60969
  Removed some redundant code from pngwrite.c, in png_destroy_write_struct().
fukasawa e60969
  Changed chunk reading/writing code to use png_uint_32 instead of png_byte[4].
fukasawa e60969
    This removes the need to allocate temporary strings for chunk names on
fukasawa e60969
    the stack in the read/write code.  Unknown chunk handling still uses the
fukasawa e60969
    string form because this is exposed in the API.
fukasawa e60969
fukasawa e60969
Version 1.5.6beta02 [September 26, 2011]
fukasawa e60969
  Added a note in the manual the png_read_update_info() must be called only
fukasawa e60969
    once with a particular info_ptr.
fukasawa e60969
  Fixed a typo in the definition of the new PNG_STRING_FROM_CHUNK(s,c) macro.
fukasawa e60969
fukasawa e60969
Version 1.5.6beta03 [September 28, 2011]
fukasawa e60969
  Revised test-pngtest.sh to report FAIL when pngtest fails.
fukasawa e60969
  Added "--strict" option to pngtest, to report FAIL when the failure is
fukasawa e60969
    only because the resulting valid files are different.
fukasawa e60969
  Revised CMakeLists.txt to work with mingw and removed some material from
fukasawa e60969
    CMakeLists.txt that is no longer useful in libpng-1.5.
fukasawa e60969
fukasawa e60969
Version 1.5.6beta04 [October 5, 2011]
fukasawa e60969
  Fixed typo in Makefile.in and Makefile.am ("-M Wl" should be "-M -Wl")."
fukasawa e60969
fukasawa e60969
Version 1.5.6beta05 [October 12, 2011]
fukasawa e60969
  Speed up png_combine_row() for interlaced images. This reduces the generality
fukasawa e60969
    of the code, allowing it to be optimized for Adam7 interlace.  The masks
fukasawa e60969
    passed to png_combine_row() are now generated internally, avoiding
fukasawa e60969
    some code duplication and localizing the interlace handling somewhat.
fukasawa e60969
  Align png_struct::row_buf - previously it was always unaligned, caused by
fukasawa e60969
    a bug in the code that attempted to align it; the code needs to subtract
fukasawa e60969
    one from the pointer to take account of the filter byte prepended to
fukasawa e60969
    each row.
fukasawa e60969
  Optimized png_combine_row() when rows are aligned. This gains a small
fukasawa e60969
    percentage for 16-bit and 32-bit pixels in the typical case where the
fukasawa e60969
    output row buffers are appropriately aligned. The optimization was not
fukasawa e60969
    previously possible because the png_struct buffer was always misaligned.
fukasawa e60969
  Fixed bug in png_write_chunk_header() debug print, introduced in 1.5.6beta01.
fukasawa e60969
fukasawa e60969
Version 1.5.6beta06 [October 17, 2011]
fukasawa e60969
  Removed two redundant tests for unitialized row.
fukasawa e60969
  Fixed a relatively harmless memory overwrite in compressed text writing
fukasawa e60969
    with a 1 byte zlib buffer.
fukasawa e60969
  Add ability to call png_read_update_info multiple times to pngvalid.c.
fukasawa e60969
  Fixes for multiple calls to png_read_update_info. These fixes attend to
fukasawa e60969
    most of the errors revealed in pngvalid, however doing the gamma work
fukasawa e60969
    twice results in inaccuracies that can't be easily fixed.  There is now
fukasawa e60969
    a warning in the code if this is going to happen.
fukasawa e60969
  Turned on multiple png_read_update_info in pngvalid transform tests.
fukasawa e60969
  Prevent libpng from overwriting unused bits at the end of the image when
fukasawa e60969
    it is not byte aligned, while reading. Prior to libpng-1.5.6 libpng would
fukasawa e60969
    overwrite the partial byte at the end of each row if the row width was not
fukasawa e60969
    an exact multiple of 8 bits and the image is not interlaced.
fukasawa e60969
fukasawa e60969
Version 1.5.6beta07 [October 21, 2011]
fukasawa e60969
  Made png_ptr->prev_row an aligned pointer into png_ptr->big_prev_row
fukasawa e60969
    (Mans Rullgard).
fukasawa e60969
fukasawa e60969
Version 1.5.6rc01 [October 26, 2011]
fukasawa e60969
  Changed misleading "Missing PLTE before cHRM" warning to "Out of place cHRM"
fukasawa e60969
fukasawa e60969
Version 1.5.6rc02 [October 27, 2011]
fukasawa e60969
  Added LSR() macro to defend against buggy compilers that evaluate non-taken
fukasawa e60969
    code branches and complain about out-of-range shifts.
fukasawa e60969
fukasawa e60969
Version 1.5.6rc03 [October 28, 2011]
fukasawa e60969
  Renamed the LSR() macro to PNG_LSR() and added PNG_LSL() macro.
fukasawa e60969
  Fixed compiler warnings with Intel and MSYS compilers. The logical shift
fukasawa e60969
    fix for Microsoft Visual C is required by other compilers, so this
fukasawa e60969
    enables that fix for all compilers when using compile-time constants.
fukasawa e60969
    Under MSYS 'byte' is a name declared in a system header file, so we
fukasawa e60969
    changed the name of a local variable to avoid the warnings that result.
fukasawa e60969
  Added #define PNG_ALIGN_TYPE PNG_ALIGN_NONE to contrib/pngminim/*/pngusr.h
fukasawa e60969
fukasawa e60969
Version 1.5.6 [November 3, 2011]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.5.7beta01 [November 4, 2011]
fukasawa e60969
  Added support for ARM processor, when decoding all PNG up-filtered rows
fukasawa e60969
    and any other-filtered rows with 3 or 4 bytes per pixel (Mans Rullgard).
fukasawa e60969
  Fixed bug in pngvalid on early allocation failure; fixed type cast in
fukasawa e60969
    pngmem.c; pngvalid would attempt to call png_error() if the allocation
fukasawa e60969
    of a png_struct or png_info failed. This would probably have led to a
fukasawa e60969
    crash.  The pngmem.c implementation of png_malloc() included a cast
fukasawa e60969
    to png_size_t which would fail on large allocations on 16-bit systems.
fukasawa e60969
  Fix for the preprocessor of the Intel C compiler. The preprocessor
fukasawa e60969
    splits adjacent @ signs with a space; this changes the concatentation
fukasawa e60969
    token from @-@-@ to PNG_JOIN; that should work with all compiler
fukasawa e60969
    preprocessors.
fukasawa e60969
  Paeth filter speed improvements from work by Siarhei Siamashka. This
fukasawa e60969
    changes the 'Paeth' reconstruction function to improve the GCC code
fukasawa e60969
    generation on x86. The changes are only part of the suggested ones;
fukasawa e60969
    just the changes that definitely improve speed and remain simple.
fukasawa e60969
    The changes also slightly increase the clarity of the code.
fukasawa e60969
fukasawa e60969
Version 1.5.7beta02 [November 11, 2011]
fukasawa e60969
  Check compression_type parameter in png_get_iCCP and remove spurious
fukasawa e60969
    casts. The compression_type parameter is always assigned to, so must
fukasawa e60969
    be non-NULL. The cast of the profile length potentially truncated the
fukasawa e60969
    value unnecessarily on a 16-bit int system, so the cast of the (byte)
fukasawa e60969
    compression type to (int) is specified by ANSI-C anyway.
fukasawa e60969
  Fixed FP division by zero in pngvalid.c; the 'test_pixel' code left
fukasawa e60969
    the sBIT fields in the test pixel as 0, which resulted in a floating
fukasawa e60969
    point division by zero which was irrelevant but causes systems where
fukasawa e60969
    FP exceptions cause a crash. Added code to pngvalid to turn on FP
fukasawa e60969
    exceptions if the appropriate glibc support is there to ensure this is
fukasawa e60969
    tested in the future.
fukasawa e60969
  Updated scripts/pnglibconf.mak and scripts/makefile.std to handle the
fukasawa e60969
    new PNG_JOIN macro.
fukasawa e60969
  Added versioning to pnglibconf.h comments.
fukasawa e60969
  Simplified read/write API initial version; basic read/write tested on
fukasawa e60969
    a variety of images, limited documentation (in the header file.)
fukasawa e60969
  Installed more accurate linear to sRGB conversion tables. The slightly
fukasawa e60969
    modified tables reduce the number of 16-bit values that
fukasawa e60969
    convert to an off-by-one 8-bit value.  The "makesRGB.c" code that was used
fukasawa e60969
    to generate the tables is now in a contrib/sRGBtables sub-directory.
fukasawa e60969
fukasawa e60969
Version 1.5.7beta03 [November 17, 2011]
fukasawa e60969
  Removed PNG_CONST from the sRGB table declarations in pngpriv.h and png.c
fukasawa e60969
  Added run-time detection of NEON support.
fukasawa e60969
  Added contrib/libtests; includes simplified API test and timing test and
fukasawa e60969
    a color conversion utility for rapid checking of failed 'pngstest' results.
fukasawa e60969
  Multiple transform bug fixes plus a work-round for double gamma correction.
fukasawa e60969
    libpng does not support more than one transform that requires linear data
fukasawa e60969
    at once - if this is tried typically the results is double gamma
fukasawa e60969
    correction. Since the simplified APIs can need rgb to gray combined with
fukasawa e60969
    a compose operation it is necessary to do one of these outside the main
fukasawa e60969
    libpng transform code. This check-in also contains fixes to various bugs
fukasawa e60969
    in the simplified APIs themselves and to some bugs in compose and rgb to
fukasawa e60969
    gray (on palette) itself.
fukasawa e60969
  Fixes for C++ compilation using g++ When libpng source is compiled
fukasawa e60969
    using g++. The compiler imposes C++ rules on the C source; thus it
fukasawa e60969
    is desireable to make the source work with either C or C++ rules
fukasawa e60969
    without throwing away useful error information.  This change adds
fukasawa e60969
    png_voidcast to allow C semantic (void*) cases or the corresponding
fukasawa e60969
    C++ static_cast operation, as appropriate.
fukasawa e60969
  Added --noexecstack to assembler file compilation. GCC does not set
fukasawa e60969
    this on assembler compilation, even though it does on C compilation.
fukasawa e60969
    This creates security issues if assembler code is enabled; the
fukasawa e60969
    work-around is to set it by default in the flags for $(CCAS)
fukasawa e60969
  Work around compilers that don't support declaration of const data. Some
fukasawa e60969
    compilers fault 'extern const' data declarations (because the data is
fukasawa e60969
    not initialized); this turns on const-ness only for compilers where
fukasawa e60969
    this is known to work.
fukasawa e60969
fukasawa e60969
Version 1.5.7beta04 [November 17, 2011]
fukasawa e60969
  Since the gcc driver does not recognize the --noexecstack flag, we must
fukasawa e60969
    use the -Wa prefix to have it passed through to the assembler.
fukasawa e60969
    Also removed a duplicate setting of this flag.
fukasawa e60969
  Added files that were omitted from the libpng-1.5.7beta03 zip distribution.
fukasawa e60969
fukasawa e60969
Version 1.5.7beta05 [November 25, 2011]
fukasawa e60969
  Removed "zTXt" from warning in generic chunk decompression function.
fukasawa e60969
  Validate time settings passed to png_set_tIME() and png_convert_to_rfc1123()
fukasawa e60969
    (Frank Busse). Note: This prevented CVE-2015-7981 from affecting
fukasawa e60969
    libpng-1.5.7 and later.
fukasawa e60969
  Added MINGW support to CMakeLists.txt
fukasawa e60969
  Reject invalid compression flag or method when reading the iTXt chunk.
fukasawa e60969
  Backed out 'simplified' API changes. The API seems too complex and there
fukasawa e60969
    is a lack of consensus or enthusiasm for the proposals.  The API also
fukasawa e60969
    reveals significant bugs inside libpng (double gamma correction and the
fukasawa e60969
    known bug of being unable to retrieve a corrected palette). It seems
fukasawa e60969
    better to wait until the bugs, at least, are corrected.
fukasawa e60969
  Moved pngvalid.c into contrib/libtests
fukasawa e60969
  Rebuilt Makefile.in, configure, etc., with autoconf-2.68
fukasawa e60969
fukasawa e60969
Version 1.5.7rc01 [December 1, 2011]
fukasawa e60969
  Replaced an "#if" with "#ifdef" in pngrtran.c
fukasawa e60969
  Revised #if PNG_DO_BC block in png.c (use #ifdef and add #else)
fukasawa e60969
fukasawa e60969
Version 1.5.7rc02 [December 5, 2011]
fukasawa e60969
  Revised project files and contrib/pngvalid/pngvalid.c to account for
fukasawa e60969
    the relocation of pngvalid into contrib/libtests.
fukasawa e60969
  Revised pngconf.h to use " __declspec(restrict)" only when MSC_VER >= 1400,
fukasawa e60969
    as in libpng-1.5.4.
fukasawa e60969
  Put CRLF line endings in the owatcom project files.
fukasawa e60969
fukasawa e60969
Version 1.5.7rc03 [December 7, 2011]
fukasawa e60969
  Updated CMakeLists.txt to account for the relocation of pngvalid.c
fukasawa e60969
fukasawa e60969
Version 1.5.7 [December 15, 2011]
fukasawa e60969
  Minor fixes to pngvalid.c for gcc 4.6.2 compatibility to remove warnings
fukasawa e60969
    reported by earlier versions.
fukasawa e60969
  Fixed minor memset/sizeof errors in pngvalid.c.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta01 [December 15, 2011]
fukasawa e60969
  Removed machine-generated configure files from the GIT repository (they will
fukasawa e60969
    continue to appear in the tarball distributions and in the libpng15 and
fukasawa e60969
    earlier GIT branches).
fukasawa e60969
  Restored the new 'simplified' API, which was started in libpng-1.5.7beta02
fukasawa e60969
    but later deleted from libpng-1.5.7beta05.
fukasawa e60969
  Added example programs for the new 'simplified' API.
fukasawa e60969
  Added ANSI-C (C90) headers and require them, and take advantage of the
fukasawa e60969
    change. Also fixed some of the projects/* and contrib/* files that needed
fukasawa e60969
    updates for libpng16 and the move of pngvalid.c.
fukasawa e60969
    With this change the required ANSI-C header files are assumed to exist: the
fukasawa e60969
    implementation must provide float.h, limits.h, stdarg.h and stddef.h and
fukasawa e60969
    libpng relies on limits.h and stddef.h existing and behaving as defined
fukasawa e60969
    (the other two required headers aren't used).  Non-ANSI systems that don't
fukasawa e60969
    have stddef.h or limits.h will have to provide an appropriate fake
fukasawa e60969
    containing the relevant types and #defines.
fukasawa e60969
  Dropped support for 16-bit platforms. The use of FAR/far has been eliminated
fukasawa e60969
    and the definition of png_alloc_size_t is now controlled by a flag so
fukasawa e60969
    that 'small size_t' systems can select it if necessary.  Libpng 1.6 may
fukasawa e60969
    not currently work on such systems -- it seems likely that it will
fukasawa e60969
    ask 'malloc' for more than 65535 bytes with any image that has a
fukasawa e60969
    sufficiently large row size (rather than simply failing to read such
fukasawa e60969
    images).
fukasawa e60969
  New tools directory containing tools used to generate libpng code.
fukasawa e60969
  Fixed race conditions in parallel make builds. With higher degrees of
fukasawa e60969
    parallelism during 'make' the use of the same temporary file names such
fukasawa e60969
    as 'dfn*' can result in a race where a temporary file from one arm of the
fukasawa e60969
    build is deleted or overwritten in another arm.  This changes the
fukasawa e60969
    temporary files for suffix rules to always use $* and ensures that the
fukasawa e60969
    non-suffix rules use unique file names.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta02 [December 21, 2011]
fukasawa e60969
  Correct configure builds where build and source directories are separate.
fukasawa e60969
    The include path of 'config.h' was erroneously made relative in pngvalid.c
fukasawa e60969
    in libpng 1.5.7.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta03 [December 22, 2011]
fukasawa e60969
  Start-up code size improvements, error handler flexibility. These changes
fukasawa e60969
    alter how the tricky allocation of the initial png_struct and png_info
fukasawa e60969
    structures are handled. png_info is now handled in pretty much the same
fukasawa e60969
    way as everything else, except that the allocations handle NULL return
fukasawa e60969
    silently.  png_struct is changed in a similar way on allocation and on
fukasawa e60969
    deallocation a 'safety' error handler is put in place (which should never
fukasawa e60969
    be required).  The error handler itself is changed to permit mismatches
fukasawa e60969
    in the application and libpng error buffer size; however, this means a
fukasawa e60969
    silent change to the API to return the jmp_buf if the size doesn't match
fukasawa e60969
    the size from the libpng compilation; libpng now allocates the memory and
fukasawa e60969
    this may fail.  Overall these changes result in slight code size
fukasawa e60969
    reductions; however, this is a reduction in code that is always executed
fukasawa e60969
    so is particularly valuable.  Overall on a 64-bit system the libpng DLL
fukasawa e60969
    decreases in code size by 1733 bytes.  pngerror.o increases in size by
fukasawa e60969
    about 465 bytes because of the new functionality.
fukasawa e60969
  Added png_convert_to_rfc1123_buffer() and deprecated png_convert_to_rfc1123()
fukasawa e60969
    to avoid including a spurious buffer in the png_struct.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta04 [December 30, 2011]
fukasawa e60969
  Regenerated configure scripts with automake-1.11.2
fukasawa e60969
  Eliminated png_info_destroy(). It is now used only in png.c and only calls
fukasawa e60969
    one other internal function and memset().
fukasawa e60969
  Enabled png_get_sCAL_fixed() if floating point APIs are enabled. Previously
fukasawa e60969
    it was disabled whenever internal fixed point arithmetic was selected,
fukasawa e60969
    which meant it didn't exist even on systems where FP was available but not
fukasawa e60969
    preferred.
fukasawa e60969
  Added pngvalid.c compile time checks for const APIs.
fukasawa e60969
  Implemented 'restrict' for png_info and png_struct. Because of the way
fukasawa e60969
    libpng works both png_info and png_struct are always accessed via a
fukasawa e60969
    single pointer.  This means adding C99 'restrict' to the pointer gives
fukasawa e60969
    the compiler some opportunity to optimize the code.  This change allows
fukasawa e60969
    that.
fukasawa e60969
  Moved AC_MSG_CHECKING([if libraries can be versioned]) later to the proper
fukasawa e60969
    location in configure.ac (Gilles Espinasse).
fukasawa e60969
  Changed png_memcpy to C assignment where appropriate. Changed all those
fukasawa e60969
    uses of png_memcpy that were doing a simple assignment to assignments
fukasawa e60969
    (all those cases where the thing being copied is a non-array C L-value).
fukasawa e60969
  Added some error checking to png_set_*() routines.
fukasawa e60969
  Removed the reference to the non-exported function png_memcpy() from
fukasawa e60969
    example.c.
fukasawa e60969
  Fixed the Visual C 64-bit build - it requires jmp_buf to be aligned, but
fukasawa e60969
    it had become misaligned.
fukasawa e60969
  Revised contrib/pngminus/pnm2png.c to avoid warnings when png_uint_32
fukasawa e60969
    and unsigned long are of different sizes.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta05 [January 15, 2012]
fukasawa e60969
  Updated manual with description of the simplified API (copied from png.h)
fukasawa e60969
  Fix bug in pngerror.c: some long warnings were being improperly truncated
fukasawa e60969
    (CVE-2011-3464, bug introduced in libpng-1.5.3beta05).
fukasawa e60969
fukasawa e60969
Version 1.6.0beta06 [January 24, 2012]
fukasawa e60969
  Added palette support to the simplified APIs. This commit
fukasawa e60969
    changes some of the macro definitions in png.h, app code
fukasawa e60969
    may need corresponding changes.
fukasawa e60969
  Increased the formatted warning buffer to 192 bytes.
fukasawa e60969
  Added color-map support to simplified API. This is an initial version for
fukasawa e60969
    review; the documentation has not yet been updated.
fukasawa e60969
  Fixed Min/GW uninstall to remove libpng.dll.a
fukasawa e60969
fukasawa e60969
Version 1.6.0beta07 [January 28, 2012]
fukasawa e60969
  Eliminated Intel icc/icl compiler warnings. The Intel (GCC derived)
fukasawa e60969
    compiler issues slightly different warnings from those issued by the
fukasawa e60969
    current vesions of GCC. This eliminates those warnings by
fukasawa e60969
    adding/removing casts and small code rewrites.
fukasawa e60969
  Updated configure.ac from autoupdate: added --enable-werror option.
fukasawa e60969
    Also some layout regularization and removal of introduced tab characters
fukasawa e60969
    (replaced with 3-character indentation).  Obsolete macros identified by
fukasawa e60969
    autoupdate have been removed; the replacements are all in 2.59 so
fukasawa e60969
    the pre-req hasn't been changed.  --enable-werror checks for support
fukasawa e60969
    for -Werror (or the given argument) in the compiler.  This mimics the
fukasawa e60969
    gcc configure option by allowing -Werror to be turned on safely; without
fukasawa e60969
    the option the tests written in configure itself fail compilation because
fukasawa e60969
    they cause compiler warnings.
fukasawa e60969
  Rewrote autogen.sh to run autoreconf instead of running tools one-by-one.
fukasawa e60969
  Conditionalize the install rules for MINGW and CYGWIN in CMakeLists.txt and
fukasawa e60969
    set CMAKE_LIBRARY_OUTPUT_DIRECTORY to "lib" on all platforms (C. Yapp).
fukasawa e60969
  Freeze libtool files in the 'scripts' directory. This version of autogen.sh
fukasawa e60969
    attempts to dissuade people from running it when it is not, or should not,
fukasawa e60969
    be necessary.  In fact, autogen.sh does not work when run in a libpng
fukasawa e60969
    directory extracted from a tar distribution anymore. You must run it in
fukasawa e60969
    a GIT clone instead.
fukasawa e60969
  Added two images to contrib/pngsuite (1-bit and 2-bit transparent grayscale),
fukasawa e60969
    and renamed three whose names were inconsistent with those in
fukasawa e60969
    pngsuite/README.txt.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta08 [February 1, 2012]
fukasawa e60969
  Fixed Image::colormap misalignment in pngstest.c
fukasawa e60969
  Check libtool/libtoolize version number (2.4.2) in configure.ac
fukasawa e60969
  Divide test-pngstest.sh into separate pngstest runs for basic and
fukasawa e60969
    transparent images.
fukasawa e60969
  Moved automake options to AM_INIT_AUTOMAKE in configure.ac
fukasawa e60969
  Added color-tests, silent-rules (Not yet implemented in Makefile.am) and
fukasawa e60969
    version checking to configure.ac
fukasawa e60969
  Improved pngstest speed by not doing redundant tests and add const to
fukasawa e60969
    the background parameter of png_image_finish_read. The --background
fukasawa e60969
    option is now done automagically only when required, so that commandline
fukasawa e60969
    option no longer exists.
fukasawa e60969
  Cleaned up pngpriv.h to consistently declare all functions and data.
fukasawa e60969
    Also eliminated PNG_CONST_DATA, which is apparently not needed but we
fukasawa e60969
    can't be sure until it is gone.
fukasawa e60969
  Added symbol prefixing that allows all the libpng external symbols
fukasawa e60969
    to be prefixed (suggested by Reuben Hawkins).
fukasawa e60969
  Updated "ftbb*.png" list in the owatcom and vstudio projects.
fukasawa e60969
  Fixed 'prefix' builds on clean systems. The generation of pngprefix.h
fukasawa e60969
    should not require itself.
fukasawa e60969
  Updated INSTALL to explain that autogen.sh must be run in a GIT clone,
fukasawa e60969
    not in a libpng directory extracted from a tar distribution.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta09 [February 1, 2012]
fukasawa e60969
  Reverted the prebuilt configure files to libpng-1.6.0beta05 condition.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta10 [February 3, 2012]
fukasawa e60969
  Added Z_SOLO for zlib-1.2.6+ and correct pngstest tests
fukasawa e60969
  Updated list of test images in CMakeLists.txt
fukasawa e60969
  Updated the prebuilt configure files to current condition.
fukasawa e60969
  Revised INSTALL information about autogen.sh; it works in tar distributions.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta11 [February 16, 2012]
fukasawa e60969
  Fix character count in pngstest command in projects/owatcom/pngstest.tgt
fukasawa e60969
  Revised test-pngstest.sh to report PASS/FAIL for each image.
fukasawa e60969
  Updated documentation about the simplified API.
fukasawa e60969
  Corrected estimate of error in libpng png_set_rgb_to_gray API.  The API is
fukasawa e60969
    extremely inaccurate for sRGB conversions because it uses an 8-bit
fukasawa e60969
    intermediate linear value and it does not use the sRGB transform, so it
fukasawa e60969
    suffers from the known instability in gamma transforms for values close
fukasawa e60969
    to 0 (see Poynton).  The net result is that the calculation has a maximum
fukasawa e60969
    error of 14.99/255; 0.5/255^(1/2.2).  pngstest now uses 15 for the
fukasawa e60969
    permitted 8-bit error. This may still not be enough because of arithmetic
fukasawa e60969
    error.
fukasawa e60969
  Removed some unused arrays (with #ifdef) from png_read_push_finish_row().
fukasawa e60969
  Fixed a memory overwrite bug in simplified read of RGB PNG with
fukasawa e60969
    non-linear gamma Also bugs in the error checking in pngread.c and changed
fukasawa e60969
    quite a lot of the checks in pngstest.c to be correct; either correctly
fukasawa e60969
    written or not over-optimistic.  The pngstest changes are insufficient to
fukasawa e60969
    allow all possible RGB transforms to be passed; pngstest cmppixel needs
fukasawa e60969
    to be rewritten to make it clearer which errors it allows and then changed
fukasawa e60969
    to permit known inaccuracies.
fukasawa e60969
  Removed tests for no-longer-used *_EMPTY_PLTE_SUPPORTED from pngstruct.h
fukasawa e60969
  Fixed fixed/float API export conditionals. 1) If FIXED_POINT or
fukasawa e60969
    FLOATING_POINT options were switched off, png.h ended up with lone ';'
fukasawa e60969
    characters.  This is not valid ANSI-C outside a function.  The ';'
fukasawa e60969
    characters have been moved inside the definition of PNG_FP_EXPORT and
fukasawa e60969
    PNG_FIXED_EXPORT. 2) If either option was switched off, the declaration
fukasawa e60969
    of the corresponding functions were completely omitted, even though some
fukasawa e60969
    of them are still used internally.  The result is still valid, but
fukasawa e60969
    produces warnings from gcc with some warning options (including -Wall). The
fukasawa e60969
    fix is to cause png.h to declare the functions with PNG_INTERNAL_FUNCTION
fukasawa e60969
    when png.h is included from pngpriv.h.
fukasawa e60969
  Check for invalid palette index while reading paletted PNG.  When one is
fukasawa e60969
    found, issue a warning and increase png_ptr->num_palette accordingly.
fukasawa e60969
    Apps are responsible for checking to see if that happened.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta12 [February 18, 2012]
fukasawa e60969
  Do not increase num_palette on invalid_index.
fukasawa e60969
  Relocated check for invalid palette index to pngrtran.c, after unpacking
fukasawa e60969
    the sub-8-bit pixels.
fukasawa e60969
  Fixed CVE-2011-3026 buffer overrun bug.  This bug was introduced when
fukasawa e60969
    iCCP chunk support was added at libpng-1.0.6. Deal more correctly with the
fukasawa e60969
    test on iCCP chunk length. Also removed spurious casts that may hide
fukasawa e60969
    problems on 16-bit systems.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta13 [February 24, 2012]
fukasawa e60969
  Eliminated redundant png_push_read_tEXt|zTXt|iTXt|unknown code from
fukasawa e60969
    pngpread.c and use the sequential png_handle_tEXt, etc., in pngrutil.c;
fukasawa e60969
    now that png_ptr->buffer is inaccessible to applications, the special
fukasawa e60969
    handling is no longer useful.
fukasawa e60969
  Added PNG_SAFE_LIMITS feature to pnglibconf.dfa, pngpriv.h, and new
fukasawa e60969
    pngusr.dfa to reset the user limits to safe ones if PNG_SAFE_LIMITS is
fukasawa e60969
    defined.  To enable, use "CPPFLAGS=-DPNG_SAFE_LIMITS_SUPPORTED=1" on the
fukasawa e60969
    configure command or put #define PNG_SAFE_LIMITS_SUPPORTED in
fukasawa e60969
    pnglibconf.h.prebuilt and pnglibconf.h.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta14 [February 27, 2012]
fukasawa e60969
  Added information about the new limits in the manual.
fukasawa e60969
  Updated Makefile.in
fukasawa e60969
fukasawa e60969
Version 1.6.0beta15 [March 2, 2012]
fukasawa e60969
  Removed unused "current_text" members of png_struct and the png_free()
fukasawa e60969
    of png_ptr->current_text from pngread.c
fukasawa e60969
  Rewrote pngstest.c for substantial speed improvement.
fukasawa e60969
  Fixed transparent pixel and 16-bit rgb tests in pngstest and removed a
fukasawa e60969
    spurious check in pngwrite.c
fukasawa e60969
  Added PNG_IMAGE_FLAG_FAST for the benefit of applications that store
fukasawa e60969
    intermediate files, or intermediate in-memory data, while processing
fukasawa e60969
    image data with the simplified API.  The option makes the files larger
fukasawa e60969
    but faster to write and read.  pngstest now uses this by default; this
fukasawa e60969
    can be disabled with the --slow option.
fukasawa e60969
  Improved pngstest fine tuning of error numbers, new test file generator.
fukasawa e60969
    The generator generates images that test the full range of sample values,
fukasawa e60969
    allow the error numbers in pngstest to be tuned and checked.  makepng
fukasawa e60969
    also allows generation of images with extra chunks, although this is
fukasawa e60969
    still work-in-progress.
fukasawa e60969
  Added check for invalid palette index while reading.
fukasawa e60969
  Fixed some bugs in ICC profile writing. The code should now accept
fukasawa e60969
    all potentially valid ICC profiles and reject obviously invalid ones.
fukasawa e60969
    It now uses png_error() to do so rather than casually writing a PNG
fukasawa e60969
    without the necessary color data.
fukasawa e60969
  Removed whitespace from the end of lines in all source files and scripts.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta16 [March 6, 2012]
fukasawa e60969
  Relocated palette-index checking function from pngrutil.c to pngtrans.c
fukasawa e60969
  Added palette-index checking while writing.
fukasawa e60969
  Changed png_inflate() and calling routines to avoid overflow problems.
fukasawa e60969
    This is an intermediate check-in that solves the immediate problems and
fukasawa e60969
    introduces one performance improvement (avoiding a copy via png_ptr->zbuf.)
fukasawa e60969
    Further changes will be made to make ICC profile handling more secure.
fukasawa e60969
  Fixed build warnings (MSVC, GCC, GCC v3). Cygwin GCC with default options
fukasawa e60969
    declares 'index' as a global, causing a warning if it is used as a local
fukasawa e60969
    variable.  GCC 64-bit warns about assigning a (size_t) (unsigned 64-bit)
fukasawa e60969
    to an (int) (signed 32-bit).  MSVC, however, warns about using the
fukasawa e60969
    unary '-' operator on an unsigned value (even though it is well defined
fukasawa e60969
    by ANSI-C to be ~x+1).  The padding calculation was changed to use a
fukasawa e60969
    different method.  Removed the tests on png_ptr->pass.
fukasawa e60969
  Added contrib/libtests/tarith.c to test internal arithmetic functions from
fukasawa e60969
    png.c. This is a libpng maintainer program used to validate changes to the
fukasawa e60969
    internal arithmetic functions.
fukasawa e60969
  Made read 'inflate' handling like write 'deflate' handling. The read
fukasawa e60969
    code now claims and releases png_ptr->zstream, like the write code.
fukasawa e60969
    The bug whereby the progressive reader failed to release the zstream
fukasawa e60969
    is now fixed, all initialization is delayed, and the code checks for
fukasawa e60969
    changed parameters on deflate rather than always calling
fukasawa e60969
    deflatedEnd/deflateInit.
fukasawa e60969
  Validate the zTXt strings in pngvalid.
fukasawa e60969
  Added code to validate the windowBits value passed to deflateInit2().
fukasawa e60969
    If the call to deflateInit2() is wrong a png_warning will be issued
fukasawa e60969
    (in fact this is harmless, but the PNG data produced may be sub-optimal).
fukasawa e60969
fukasawa e60969
Version 1.6.0beta17 [March 10, 2012]
fukasawa e60969
  Fixed PNG_LIBPNG_BUILD_BASE_TYPE definition. 
fukasawa e60969
  Reject all iCCP chunks after the first, even if the first one is invalid.
fukasawa e60969
  Deflate/inflate was reworked to move common zlib calls into single
fukasawa e60969
    functions [rw]util.c.  A new shared keyword check routine was also added
fukasawa e60969
    and the 'zbuf' is no longer allocated on progressive read.  It is now
fukasawa e60969
    possible to call png_inflate() incrementally.  A warning is no longer
fukasawa e60969
    issued if the language tag or translated keyword in the iTXt chunk
fukasawa e60969
    has zero length.
fukasawa e60969
  If benign errors are disabled use maximum window on ancilliary inflate.
fukasawa e60969
    This works round a bug introduced in 1.5.4 where compressed ancillary
fukasawa e60969
    chunks could end up with a too-small windowBits value in the deflate
fukasawa e60969
    header.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta18 [March 16, 2012]
fukasawa e60969
  Issue a png_benign_error() instead of png_warning() about bad palette index.
fukasawa e60969
  In pngtest, treat benign errors as errors if "-strict" is present.
fukasawa e60969
  Fixed an off-by-one error in the palette index checking function.
fukasawa e60969
  Fixed a compiler warning under Cygwin (Windows-7, 32-bit system)
fukasawa e60969
  Revised example.c to put text strings in a temporary character array
fukasawa e60969
    instead of directly assigning string constants to png_textp members.
fukasawa e60969
    This avoids compiler warnings when -Wwrite-strings is enabled.
fukasawa e60969
  Added output flushing to aid debugging under Visual Studio. Unfortunately
fukasawa e60969
    this is necessary because the VS2010 output window otherwise simply loses
fukasawa e60969
    the error messages on error (they weren't flushed to the window before
fukasawa e60969
    the process exited, apparently!)
fukasawa e60969
  Added configuration support for benign errors and changed the read
fukasawa e60969
    default. Also changed some warnings in the iCCP and sRGB handling
fukasawa e60969
    from to benign errors. Configuration now makes read benign
fukasawa e60969
    errors warnings and write benign errors to errors by default (thus
fukasawa e60969
    changing the behavior on read).  The simplified API always forces
fukasawa e60969
    read benign errors to warnings (regardless of the system default, unless
fukasawa e60969
    this is disabled in which case the simplified API can't be built.)
fukasawa e60969
fukasawa e60969
Version 1.6.0beta19 [March 18, 2012]
fukasawa e60969
  Work around for duplicate row start calls; added warning messages.
fukasawa e60969
    This turns on PNG_FLAG_DETECT_UNINITIALIZED to detect app code that
fukasawa e60969
    fails to call one of the 'start' routines (not enabled in libpng-1.5
fukasawa e60969
    because it is technically an API change, since it did normally work
fukasawa e60969
    before.)  It also makes duplicate calls to png_read_start_row (an
fukasawa e60969
    internal function called at the start of the image read) benign, as
fukasawa e60969
    they were before changes to use png_inflate_claim. Somehow webkit is
fukasawa e60969
    causing this to happen; this is probably a mis-feature in the zlib
fukasawa e60969
    changes so this commit is only a work-round.
fukasawa e60969
  Removed erroneous setting of DETECT_UNINITIALIZED and added more
fukasawa e60969
    checks. The code now does a png_error if an attempt is made to do the
fukasawa e60969
    row initialization twice; this is an application error and it has
fukasawa e60969
    serious consequences because the transform data in png_struct is
fukasawa e60969
    changed by each call.
fukasawa e60969
  Added application error reporting and added chunk names to read
fukasawa e60969
    benign errors; also added --strict to pngstest - not enabled
fukasawa e60969
    yet because a warning is produced.
fukasawa e60969
  Avoid the double gamma correction warning in the simplified API.
fukasawa e60969
    This allows the --strict option to pass in the pngstest checks
fukasawa e60969
fukasawa e60969
Version 1.6.0beta20 [March 29, 2012]
fukasawa e60969
  Changed chunk handler warnings into benign errors, incrementally load iCCP
fukasawa e60969
  Added checksum-icc.c to contrib/tools
fukasawa e60969
  Prevent PNG_EXPAND+PNG_SHIFT doing the shift twice.
fukasawa e60969
  Recognize known sRGB ICC profiles while reading; prefer writing the
fukasawa e60969
    iCCP profile over writing the sRGB chunk, controlled by the
fukasawa e60969
    PNG_sRGB_PROFILE_CHECKS option.
fukasawa e60969
  Revised png_set_text_2() to avoid potential memory corruption (fixes
fukasawa e60969
    CVE-2011-3048, also known as CVE-2012-3425).
fukasawa e60969
fukasawa e60969
Version 1.6.0beta21 [April 27, 2012]
fukasawa e60969
  Revised scripts/makefile.darwin: use system zlib; remove quotes around
fukasawa e60969
    architecture list; add missing ppc architecture; add architecture options
fukasawa e60969
    to shared library link; don't try to create a shared lib based on missing
fukasawa e60969
    RELEASE variable.
fukasawa e60969
  Enable png_set_check_for_invalid_index() for both read and write.
fukasawa e60969
  Removed #ifdef PNG_HANDLE_AS_UNKNOWN_SUPPORTED in pngpriv.h around
fukasawa e60969
    declaration of png_handle_unknown().
fukasawa e60969
  Added -lssp_nonshared in a comment in scripts/makefile.freebsd
fukasawa e60969
    and changed deprecated NOOBJ and NOPROFILE to NO_OBJ and NO_PROFILE.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta22 [May 23, 2012]
fukasawa e60969
  Removed need for -Wno-cast-align with clang.  clang correctly warns on
fukasawa e60969
    alignment increasing pointer casts when -Wcast-align is passed. This
fukasawa e60969
    fixes the cases that clang warns about either by eliminating the
fukasawa e60969
    casts from png_bytep to png_uint_16p (pngread.c), or, for pngrutil.c
fukasawa e60969
    where the cast is previously verified or pngstest.c where it is OK, by
fukasawa e60969
    introducing new png_aligncast macros to do the cast in a way that clang
fukasawa e60969
    accepts.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta23 [June 6, 2012]
fukasawa e60969
  Revised CMakeLists.txt to not attempt to make a symlink under mingw.
fukasawa e60969
  Made fixes for new optimization warnings from gcc 4.7.0. The compiler
fukasawa e60969
    performs an optimization which is safe; however it then warns about it.
fukasawa e60969
    Changing the type of 'palette_number' in pngvalid.c removes the warning.
fukasawa e60969
  Do not depend upon a GCC feature macro being available for use in generating
fukasawa e60969
    the linker mapfile symbol prefix.
fukasawa e60969
  Improved performance of new do_check_palette_indexes() function (only
fukasawa e60969
    update the value when it actually increases, move test for whether
fukasawa e60969
    the check is wanted out of the function.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta24 [June 7, 2012]
fukasawa e60969
  Don't check palette indexes if num_palette is 0 (as it can be in MNG files).
fukasawa e60969
fukasawa e60969
Version 1.6.0beta25 [June 16, 2012]
fukasawa e60969
  Revised png_set_keep_unknown_chunks() so num_chunks < 0 means ignore all
fukasawa e60969
    unknown chunks and all known chunks except for IHDR, PLTE, tRNS, IDAT,
fukasawa e60969
    and IEND.  Previously it only meant ignore all unknown chunks, the
fukasawa e60969
    same as num_chunks == 0. Revised png_image_skip_unused_chunks() to
fukasawa e60969
    provide a list of chunks to be processed instead of a list of chunks to
fukasawa e60969
    ignore.  Revised contrib/gregbook/readpng2.c accordingly.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta26 [July 10, 2012]
fukasawa e60969
  Removed scripts/makefile.cegcc from the *.zip and *.7z distributions; it
fukasawa e60969
    depends on configure, which is not included in those archives.
fukasawa e60969
  Moved scripts/chkfmt to contrib/tools.
fukasawa e60969
  Changed "a+w" to "u+w" in Makefile.in to fix CVE-2012-3386.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta27 [August 11, 2012]
fukasawa e60969
  Do not compile PNG_DEPRECATED, PNG_ALLOC and PNG_PRIVATE when __GNUC__ < 3.
fukasawa e60969
  Do not use __restrict when GNUC is <= 3.1
fukasawa e60969
  Removed references to png_zalloc() and png_zfree() from the manual.
fukasawa e60969
  Fixed configurations where floating point is completely disabled.  Because
fukasawa e60969
    of the changes to support symbol prefixing PNG_INTERNAL_FUNCTION declares
fukasawa e60969
    floating point APIs during libpng builds even if they are completely
fukasawa e60969
    disabled. This requires the png floating point types (png_double*) to be
fukasawa e60969
    declared even though the functions are never actually defined.  This
fukasawa e60969
    change provides a dummy definition so that the declarations work, yet any
fukasawa e60969
    implementation will fail to compile because of an incomplete type.
fukasawa e60969
  Re-eliminated the use of strcpy() in pngtest.c.  An unncessary use of
fukasawa e60969
    strcpy() was accidentally re-introduced in libpng16; this change replaces
fukasawa e60969
    it with strncpy().
fukasawa e60969
  Eliminated use of png_sizeof(); use sizeof() instead.
fukasawa e60969
  Use a consistent style for (sizeof type) and (sizeof (array))
fukasawa e60969
  Cleanup of png_set_filler().  This function does very different things on
fukasawa e60969
    read and write.  In libpng 1.6 the two cases can be distinguished and
fukasawa e60969
    considerable code cleanup, and extra error checking, is possible.  This
fukasawa e60969
    makes calls on the write side that have no effect be ignored with a
fukasawa e60969
    png_app_error(), which can be disabled in the app using
fukasawa e60969
    png_set_benign_errors(), and removes the spurious use of usr_channels
fukasawa e60969
    on the read side.
fukasawa e60969
  Insist on autotools 1.12.1 for git builds because there are security issues
fukasawa e60969
    with 1.12 and insisting on anything less would allow 1.12 to be used.
fukasawa e60969
  Removed info_ptr->signature[8] from WRITE-only builds.
fukasawa e60969
  Add some conditions for compiling png_fixed().  This is a small function
fukasawa e60969
    but it requires "-lm" on some platforms.
fukasawa e60969
  Cause pngtest --strict to fail on any warning from libpng (not just errors)
fukasawa e60969
    and cause it not to fail at the comparison step if libpng lacks support
fukasawa e60969
    for writing chunks that it reads from the input (currently only implemented
fukasawa e60969
    for compressed text chunks).
fukasawa e60969
  Make all three "make check" test programs work without READ or WRITE support.
fukasawa e60969
    Now "make check" will succeed even if libpng is compiled with -DPNG_NO_READ
fukasawa e60969
    or -DPNG_NO_WRITE.  The tests performed are reduced, but the basic reading
fukasawa e60969
    and writing of a PNG file is always tested by one or more of the tests.
fukasawa e60969
  Consistently use strlen(), memset(), memcpy(), and memcmp() instead of the
fukasawa e60969
    png_strlen(), png_memset(), png_memcpy(), and png_memcmp() macros.
fukasawa e60969
  Removed the png_sizeof(), png_strlen(), png_memset(), png_memcpy(), and
fukasawa e60969
    png_memcmp() macros.
fukasawa e60969
  Work around gcc 3.x and Microsoft Visual Studio 2010 complaints. Both object
fukasawa e60969
    to the split initialization of num_chunks.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta28 [August 29, 2012]
fukasawa e60969
  Unknown handling fixes and clean up. This adds more correct option
fukasawa e60969
    control of the unknown handling, corrects the pre-existing bug where
fukasawa e60969
    the per-chunk 'keep' setting is ignored and makes it possible to skip
fukasawa e60969
    IDAT chunks in the sequential reader (broken in earlier 1.6 versions).
fukasawa e60969
    There is a new test program, test-unknown.c, which is a work in progress
fukasawa e60969
    (not currently part of the test suite).  Comments in the header files now
fukasawa e60969
    explain how the unknown handling works.
fukasawa e60969
  Allow fine grain control of unknown chunk APIs. This change allows
fukasawa e60969
    png_set_keep_unknown_chunks() to be turned off if not required and causes
fukasawa e60969
    both read and write to behave appropriately (on read this is only possible
fukasawa e60969
    if the user callback is used to handle unknown chunks).  The change
fukasawa e60969
    also removes the support for storing unknown chunks in the info_struct
fukasawa e60969
    if the only unknown handling enabled is via the callback, allowing libpng
fukasawa e60969
    to be configured with callback reading and none of the unnecessary code.
fukasawa e60969
  Corrected fix for unknown handling in pngtest. This reinstates the
fukasawa e60969
    libpng handling of unknown chunks other than vpAg and sTER (including
fukasawa e60969
    unsafe-to-copy chunks which were dropped before) and eliminates the
fukasawa e60969
    repositioning of vpAg and sTER in pngtest.png by changing pngtest.png
fukasawa e60969
    (so the chunks are where libpng would put them).
fukasawa e60969
  Added "tunknown" test and corrected a logic error in png_handle_unknown()
fukasawa e60969
    when SAVE support is absent.  Moved the shell test scripts for
fukasawa e60969
    contrib/libtests from the libpng top directory to contrib/libtests.
fukasawa e60969
    png_handle_unknown() must always read or skip the chunk, if
fukasawa e60969
    SAVE_UNKNOWN_CHUNKS is turned off *and* the application does not set
fukasawa e60969
    a user callback an unknown chunk will not be read, leading to a read
fukasawa e60969
    error, which was revealed by the "tunknown" test.
fukasawa e60969
  Cleaned up and corrected ICC profile handling.
fukasawa e60969
    contrib/libtests/makepng: corrected 'rgb' and 'gray' cases.  profile_error
fukasawa e60969
    messages could be truncated; made a correct buffer size calculation and
fukasawa e60969
    adjusted pngerror.c appropriately. png_icc_check_* checking improved;
fukasawa e60969
    changed the functions to receive the correct color type of the PNG on read
fukasawa e60969
    or write and check that it matches the color space of the profile (despite
fukasawa e60969
    what the comments said before, there is danger in assuming the app will
fukasawa e60969
    cope correctly with an RGB profile on a grayscale image and, since it
fukasawa e60969
    violates the PNG spec, allowing it is certain to produce inconsistent
fukasawa e60969
    app behavior and might even cause app crashes.) Check that profiles
fukasawa e60969
    contain the tags needed to process the PNG (tags all required by the ICC
fukasawa e60969
    spec). Removed unused PNG_STATIC from pngpriv.h.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta29 [September 4, 2012]
fukasawa e60969
  Fixed the simplified API example programs to add the *colormap parameter
fukasawa e60969
    to several of he API and improved the error message if the version field
fukasawa e60969
    is not set.
fukasawa e60969
  Added contrib/examples/* to the *.zip and *.7z distributions.
fukasawa e60969
  Updated simplified API synopses and description of the png_image structure
fukasawa e60969
    in the manual.
fukasawa e60969
  Made makepng and pngtest produce identical PNGs, add "--relaxed" option
fukasawa e60969
    to pngtest. The "--relaxed" option turns off the benign errors that are
fukasawa e60969
    enabled by default in pre-RC builds. makepng can now write ICC profiles
fukasawa e60969
    where the length has not been extended to a multiple of 4, and pngtest
fukasawa e60969
    now intercepts all libpng errors, allowing the previously-introduced
fukasawa e60969
    "--strict test" on no warnings to actually work.
fukasawa e60969
  Improved ICC profile handling including cHRM chunk generation and fixed
fukasawa e60969
    Cygwin+MSVC build errors. The ICC profile handling now includes more
fukasawa e60969
    checking.  Several errors that caused rejection of the profile are now
fukasawa e60969
    handled with a warning in such a way that the invalid profiles will be
fukasawa e60969
    read by default in release (but not pre-RC) builds but will not be
fukasawa e60969
    written by default.  The easy part of handling the cHRM chunk is written,
fukasawa e60969
    where the ICC profile contains the required data.  The more difficult
fukasawa e60969
    part plus guessing a gAMA value requires code to pass selected RGB values
fukasawa e60969
    through the profile.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta30 [October 24, 2012]
fukasawa e60969
  Changed ICC profile matrix/vector types to not depend on array type rules.
fukasawa e60969
    By the ANSI-C standard the new types should be identical to the previous
fukasawa e60969
    versions, and all known versions of gcc tested with the previous versions
fukasawa e60969
    except for GCC-4.2.1 work with this version.  The change makes the ANSI-C
fukasawa e60969
    rule that const applied to an array of elements applies instead to the
fukasawa e60969
    elements in the array moot by explicitly applying const to the base
fukasawa e60969
    elements of the png_icc_matrix and png_icc_vector types. The accidental
fukasawa e60969
    (harmless) 'const' previously applied to the parameters of two of the
fukasawa e60969
    functions have also been removed.
fukasawa e60969
  Added a work around for GCC 4.2 optimization bug.
fukasawa e60969
  Marked the broken (bad white point) original HP sRGB profiles correctly and
fukasawa e60969
    correct comments.
fukasawa e60969
  Added -DZ_SOLO to contrib/pngminim/*/makefile to work with zlib-1.2.7
fukasawa e60969
  Use /MDd for vstudio debug builds. Also added pngunkown to the vstudio
fukasawa e60969
    builds, fixed build errors and corrected a minor exit code error in
fukasawa e60969
    pngvalid if the 'touch' file name is invalid.
fukasawa e60969
  Add updated WARNING file to projects/vstudio from libpng 1.5/vstudio
fukasawa e60969
  Fixed build when using #define PNG_NO_READ_GAMMA in png_do_compose() in
fukasawa e60969
    pngrtran.c (Domani Hannes).
fukasawa e60969
fukasawa e60969
Version 1.6.0beta31 [November 1, 2012]
fukasawa e60969
  Undid the erroneous change to vstudio/pngvalid build in libpng-1.6.0beta30.
fukasawa e60969
  Made pngvalid so that it will build outside the libpng source tree.
fukasawa e60969
  Made builds -DPNG_NO_READ_GAMMA compile (the unit tests still fail).
fukasawa e60969
  Made PNG_NO_READ_GAMMA switch off interfaces that depend on READ_GAMMA.
fukasawa e60969
    Prior to 1.6.0 switching off READ_GAMMA did unpredictable things to the
fukasawa e60969
    interfaces that use it (specifically, png_do_background in 1.4 would
fukasawa e60969
    simply display composite for grayscale images but do composition
fukasawa e60969
    with the incorrect arithmetic for color ones). In 1.6 the semantic
fukasawa e60969
    of -DPNG_NO_READ_GAMMA is changed to simply disable any interface that
fukasawa e60969
    depends on it; this obliges people who set it to consider whether they
fukasawa e60969
    really want it off if they happen to use any of the interfaces in
fukasawa e60969
    question (typically most users who disable it won't).
fukasawa e60969
  Fixed GUIDs in projects/vstudio. Some were duplicated or missing,
fukasawa e60969
    resulting in VS2010 having to update the files.
fukasawa e60969
  Removed non-working ICC profile support code that was mostly added to
fukasawa e60969
    libpng-1.6.0beta29 and beta30. There was too much code for too little
fukasawa e60969
    gain; implementing full ICC color correction may be desireable but is left
fukasawa e60969
    up to applications.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta32 [November 25, 2012]
fukasawa e60969
  Fixed an intermittent SEGV in pngstest due to an uninitialized array element.
fukasawa e60969
  Added the ability for contrib/libtests/makepng.c to make a PNG with just one
fukasawa e60969
    color. This is useful for debugging pngstest color inaccuracy reports.
fukasawa e60969
  Fixed error checking in the simplified write API (Olaf van der Spek)
fukasawa e60969
  Made png_user_version_check() ok to use with libpng version 1.10.x and later.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta33 [December 15, 2012]
fukasawa e60969
  Fixed typo in png.c (PNG_SET_CHUNK_MALLOC_MAX should be PNG_CHUNK_MALLOC_MAX)
fukasawa e60969
    that causes the MALLOC_MAX limit not to work (John Bowler)
fukasawa e60969
  Change png_warning() to png_app_error() in pngwrite.c and comment the
fukasawa e60969
    fall-through condition.
fukasawa e60969
  Change png_warning() to png_app_warning() in png_write_tRNS().
fukasawa e60969
  Rearranged the ARM-NEON optimizations: Isolated the machine specific code
fukasawa e60969
    to the hardware subdirectory and added comments to pngrutil.c so that
fukasawa e60969
    implementors of other optimizations know what to do.
fukasawa e60969
  Fixed cases of unquoted DESTDIR in Makefile.am
fukasawa e60969
  Rebuilt Makefile.in, etc., with autoconf-2.69 and automake-1.12.5.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta34 [December 19, 2012]
fukasawa e60969
  Cleaned up whitespace in the synopsis portion of the manpage "libpng.3"
fukasawa e60969
  Disassembled the version number in scripts/options.awk (necessary for
fukasawa e60969
    building on SunOs).
fukasawa e60969
fukasawa e60969
Version 1.6.0beta35 [December 23, 2012]
fukasawa e60969
  Made default Zlib compression settings be configurable. This adds #defines to
fukasawa e60969
    pnglibconf.h to control the defaults.
fukasawa e60969
  Fixed Windows build issues, enabled ARM compilation. Various warnings issued
fukasawa e60969
    by earlier versions of GCC fixed for Cygwin and Min/GW (which both use old
fukasawa e60969
    GCCs.) ARM support is enabled by default in zlib.props (unsupported by
fukasawa e60969
    Microsoft) and ARM compilation is made possible by deleting the check for
fukasawa e60969
    x86. The test programs cannot be run because they are not signed.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta36 [January 2, 2013]
fukasawa e60969
  Discontinued distributing libpng-1.x.x.tar.bz2.
fukasawa e60969
  Discontinued distributing libpng-1.7.0-1.6.0-diff.txt and similar.
fukasawa e60969
  Rebuilt configure with autoconf-2.69 (inadvertently not done in beta33)
fukasawa e60969
  Fixed 'make distcheck' on SUN OS - libpng.so was not being removed
fukasawa e60969
fukasawa e60969
Version 1.6.0beta37 [January 10, 2013]
fukasawa e60969
  Fixed conceivable but difficult to repro overflow. Also added two test
fukasawa e60969
    programs to generate and test a PNG which should have the problem.
fukasawa e60969
fukasawa e60969
Version 1.6.0beta39 [January 19, 2013]
fukasawa e60969
  Again corrected attempt at overflow detection in png_set_unknown_chunks()
fukasawa e60969
  (CVE-2013-7353).  Added overflow detection in png_set_sPLT() and
fukasawa e60969
  png_set_text_2() (CVE-2013-7354).
fukasawa e60969
fukasawa e60969
Version 1.6.0beta40 [January 20, 2013]
fukasawa e60969
  Use consistent handling of overflows in text, sPLT and unknown png_set_* APIs
fukasawa e60969
fukasawa e60969
Version 1.6.0rc01 [January 26, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.0rc02 [February 4, 2013]
fukasawa e60969
  Added png_get_palette_max() function.
fukasawa e60969
fukasawa e60969
Version 1.6.0rc03 [February 5, 2013]
fukasawa e60969
  Fixed the png_get_palette_max API.
fukasawa e60969
fukasawa e60969
Version 1.6.0rc04 [February 7, 2013]
fukasawa e60969
  Turn serial tests back on (recently turned off by autotools upgrade).
fukasawa e60969
fukasawa e60969
Version 1.6.0rc05 [February 8, 2013]
fukasawa e60969
  Update manual about png_get_palette_max().
fukasawa e60969
fukasawa e60969
Version 1.6.0rc06 [February 9, 2013]
fukasawa e60969
  Fixed missing dependency in --prefix builds The intermediate
fukasawa e60969
    internal 'prefix.h' file can only be generated correctly after
fukasawa e60969
    pnglibconf.h, however the dependency was not in Makefile.am.  The
fukasawa e60969
    symptoms are unpredictable depending on the order make chooses to
fukasawa e60969
    build pngprefix.h and pnglibconf.h, often the error goes unnoticed
fukasawa e60969
    because there is a system pnglibconf.h to use instead.
fukasawa e60969
fukasawa e60969
Version 1.6.0rc07 [February 10, 2013]
fukasawa e60969
  Enclosed the new png_get_palette_max in #ifdef PNG_GET_PALETTE_MAX_SUPPORTED
fukasawa e60969
    block, and revised pnglibconf.h and pnglibconf.h.prebuilt accordingly.
fukasawa e60969
fukasawa e60969
Version 1.6.0rc08 [February 10, 2013]
fukasawa e60969
  Fix typo in png.h #ifdef
fukasawa e60969
fukasawa e60969
Version 1.6.0 [February 14, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.1beta01 [February 16, 2013]
fukasawa e60969
  Made symbol prefixing work with the ARM neon optimizations. Also allow
fukasawa e60969
    pngpriv.h to be included for preprocessor definitions only, so it can
fukasawa e60969
    be used in non-C/C++ files. Back ported from libpng 1.7.
fukasawa e60969
  Made sRGB check numbers consistent.
fukasawa e60969
  Ported libpng 1.5 options.awk/dfn file handling to 1.6, fixed one bug.
fukasawa e60969
  Removed cc -E workround, corrected png_get_palette_max API Tested on
fukasawa e60969
    SUN OS cc 5.9, which demonstrates the tokenization problem previously
fukasawa e60969
    avoided by using /lib/cpp.  Since all .dfn output is now protected in
fukasawa e60969
    double quotes unless it is to be macro substituted the fix should
fukasawa e60969
    work everywhere.
fukasawa e60969
  Enabled parallel tests - back ported from libpng-1.7.
fukasawa e60969
  scripts/pnglibconf.dfa formatting improvements back ported from libpng17.
fukasawa e60969
  Fixed a race condition in the creation of the build 'scripts' directory
fukasawa e60969
    while building with a parallel make.
fukasawa e60969
  Use approved/supported Android method to check for NEON, use Linux/POSIX
fukasawa e60969
    1003.1 API to check /proc/self/auxv avoiding buffer allocation and other
fukasawa e60969
    library calls (ported from libpng15).
fukasawa e60969
fukasawa e60969
Version 1.6.1beta02 [February 19, 2013]
fukasawa e60969
  Use parentheses more consistently in "#if defined(MACRO)" tests.
fukasawa e60969
  Folded long lines.
fukasawa e60969
  Reenabled code to allow zero length PLTE chunks for MNG.
fukasawa e60969
fukasawa e60969
Version 1.6.1beta03 [February 22, 2013]
fukasawa e60969
  Fixed ALIGNED_MEMORY support.
fukasawa e60969
  Added a new configure option:
fukasawa e60969
    --enable-arm-neon=always will stop the run-time checks. New checks
fukasawa e60969
    within arm/arm_init.c will cause the code not to be compiled unless
fukasawa e60969
    __ARM_NEON__ is set. This should make it fail safe (if someone asks
fukasawa e60969
    for it on then the build will fail if it can't be done.)
fukasawa e60969
  Updated the INSTALL document.
fukasawa e60969
fukasawa e60969
Version 1.6.1beta04 [February 27, 2013]
fukasawa e60969
  Revised INSTALL to recommend using CPPFLAGS instead of INCLUDES.
fukasawa e60969
  Revised scripts/makefile.freebsd to respect ZLIBLIB and ZLIBINC.
fukasawa e60969
  Revised scripts/dfn.awk to work with the buggy MSYS awk that has trouble
fukasawa e60969
    with CRLF line endings.
fukasawa e60969
fukasawa e60969
Version 1.6.1beta05 [March 1, 2013]
fukasawa e60969
  Avoid a possible memory leak in contrib/gregbook/readpng.c
fukasawa e60969
fukasawa e60969
Version 1.6.1beta06 [March 4, 2013]
fukasawa e60969
  Better documentation of unknown handling API interactions.
fukasawa e60969
  Corrected Android builds and corrected libpng.vers with symbol
fukasawa e60969
    prefixing.  It also makes those tests compile and link on Android.
fukasawa e60969
  Added an API png_set_option() to set optimization options externally,
fukasawa e60969
    providing an alternative and general solution for the non-portable
fukasawa e60969
    run-time tests used by the ARM Neon code, using the PNG_ARM_NEON option.
fukasawa e60969
  The order of settings vs options in pnglibconf.h is reversed to allow
fukasawa e60969
    settings to depend on options and options can now set (or override) the
fukasawa e60969
    defaults for settings.
fukasawa e60969
fukasawa e60969
Version 1.6.1beta07 [March 7, 2013]
fukasawa e60969
  Corrected simplified API default gamma for color-mapped output, added
fukasawa e60969
    a flag to change default. In 1.6.0 when the simplified API was used
fukasawa e60969
    to produce color-mapped output from an input image with no gamma
fukasawa e60969
    information the gamma assumed for the input could be different from
fukasawa e60969
    that assumed for non-color-mapped output.  In particular 16-bit depth
fukasawa e60969
    input files were assumed to be sRGB encoded, whereas in the 'direct'
fukasawa e60969
    case they were assumed to have linear data.  This was an error.  The
fukasawa e60969
    fix makes the simplified API treat all input files the same way and
fukasawa e60969
    adds a new flag to the png_image::flags member to allow the
fukasawa e60969
    application/user to specify that 16-bit files contain sRGB data
fukasawa e60969
    rather than the default linear.
fukasawa e60969
  Fixed bugs in the pngpixel and makepng test programs.
fukasawa e60969
fukasawa e60969
Version 1.6.1beta08 [March 7, 2013]
fukasawa e60969
  Fixed CMakelists.txt to allow building a single variant of the library
fukasawa e60969
    (Claudio Bley):
fukasawa e60969
  Introduced a PNG_LIB_TARGETS variable that lists all activated library
fukasawa e60969
    targets.  It is an error if this variable ends up empty, ie. you have
fukasawa e60969
    to build at least one library variant.
fukasawa e60969
  Made the *_COPY targets only depend on library targets actually being build.
fukasawa e60969
  Use PNG_LIB_TARGETS to unify a code path.
fukasawa e60969
  Changed the CREATE_SYMLINK macro to expect the full path to a file as the
fukasawa e60969
    first argument. When symlinking the filename component of that path is
fukasawa e60969
    determined and used as the link target.
fukasawa e60969
  Use copy_if_different in the CREATE_SYMLINK macro.
fukasawa e60969
fukasawa e60969
Version 1.6.1beta09 [March 13, 2013]
fukasawa e60969
  Eliminated two warnings from the Intel C compiler. The warnings are
fukasawa e60969
    technically valid, although a reasonable treatment of division would
fukasawa e60969
    show it to be incorrect.
fukasawa e60969
fukasawa e60969
Version 1.6.1rc01 [March 21, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.1 [March 28, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.2beta01 [April 14, 2013]
fukasawa e60969
  Updated documentation of 1.5.x to 1.6.x changes in iCCP chunk handling.
fukasawa e60969
  Fixed incorrect warning of excess deflate data. End condition - the
fukasawa e60969
    warning would be produced if the end of the deflate stream wasn't read
fukasawa e60969
    in the last row.  The warning is harmless.
fukasawa e60969
  Corrected the test on user transform changes on read. It was in the
fukasawa e60969
    png_set of the transform function, but that doesn't matter unless the
fukasawa e60969
    transform function changes the rowbuf size, and that is only valid if
fukasawa e60969
    transform_info is called.
fukasawa e60969
  Corrected a misplaced closing bracket in contrib/libtests/pngvalid.c
fukasawa e60969
    (Flavio Medeiros).
fukasawa e60969
  Corrected length written to uncompressed iTXt chunks (Samuli Suominen).
fukasawa e60969
    Bug was introduced in libpng-1.6.0.
fukasawa e60969
fukasawa e60969
Version 1.6.2rc01 [April 18, 2013]
fukasawa e60969
  Added contrib/tools/fixitxt.c, to repair the erroneous iTXt chunk length
fukasawa e60969
    written by libpng-1.6.0 and 1.6.1.
fukasawa e60969
  Disallow storing sRGB information when the sRGB is not supported.
fukasawa e60969
fukasawa e60969
Version 1.6.2rc02 [April 18, 2013]
fukasawa e60969
  Merge pngtest.c with libpng-1.7.0
fukasawa e60969
fukasawa e60969
Version 1.6.2rc03 [April 22, 2013]
fukasawa e60969
  Trivial spelling cleanup.
fukasawa e60969
fukasawa e60969
Version 1.6.2rc04 and 1.6.2rc05 [omitted]
fukasawa e60969
fukasawa e60969
Version 1.6.2rc06 [April 24, 2013]
fukasawa e60969
  Reverted to version 1.6.2rc03.  Recent changes to arm/neon support
fukasawa e60969
    have been ported to libpng-1.7.0beta09 and will reappear in version
fukasawa e60969
    1.6.3beta01.
fukasawa e60969
fukasawa e60969
Version 1.6.2 [April 25, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.3beta01 [April 25, 2013]
fukasawa e60969
  Revised stack marking in arm/filter_neon.S and configure.ac.
fukasawa e60969
  Ensure that NEON filter stuff is completely disabled when switched 'off'.
fukasawa e60969
    Previously the ARM NEON specific files were still built if the option
fukasawa e60969
    was switched 'off' as opposed to being explicitly disabled.
fukasawa e60969
fukasawa e60969
Version 1.6.3beta02 [April 26, 2013]
fukasawa e60969
  Test for 'arm*' not just 'arm' in the host_cpu configure variable.
fukasawa e60969
  Rebuilt the configure scripts.
fukasawa e60969
fukasawa e60969
Version 1.6.3beta03 [April 30, 2013]
fukasawa e60969
  Expanded manual paragraph about writing private chunks, particularly
fukasawa e60969
    the need to call png_set_keep_unknown_chunks() when writing them.
fukasawa e60969
  Avoid dereferencing NULL pointer possibly returned from
fukasawa e60969
    png_create_write_struct() (Andrew Church).
fukasawa e60969
fukasawa e60969
Version 1.6.3beta05 [May 9, 2013]
fukasawa e60969
  Calculate our own zlib windowBits when decoding rather than trusting the
fukasawa e60969
    CMF bytes in the PNG datastream.
fukasawa e60969
  Added an option to force maximum window size for inflating, which was
fukasawa e60969
    the behavior of libpng15 and earlier, via a new PNG_MAXIMUM_INFLATE_WINDOW
fukasawa e60969
    option for png_set_options().
fukasawa e60969
  Added png-fix-itxt and png-fix-too-far-back to the built programs and
fukasawa e60969
    removed warnings from the source code and timepng that are revealed as
fukasawa e60969
    a result.
fukasawa e60969
  Detect wrong libpng versions linked to png-fix-too-far-back, which currently
fukasawa e60969
    only works with libpng versions that can be made to reliably fail when
fukasawa e60969
    the deflate data contains an out-of-window reference.  This means only
fukasawa e60969
    1.6 and later.
fukasawa e60969
  Fixed gnu issues: g++ needs a static_cast, gcc 4.4.7 has a broken warning
fukasawa e60969
    message which it is easier to work round than ignore.
fukasawa e60969
  Updated contrib/pngminus/pnm2png.c (Paul Stewart):
fukasawa e60969
    Check for EOF
fukasawa e60969
    Ignore "#" delimited comments in input file to pnm2png.c.
fukasawa e60969
    Fixed whitespace handling
fukasawa e60969
    Added a call to png_set_packing()
fukasawa e60969
    Initialize dimension values so if sscanf fails at least we have known
fukasawa e60969
      invalid values.
fukasawa e60969
  Attempt to detect configuration issues with png-fix-too-far-back, which
fukasawa e60969
    requires both the correct libpng and the correct zlib to function
fukasawa e60969
    correctly.
fukasawa e60969
  Check ZLIB_VERNUM for mismatches, enclose #error in quotes
fukasawa e60969
  Added information in the documentation about problems with and fixes for
fukasawa e60969
    the bad CRC and bad iTXt chunk situations.
fukasawa e60969
fukasawa e60969
Version 1.6.3beta06 [May 12, 2013]
fukasawa e60969
  Allow contrib/pngminus/pnm2png.c to compile without WRITE_INVERT and
fukasawa e60969
    WRITE_PACK supported (writes error message that it can't read P1 or
fukasawa e60969
    P4 PBM files).
fukasawa e60969
  Improved png-fix-too-far-back usage message, added --suffix option.
fukasawa e60969
  Revised contrib/pngminim/*/makefile to generate pnglibconf.h with the
fukasawa e60969
    right zlib header files.
fukasawa e60969
  Separated CPPFLAGS and CFLAGS in contrib/pngminim/*/makefile
fukasawa e60969
fukasawa e60969
Version 1.6.3beta07 [June 8, 2013]
fukasawa e60969
  Removed a redundant test in png_set_IHDR().
fukasawa e60969
  Added set(CMAKE_CONFIGURATION_TYPES ...) to CMakeLists.txt (Andrew Hundt)
fukasawa e60969
  Deleted set(CMAKE_BUILD_TYPE) block from CMakeLists.txt
fukasawa e60969
  Enclose the prototypes for the simplified write API in
fukasawa e60969
    #ifdef PNG_STDIO_SUPPORTED/#endif
fukasawa e60969
  Make ARM NEON support work at compile time (not just configure time).
fukasawa e60969
    This moves the test on __ARM_NEON__ into pngconf.h to avoid issues when
fukasawa e60969
    using a compiler that compiles for multiple architectures at one time.
fukasawa e60969
  Removed PNG_FILTER_OPTIMIZATIONS and PNG_ARM_NEON_SUPPORTED from
fukasawa e60969
    pnglibconf.h, allowing more of the decisions to be made internally
fukasawa e60969
    (pngpriv.h) during the compile.  Without this, symbol prefixing is broken
fukasawa e60969
    under certain circumstances on ARM platforms.  Now only the API parts of
fukasawa e60969
    the optimizations ('check' vs 'api') are exposed in the public header files
fukasawa e60969
    except that the new setting PNG_ARM_NEON_OPT documents how libpng makes the
fukasawa e60969
    decision about whether or not to use the optimizations.
fukasawa e60969
  Protect symbol prefixing against CC/CPPFLAGS/CFLAGS useage.
fukasawa e60969
    Previous iOS/Xcode fixes for the ARM NEON optimizations moved the test
fukasawa e60969
    on __ARM_NEON__ from configure time to compile time.  This breaks symbol
fukasawa e60969
    prefixing because the definition of the special png_init_filter_functions
fukasawa e60969
    call was hidden at configure time if the relevant compiler arguments are
fukasawa e60969
    passed in CFLAGS as opposed to CC.  This change attempts to avoid all
fukasawa e60969
    the confusion that would result by declaring the init function even when
fukasawa e60969
    it is not used, so that it will always get prefixed.
fukasawa e60969
fukasawa e60969
Version 1.6.3beta08 [June 18, 2013]
fukasawa e60969
  Revised libpng.3 so that "doclifter" can process it.
fukasawa e60969
fukasawa e60969
Version 1.6.3beta09 [June 27, 2013]
fukasawa e60969
  Revised example.c to illustrate use of PNG_DEFAULT_sRGB and PNG_GAMMA_MAC_18
fukasawa e60969
    as parameters for png_set_gamma().  These have been available since
fukasawa e60969
    libpng-1.5.4.
fukasawa e60969
  Renamed contrib/tools/png-fix-too-far-back.c to pngfix.c and revised it
fukasawa e60969
    to check all compressed chunks known to libpng.
fukasawa e60969
fukasawa e60969
Version 1.6.3beta10 [July 5, 2013]
fukasawa e60969
  Updated documentation to show default behavior of benign errors correctly.
fukasawa e60969
  Only compile ARM code when PNG_READ_SUPPORTED is defined.
fukasawa e60969
  Fixed undefined behavior in contrib/tools/pngfix.c and added new strip
fukasawa e60969
    option. pngfix relied on undefined behavior and even a simple change from
fukasawa e60969
    gcc to g++ caused it to fail.  The new strip option 'unsafe' has been
fukasawa e60969
    implemented and is the default if --max is given.  Option names have
fukasawa e60969
    been clarified, with --strip=transform now stripping the bKGD chunk,
fukasawa e60969
    which was stripped previously with --strip=unused.
fukasawa e60969
  Added all documented chunk types to pngpriv.h
fukasawa e60969
  Unified pngfix.c source with libpng17.
fukasawa e60969
fukasawa e60969
Version 1.6.3rc01 [July 11, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.3 [July 18, 2013]
fukasawa e60969
  Revised manual about changes in iTXt chunk handling made in libpng-1.6.0.
fukasawa e60969
  Added "/* SAFE */" comments in pngrutil.c and pngrtran.c where warnings
fukasawa e60969
    may be erroneously issued by code-checking applications.
fukasawa e60969
fukasawa e60969
Version 1.6.4beta01 [August 21, 2013]
fukasawa e60969
  Added information about png_set_options() to the manual.
fukasawa e60969
  Delay calling png_init_filter_functions() until a row with nonzero filter
fukasawa e60969
    is found.
fukasawa e60969
fukasawa e60969
Version 1.6.4beta02 [August 30, 2013]
fukasawa e60969
  Fixed inconsistent conditional compilation of png_chunk_unknown_handling()
fukasawa e60969
    prototype, definition, and usage.  Made it depend on
fukasawa e60969
    PNG_HANDLE_AS_UNKNOWN_SUPPORTED everywhere.
fukasawa e60969
fukasawa e60969
Version 1.6.4rc01 [September 5, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.4 [September 12, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.5 [September 14, 2013]
fukasawa e60969
  Removed two stray lines of code from arm/arm_init.c.
fukasawa e60969
fukasawa e60969
Version 1.6.6 [September 16, 2013]
fukasawa e60969
  Removed two stray lines of code from arm/arm_init.c, again.
fukasawa e60969
fukasawa e60969
Version 1.6.7beta01 [September 30, 2013]
fukasawa e60969
  Revised unknown chunk code to correct several bugs in the NO_SAVE_/NO_WRITE
fukasawa e60969
    combination
fukasawa e60969
  Allow HANDLE_AS_UNKNOWN to work when other options are configured off. Also
fukasawa e60969
    fixed the pngminim makefiles to work when $(MAKEFLAGS) contains stuff
fukasawa e60969
    which terminates the make options (as by default in recent versions of
fukasawa e60969
    Gentoo).
fukasawa e60969
  Avoid up-cast warnings in pngvalid.c. On ARM the alignment requirements of
fukasawa e60969
    png_modifier are greater than that of png_store and as a consequence
fukasawa e60969
    compilation of pngvalid.c results in a warning about increased alignment
fukasawa e60969
    requirements because of the bare cast to (png_modifier*). The code is safe,
fukasawa e60969
    because the pointer is known to point to a stack allocated png_modifier,
fukasawa e60969
    but this change avoids the warning.
fukasawa e60969
  Fixed default behavior of ARM_NEON_API. If the ARM NEON API option was
fukasawa e60969
    compiled without the CHECK option it defaulted to on, not off.
fukasawa e60969
  Check user callback behavior in pngunknown.c. Previous versions compiled
fukasawa e60969
    if SAVE_UNKNOWN was not available but did nothing since the callback
fukasawa e60969
    was never implemented.
fukasawa e60969
  Merged pngunknown.c with 1.7 version and back ported 1.7 improvements/fixes
fukasawa e60969
fukasawa e60969
Version 1.6.7beta02 [October 12, 2013]
fukasawa e60969
  Made changes for compatibility with automake 1.14:
fukasawa e60969
    1) Added the 'compile' program to the list of programs that must be cleaned
fukasawa e60969
       in autogen.sh
fukasawa e60969
    2) Added 'subdir-objects' which causes .c files in sub-directories to be
fukasawa e60969
       compiled such that the corresponding .o files are also in the
fukasawa e60969
       sub-directory.  This is because automake 1.14 warns that the
fukasawa e60969
       current behavior of compiling to the top level directory may be removed
fukasawa e60969
       in the future.
fukasawa e60969
    3) Updated dependencies on pnglibconf.h to match the new .o locations and
fukasawa e60969
       added all the files in contrib/libtests and contrib/tools that depend
fukasawa e60969
       on pnglibconf.h
fukasawa e60969
    4) Added 'BUILD_SOURCES = pnglibconf.h'; this is the automake recommended
fukasawa e60969
       way of handling the dependencies of sources that are machine generated;
fukasawa e60969
       unfortunately it only works if the user does 'make all' or 'make check',
fukasawa e60969
       so the dependencies (3) are still required.
fukasawa e60969
  Cleaned up (char*) casts of zlib messages. The latest version of the Intel C
fukasawa e60969
    compiler complains about casting a string literal as (char*), so copied the
fukasawa e60969
    treatment of z_const from the library code into pngfix.c
fukasawa e60969
  Simplified error message code in pngunknown. The simplification has the
fukasawa e60969
    useful side effect of avoiding a bogus warning generated by the latest
fukasawa e60969
    version of the Intel C compiler (it objects to
fukasawa e60969
    condition ? string-literal : string-literal).
fukasawa e60969
  Make autogen.sh work with automake 1.13 as well as 1.14. Do this by always
fukasawa e60969
    removing the 1.14 'compile' script but never checking for it.
fukasawa e60969
fukasawa e60969
Version 1.6.7beta03 [October 19, 2013]
fukasawa e60969
  Added ARMv8 support (James Yu <james.yu at linaro.org>).  Added file
fukasawa e60969
    arm/filter_neon_intrinsics.c; enable with -mfpu=neon.
fukasawa e60969
  Revised pngvalid to generate size images with as many filters as it can
fukasawa e60969
    manage, limited by the number of rows.
fukasawa e60969
  Cleaned up ARM NEON compilation handling. The tests are now in pngpriv.h
fukasawa e60969
    and detect the broken GCC compilers.
fukasawa e60969
fukasawa e60969
Version 1.6.7beta04 [October 26, 2013]
fukasawa e60969
  Allow clang derived from older GCC versions to use ARM intrinsics. This
fukasawa e60969
    causes all clang builds that use -mfpu=neon to use the intrinsics code,
fukasawa e60969
    not the assembler code.  This has only been tested on iOS 7. It may be
fukasawa e60969
    necessary to exclude some earlier clang versions but this seems unlikely.
fukasawa e60969
  Changed NEON implementation selection mechanism. This allows assembler
fukasawa e60969
    or intrinsics to be turned on at compile time during the build by defining
fukasawa e60969
    PNG_ARM_NEON_IMPLEMENTATION to the correct value (2 or 1).  This macro
fukasawa e60969
    is undefined by default and the build type is selected in pngpriv.h.
fukasawa e60969
fukasawa e60969
Version 1.6.7rc01 [November 2, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.7rc02 [November 7, 2013]
fukasawa e60969
  Fixed #include in filter_neon_intrinsics.c and ctype macros. The ctype char
fukasawa e60969
    checking macros take an unsigned char argument, not a signed char.
fukasawa e60969
fukasawa e60969
Version 1.6.7 [November 14, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.8beta01 [November 24, 2013]
fukasawa e60969
  Moved prototype for png_handle_unknown() in pngpriv.h outside of
fukasawa e60969
    the #ifdef PNG_SET_UNKNOWN_CHUNKS_SUPPORTED/#endif block.
fukasawa e60969
  Added "-Wall" to CFLAGS in contrib/pngminim/*/makefile
fukasawa e60969
  Conditionally compile some unused functions reported by -Wall in
fukasawa e60969
    pngminim.
fukasawa e60969
  Fixed 'minimal' builds. Various obviously useful minimal configurations
fukasawa e60969
    don't build because of missing contrib/libtests test programs and
fukasawa e60969
    overly complex dependencies in scripts/pnglibconf.dfa. This change
fukasawa e60969
    adds contrib/conftest/*.dfa files that can be used in automatic build
fukasawa e60969
    scripts to ensure that these configurations continue to build.
fukasawa e60969
  Enabled WRITE_INVERT and WRITE_PACK in contrib/pngminim/encoder.
fukasawa e60969
  Fixed pngvalid 'fail' function declaration on the Intel C Compiler.
fukasawa e60969
    This reverts to the previous 'static' implementation and works round
fukasawa e60969
    the 'unused static function' warning by using PNG_UNUSED().
fukasawa e60969
fukasawa e60969
Version 1.6.8beta02 [November 30, 2013]
fukasawa e60969
  Removed or marked PNG_UNUSED some harmless "dead assignments" reported
fukasawa e60969
    by clang scan-build.
fukasawa e60969
  Changed tabs to 3 spaces in png_debug macros and changed '"%s"m'
fukasawa e60969
    to '"%s" m' to improve portability among compilers.
fukasawa e60969
  Changed png_free_default() to free() in pngtest.c
fukasawa e60969
fukasawa e60969
Version 1.6.8rc01 [December 12, 2013]
fukasawa e60969
  Tidied up pngfix inits and fixed pngtest no-write builds.
fukasawa e60969
fukasawa e60969
Version 1.6.8rc02 [December 14, 2013]
fukasawa e60969
  Handle zero-length PLTE chunk or NULL palette with png_error()
fukasawa e60969
    instead of png_chunk_report(), which by default issues a warning
fukasawa e60969
    rather than an error, leading to later reading from a NULL pointer
fukasawa e60969
    (png_ptr->palette) in png_do_expand_palette(). This is CVE-2013-6954
fukasawa e60969
    and VU#650142.  Libpng-1.6.1 through 1.6.7 are vulnerable.
fukasawa e60969
    Libpng-1.6.0 and earlier do not have this bug.
fukasawa e60969
fukasawa e60969
Version 1.6.8 [December 19, 2013]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.9beta01 [December 26, 2013]
fukasawa e60969
  Bookkeeping: Moved functions around (no changes). Moved transform
fukasawa e60969
    function definitions before the place where they are called so that
fukasawa e60969
    they can be made static. Move the intrapixel functions and the
fukasawa e60969
    grayscale palette builder out of the png?tran.c files. The latter
fukasawa e60969
    isn't a transform function and is no longer used internally, and the
fukasawa e60969
    former MNG specific functions are better placed in pngread/pngwrite.c
fukasawa e60969
  Made transform implementation functions static. This makes the internal
fukasawa e60969
    functions called by png_do_{read|write}_transformations static. On an
fukasawa e60969
    x86-64 DLL build (Gentoo Linux) this reduces the size of the text
fukasawa e60969
    segment of the DLL by 1208 bytes, about 0.6%. It also simplifies
fukasawa e60969
    maintenance by removing the declarations from pngpriv.h and allowing
fukasawa e60969
    easier changes to the internal interfaces.
fukasawa e60969
  Rebuilt configure scripts with automake-1.14.1 and autoconf-2.69
fukasawa e60969
    in the tar distributions.
fukasawa e60969
fukasawa e60969
Version 1.6.9beta02 [January 1, 2014]
fukasawa e60969
  Added checks for libpng 1.5 to pngvalid.c.  This supports the use of
fukasawa e60969
    this version of pngvalid in libpng 1.5
fukasawa e60969
  Merged with pngvalid.c from libpng-1.7 changes to create a single
fukasawa e60969
    pngvalid.c
fukasawa e60969
  Removed #error macro from contrib/tools/pngfix.c (Thomas Klausner).
fukasawa e60969
  Merged pngrio.c, pngtrans.c, pngwio.c, and pngerror.c with libpng-1.7.0
fukasawa e60969
  Merged libpng-1.7.0 changes to make no-interlace configurations work
fukasawa e60969
    with test programs.
fukasawa e60969
  Revised pngvalid.c to support libpng 1.5, which does not support the
fukasawa e60969
    PNG_MAXIMUM_INFLATE_WINDOW option, so #define it out when appropriate in
fukasawa e60969
    pngvalid.c
fukasawa e60969
  Allow unversioned links created on install to be disabled in configure.
fukasawa e60969
    In configure builds 'make install' changes/adds links like png.h
fukasawa e60969
    and libpng.a to point to the newly installed, versioned, files (e.g.
fukasawa e60969
    libpng17/png.h and libpng17.a). Three new configure options and some
fukasawa e60969
    rearrangement of Makefile.am allow creation of these links to be disabled.
fukasawa e60969
fukasawa e60969
Version 1.6.9beta03 [January 10, 2014]
fukasawa e60969
  Removed potentially misleading warning from png_check_IHDR().
fukasawa e60969
fukasawa e60969
Version 1.6.9beta04 [January 20, 2014]
fukasawa e60969
  Updated scripts/makefile.* to use CPPFLAGS (Cosmin).
fukasawa e60969
  Added clang attribute support (Cosmin).
fukasawa e60969
fukasawa e60969
Version 1.6.9rc01 [January 28, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.9rc02 [January 30, 2014]
fukasawa e60969
  Quiet an uninitialized memory warning from VC2013 in png_get_png().
fukasawa e60969
fukasawa e60969
Version 1.6.9 [February 6, 2014]
fukasawa e60969
fukasawa e60969
Version 1.6.10beta01 [February 9, 2014]
fukasawa e60969
  Backported changes from libpng-1.7.0beta30 and beta31:
fukasawa e60969
  Fixed a large number of instances where PNGCBAPI was omitted from
fukasawa e60969
    function definitions.
fukasawa e60969
  Added pngimage test program for png_read_png() and png_write_png()
fukasawa e60969
    with two new test scripts.
fukasawa e60969
  Removed dependence on !PNG_READ_EXPAND_SUPPORTED for calling
fukasawa e60969
    png_set_packing() in png_read_png().
fukasawa e60969
  Fixed combination of ~alpha with shift. On read invert alpha, processing
fukasawa e60969
    occurred after shift processing, which causes the final values to be
fukasawa e60969
    outside the range that should be produced by the shift. Reversing the
fukasawa e60969
    order on read makes the two transforms work together correctly and mirrors
fukasawa e60969
    the order used on write.
fukasawa e60969
  Do not read invalid sBIT chunks. Previously libpng only checked sBIT
fukasawa e60969
    values on write, so a malicious PNG writer could therefore cause
fukasawa e60969
    the read code to return an invalid sBIT chunk, which might lead to
fukasawa e60969
    application errors or crashes.  Such chunks are now skipped (with
fukasawa e60969
    chunk_benign_error).
fukasawa e60969
  Make png_read_png() and png_write_png() prototypes in png.h depend
fukasawa e60969
    upon PNG_READ_SUPPORTED and PNG_WRITE_SUPPORTED.
fukasawa e60969
  Support builds with unsupported PNG_TRANSFORM_* values.  All of the
fukasawa e60969
    PNG_TRANSFORM_* values are always defined in png.h and, because they
fukasawa e60969
    are used for both read and write in some cases, it is not reliable
fukasawa e60969
    to #if out ones that are totally unsupported. This change adds error
fukasawa e60969
    detection in png_read_image() and png_write_image() to do a
fukasawa e60969
    png_app_error() if the app requests something that cannot be done
fukasawa e60969
    and it adds corresponding code to pngimage.c to handle such options
fukasawa e60969
    by not attempting to test them.
fukasawa e60969
fukasawa e60969
Version 1.6.10beta02 [February 23, 2014]
fukasawa e60969
  Moved redefines of png_error(), png_warning(), png_chunk_error(),
fukasawa e60969
    and png_chunk_warning() from pngpriv.h to png.h to make them visible
fukasawa e60969
    to libpng-calling applications.
fukasawa e60969
  Moved OS dependent code from arm/arm_init.c, to allow the included
fukasawa e60969
    implementation of the ARM NEON discovery function to be set at
fukasawa e60969
    build-time and provide sample implementations from the current code in the
fukasawa e60969
    contrib/arm-neon subdirectory. The __linux__ code has also been changed to
fukasawa e60969
    compile and link on Android by using /proc/cpuinfo, and the old linux code
fukasawa e60969
    is in contrib/arm-neon/linux-auxv.c.  The new code avoids POSIX and Linux
fukasawa e60969
    dependencies apart from opening /proc/cpuinfo and is C90 compliant.
fukasawa e60969
  Check for info_ptr == NULL early in png_read_end() so we don't need to
fukasawa e60969
    run all the png_handle_*() and depend on them to return if info_ptr == NULL.
fukasawa e60969
    This improves the performance of png_read_end(png_ptr, NULL) and makes
fukasawa e60969
    it more robust against future programming errors.
fukasawa e60969
  Check for __has_extension before using it in pngconf.h, to
fukasawa e60969
    support older Clang versions (Jeremy Sequoia).
fukasawa e60969
  Treat CRC error handling with png_set_crc_action(), instead of with
fukasawa e60969
    png_set_benign_errors(), which has been the case since libpng-1.6.0beta18.
fukasawa e60969
  Use a user warning handler in contrib/gregbook/readpng2.c instead of default,
fukasawa e60969
    so warnings will be put on stderr even if libpng has CONSOLE_IO disabled.
fukasawa e60969
  Added png_ptr->process_mode = PNG_READ_IDAT_MODE in png_push_read_chunk
fukasawa e60969
    after recognizing the IDAT chunk, which avoids an infinite loop while
fukasawa e60969
    reading a datastream whose first IDAT chunk is of zero-length.
fukasawa e60969
    This fixes CERT VU#684412 and CVE-2014-0333.
fukasawa e60969
  Don't recognize known sRGB profiles as sRGB if they have been hacked,
fukasawa e60969
    but don't reject them and don't issue a copyright violation warning.
fukasawa e60969
fukasawa e60969
Version 1.6.10beta03 [February 25, 2014]
fukasawa e60969
  Moved some documentation from png.h to libpng.3 and libpng-manual.txt
fukasawa e60969
  Minor editing of contrib/arm-neon/README and contrib/examples/*.c
fukasawa e60969
fukasawa e60969
Version 1.6.10rc01 [February 27, 2014]
fukasawa e60969
  Fixed typos in the manual and in scripts/pnglibconf.dfa (CFLAGS -> CPPFLAGS
fukasawa e60969
    and PNG_USR_CONFIG -> PNG_USER_CONFIG).
fukasawa e60969
fukasawa e60969
Version 1.6.10rc02 [February 28, 2014]
fukasawa e60969
  Removed unreachable return statement after png_chunk_error()
fukasawa e60969
    in pngrutil.c
fukasawa e60969
fukasawa e60969
Version 1.6.10rc03 [March 4, 2014]
fukasawa e60969
  Un-deprecated png_data_freer().
fukasawa e60969
fukasawa e60969
Version 1.6.10 [March 6, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.11beta01 [March 17, 2014]
fukasawa e60969
  Use "if (value != 0)" instead of "if (value)" consistently.
fukasawa e60969
  Changed ZlibSrcDir from 1.2.5 to 1.2.8 in projects/vstudio.
fukasawa e60969
  Moved configuration information from the manual to the INSTALL file.
fukasawa e60969
fukasawa e60969
Version 1.6.11beta02 [April 6, 2014]
fukasawa e60969
  Removed #if/#else/#endif from inside two pow() calls in pngvalid.c because
fukasawa e60969
    they were handled improperly by Portland Group's PGI-14.1 - PGI-14.3
fukasawa e60969
    when using its "__builtin_pow()" function.
fukasawa e60969
  Silence 'unused parameter' build warnings (Cosmin Truta).
fukasawa e60969
  $(CP) is now used alongside $(RM_F).  Also, use 'copy' instead of 'cp'
fukasawa e60969
    where applicable, and applied other minor makefile changes (Cosmin).
fukasawa e60969
  Don't warn about invalid dimensions exceeding user limits (Cosmin).
fukasawa e60969
  Allow an easy replacement of the default pre-built configuration
fukasawa e60969
    header with a custom header, via the make PNGLIBCONF_H_PREBUILT
fukasawa e60969
    macro (Cosmin).
fukasawa e60969
fukasawa e60969
Version 1.6.11beta03 [April 6, 2014]
fukasawa e60969
  Fixed a typo in pngrutil.c, introduced in libpng-1.5.6, that interferes
fukasawa e60969
    with "blocky" expansion of sub-8-bit interlaced PNG files (Eric Huss).
fukasawa e60969
  Optionally use  __builtin_bswap16() in png_do_swap().
fukasawa e60969
fukasawa e60969
Version 1.6.11beta04 [April 19, 2014]
fukasawa e60969
  Made progressive reading of interlaced images consistent with the
fukasawa e60969
    behavior of the sequential reader and consistent with the manual, by
fukasawa e60969
    moving some code out of the PNG_READ_INTERLACING_SUPPORTED blocks. The
fukasawa e60969
    row_callback now receives the proper pass number and unexpanded rows, when
fukasawa e60969
    png_combine_row() isn't built or used, and png_set_interlace_handling()
fukasawa e60969
    is not called.
fukasawa e60969
  Allow PNG_sRGB_PROFILE_CHECKING = (-1) to mean no sRGB profile checking.
fukasawa e60969
fukasawa e60969
Version 1.6.11beta05 [April 26, 2014]
fukasawa e60969
  Do not reject ICC V2 profiles that lack padding (Kai-Uwe Behrmann).
fukasawa e60969
  Relocated closing bracket of the sRGB profile test loop to avoid getting
fukasawa e60969
    "Not recognizing known sRGB profile that has been edited" warning for
fukasawa e60969
    ICC V2 profiles that lack the MD5 signature in the profile header.
fukasawa e60969
fukasawa e60969
Version 1.6.11beta06 [May 19, 2014]
fukasawa e60969
  Added PNG_SKIP_sRGB_CHECK_PROFILE choice for png_set_option().
fukasawa e60969
fukasawa e60969
Version 1.6.11rc01 [May 27, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.11rc02 [June 3, 2014]
fukasawa e60969
  Test ZLIB_VERNUM instead of PNG_ZLIB_VERNUM in contrib/tools/pngfix.c
fukasawa e60969
fukasawa e60969
Version 1.6.11 [June 5, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.12rc01 [June 6, 2014]
fukasawa e60969
  Relocated new code from 1.6.11beta06 in png.c to a point after the
fukasawa e60969
    declarations (Max Stepin).
fukasawa e60969
fukasawa e60969
Version 1.6.12rc02 [June 7, 2014]
fukasawa e60969
  Changed file permissions of contrib/tools/intgamma.sh,
fukasawa e60969
    test-driver, and compile from 0644 to 0755 (Cosmin).
fukasawa e60969
fukasawa e60969
Version 1.6.12rc03 [June 8, 2014]
fukasawa e60969
  Ensure "__has_attribute()" macro exists before trying to use it with
fukasawa e60969
    old clang compilers (MacPorts Ticket #43939).
fukasawa e60969
fukasawa e60969
Version 1.6.12 [June 12, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.13beta01 [July 4, 2014]
fukasawa e60969
  Quieted -Wsign-compare and -Wclobber compiler warnings in
fukasawa e60969
    contrib/pngminus/*.c
fukasawa e60969
  Added "(void) png_ptr;" where needed in contrib/gregbook to quiet
fukasawa e60969
    compiler complaints about unused pointers.
fukasawa e60969
  Split a long output string in contrib/gregbook/rpng2-x.c.
fukasawa e60969
  Added "PNG_SET_OPTION" requirement for sRGB chunk support to pnglibconf.dfa,
fukasawa e60969
    Needed for write-only support (John Bowler).
fukasawa e60969
  Changed "if defined(__ARM_NEON__)" to
fukasawa e60969
    "if (defined(__ARM_NEON__) || defined(__ARM_NEON))" (James Wu).
fukasawa e60969
  Fixed clang no-warning builds: png_digit was defined but never used.
fukasawa e60969
    
fukasawa e60969
Version 1.6.13beta02 [July 21, 2014]
fukasawa e60969
  Fixed an incorrect separator ("/" should be "\") in scripts/makefile.vcwin32
fukasawa e60969
    (bug report from Wolfgang S. Kechel).  Bug was introduced in libpng-1.6.11.
fukasawa e60969
    Also fixed makefile.bc32, makefile.bor, makefile.msc, makefile.intel, and
fukasawa e60969
    makefile.tc3 similarly.
fukasawa e60969
fukasawa e60969
Version 1.6.13beta03 [August 3, 2014]
fukasawa e60969
  Removed scripts/makefile.elf. It has not worked since libpng-1.5.0beta14
fukasawa e60969
    due to elimination of the PNG_FUNCTION_EXPORT and PNG_DATA_EXPORT
fukasawa e60969
    definitions from pngconf.h.
fukasawa e60969
  Ensure that CMakeLists.txt makes the target "lib" directory before making
fukasawa e60969
    symbolic link into it (SourceForge bug report #226 by Rolf Timmermans).
fukasawa e60969
fukasawa e60969
Version 1.6.13beta04 [August 8, 2014]
fukasawa e60969
  Added opinion that the ECCN (Export Control Classification Number) for
fukasawa e60969
    libpng is EAR99 to the README file.
fukasawa e60969
  Eliminated use of "$<" in makefile explicit rules, when copying
fukasawa e60969
    $PNGLIBCONF_H_PREBUILT.  This does not work on some versions of make;
fukasawa e60969
    bug introduced in libpng version 1.6.11.
fukasawa e60969
fukasawa e60969
Version 1.6.13rc01 [August 14, 2014]
fukasawa e60969
  Made "ccopts" agree with "CFLAGS" in scripts/makefile.hp* and makefile.*sunu
fukasawa e60969
fukasawa e60969
Version 1.6.13 [August 21, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.14beta01 [September 14, 2014]
fukasawa e60969
  Guard usage of png_ptr->options with #ifdef PNG_SET_OPTION_SUPPORTED.
fukasawa e60969
  Do not build contrib/tools/pngfix.c when PNG_SETJMP_NOT_SUPPORTED,
fukasawa e60969
    to allow "make" to complete without setjmp support (bug report by
fukasawa e60969
    Claudio Fontana)
fukasawa e60969
  Add "#include <setjmp.h>" to contrib/tools/pngfix.c (John Bowler)
fukasawa e60969
fukasawa e60969
Version 1.6.14beta02 [September 18, 2014]
fukasawa e60969
  Use nanosleep() instead of usleep() in contrib/gregbook/rpng2-x.c
fukasawa e60969
    because usleep() is deprecated.
fukasawa e60969
  Define usleep() in contrib/gregbook/rpng2-x.c if not already defined
fukasawa e60969
    in unistd.h and nanosleep() is not available; fixes error introduced
fukasawa e60969
    in libpng-1.6.13.
fukasawa e60969
  Disable floating point exception handling in pngvalid.c when
fukasawa e60969
    PNG_FLOATING_ARITHMETIC is not supported (bug report by "zootus
fukasawa e60969
    at users.sourceforge.net").
fukasawa e60969
fukasawa e60969
Version 1.6.14beta03 [September 19, 2014]
fukasawa e60969
  Define FE_DIVBYZERO, FE_INVALID, and FE_OVERFLOW in pngvalid.c if not
fukasawa e60969
    already defined.  Revert floating point exception handling in pngvalid.c
fukasawa e60969
    to version 1.6.14beta01 behavior.
fukasawa e60969
fukasawa e60969
Version 1.6.14beta04 [September 27, 2014]
fukasawa e60969
  Fixed incorrect handling of the iTXt compression flag in pngrutil.c
fukasawa e60969
    (bug report by Shunsaku Hirata).  Bug was introduced in libpng-1.6.0.
fukasawa e60969
fukasawa e60969
Version 1.6.14beta05 [October 1, 2014]
fukasawa e60969
  Added "option READ_iCCP enables READ_COMPRESSED_TEXT" to pnglibconf.dfa
fukasawa e60969
fukasawa e60969
Version 1.6.14beta06 [October 5, 2014]
fukasawa e60969
  Removed unused "text_len" parameter from private function png_write_zTXt().
fukasawa e60969
  Conditionally compile some code in png_deflate_claim(), when
fukasawa e60969
    PNG_WARNINGS_SUPPORTED and PNG_ERROR_TEXT_SUPPORTED are disabled.
fukasawa e60969
  Replaced repeated code in pngpread.c with PNG_PUSH_SAVE_BUFFER_IF_FULL.
fukasawa e60969
  Added "chunk iTXt enables TEXT" and "chunk zTXt enables TEXT"
fukasawa e60969
    to pnglibconf.dfa.
fukasawa e60969
  Removed "option READ_COMPRESSED_TEXT enables READ_TEXT" from pnglibconf.dfa,
fukasawa e60969
    to make it possible to configure a libpng that supports iCCP but not TEXT.
fukasawa e60969
fukasawa e60969
Version 1.6.14beta07 [October 7, 2014]
fukasawa e60969
  Removed "option WRITE_COMPRESSED_TEXT enables WRITE_TEXT" from pnglibconf.dfa
fukasawa e60969
  Only mark text chunks as written after successfully writing them.
fukasawa e60969
fukasawa e60969
Version 1.6.14rc01 [October 15, 2014]
fukasawa e60969
  Fixed some typos in comments.
fukasawa e60969
fukasawa e60969
Version 1.6.14rc02 [October 17, 2014]
fukasawa e60969
  Changed png_convert_to_rfc_1123() to png_convert_to_rfc_1123_buffer()
fukasawa e60969
    in the manual, to reflect the change made in libpng-1.6.0.
fukasawa e60969
  Updated README file to explain that direct access to the png_struct
fukasawa e60969
    and info_struct members has not been permitted since libpng-1.5.0.
fukasawa e60969
fukasawa e60969
Version 1.6.14 [October 23, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.15beta01 [October 29, 2014]
fukasawa e60969
  Changed "if (!x)" to "if (x == 0)" and "if (x)" to "if (x != 0)"
fukasawa e60969
  Simplified png_free_data().
fukasawa e60969
  Added missing "ptr = NULL" after some instances of png_free().
fukasawa e60969
fukasawa e60969
Version 1.6.15beta02 [November 1, 2014]
fukasawa e60969
  Changed remaining "if (!x)" to "if (x == 0)" and "if (x)" to "if (x != 0)"
fukasawa e60969
fukasawa e60969
Version 1.6.15beta03 [November 3, 2014]
fukasawa e60969
  Added PNG_USE_ARM_NEON configuration flag (Marcin Juszkiewicz).
fukasawa e60969
fukasawa e60969
Version 1.6.15beta04 [November 4, 2014]
fukasawa e60969
  Removed new PNG_USE_ARM_NEON configuration flag and made a one-line
fukasawa e60969
    revision to configure.ac to support ARM on aarch64 instead (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.15beta05 [November 5, 2014]
fukasawa e60969
  Use png_get_libpng_ver(NULL) instead of PNG_LIBPNG_VER_STRING in
fukasawa e60969
    example.c, pngtest.c, and applications in the contrib directory.
fukasawa e60969
  Avoid out-of-bounds memory access in png_user_version_check().
fukasawa e60969
  Simplified and future-proofed png_user_version_check().
fukasawa e60969
  Fixed GCC unsigned int->float warnings. Various versions of GCC
fukasawa e60969
    seem to generate warnings when an unsigned value is implicitly
fukasawa e60969
    converted to double. This is probably a GCC bug but this change
fukasawa e60969
    avoids the issue by explicitly converting to (int) where safe.
fukasawa e60969
  Free all allocated memory in pngimage. The file buffer cache was left
fukasawa e60969
    allocated at the end of the program, harmless but it causes memory
fukasawa e60969
    leak reports from clang.
fukasawa e60969
  Fixed array size calculations to avoid warnings. At various points
fukasawa e60969
    in the code the number of elements in an array is calculated using
fukasawa e60969
    sizeof.  This generates a compile time constant of type (size_t) which
fukasawa e60969
    is then typically assigned to an (unsigned int) or (int). Some versions
fukasawa e60969
    of GCC on 64-bit systems warn about the apparent narrowing, even though
fukasawa e60969
    the same compiler does apparently generate the correct, in-range,
fukasawa e60969
    numeric constant.  This adds appropriate, safe, casts to make the
fukasawa e60969
    warnings go away.
fukasawa e60969
fukasawa e60969
Version 1.6.15beta06 [November 6, 2014]
fukasawa e60969
  Reverted use png_get_libpng_ver(NULL) instead of PNG_LIBPNG_VER_STRING
fukasawa e60969
    in the manual, example.c, pngtest.c, and applications in the contrib
fukasawa e60969
    directory.  It was incorrect advice.
fukasawa e60969
fukasawa e60969
Version 1.6.15beta07 [November 7, 2014]
fukasawa e60969
  Removed #ifdef PNG_16BIT_SUPPORTED/#endif around png_product2(); it is
fukasawa e60969
    needed by png_reciprocal2().
fukasawa e60969
  Added #ifdef PNG_16BIT_SUPPORTED/#endif around png_log16bit() and
fukasawa e60969
    png_do_swap().
fukasawa e60969
  Changed all "#endif /* PNG_FEATURE_SUPPORTED */" to "#endif /* FEATURE */"
fukasawa e60969
fukasawa e60969
Version 1.6.15beta08 [November 8, 2014]
fukasawa e60969
  More housecleaning in *.h
fukasawa e60969
fukasawa e60969
Version 1.6.15rc01 [November 13, 2014]
fukasawa e60969
fukasawa e60969
Version 1.6.15rc02 [November 14, 2014]
fukasawa e60969
  The macros passed in the command line to Borland make were ignored if
fukasawa e60969
    similarly-named macros were already defined in makefiles. This behavior
fukasawa e60969
    is different from POSIX make and other make programs.  Surround the
fukasawa e60969
    macro definitions with ifndef guards (Cosmin).
fukasawa e60969
fukasawa e60969
Version 1.6.15rc03 [November 16, 2014]
fukasawa e60969
  Added "-D_CRT_SECURE_NO_WARNINGS" to CFLAGS in scripts/makefile.vcwin32.
fukasawa e60969
  Removed the obsolete $ARCH variable from scripts/makefile.darwin.
fukasawa e60969
fukasawa e60969
Version 1.6.15 [November 20, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.16beta01 [December 14, 2014]
fukasawa e60969
  Added ".align 2" to arm/filter_neon.S to support old GAS assemblers that
fukasawa e60969
    don't do alignment correctly.
fukasawa e60969
  Revised Makefile.am and scripts/symbols.dfn to work with MinGW/MSYS
fukasawa e60969
    (Bob Friesenhahn).
fukasawa e60969
fukasawa e60969
Version 1.6.16beta02 [December 15, 2014]
fukasawa e60969
  Revised Makefile.am and scripts/*.dfn again to work with MinGW/MSYS;
fukasawa e60969
    renamed scripts/*.dfn to scripts/*.c (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.16beta03 [December 21, 2014]
fukasawa e60969
  Quiet a "comparison always true" warning in pngstest.c (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.16rc01 [December 21, 2014]
fukasawa e60969
  Restored a test on width that was removed from png.c at libpng-1.6.9
fukasawa e60969
    (Bug report by Alex Eubanks, CVE-2015-0973).
fukasawa e60969
fukasawa e60969
Version 1.6.16rc02 [December 21, 2014]
fukasawa e60969
  Undid the update to pngrutil.c in 1.6.16rc01.
fukasawa e60969
fukasawa e60969
Version 1.6.16rc03 [December 21, 2014]
fukasawa e60969
  Fixed an overflow in png_combine_row() with very wide interlaced images
fukasawa e60969
    (Bug report and fix by John Bowler, CVE-2014-9495).
fukasawa e60969
fukasawa e60969
Version 1.6.16 [December 22, 2014]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.17beta01 [January 29, 2015]
fukasawa e60969
  Removed duplicate PNG_SAFE_LIMITS_SUPPORTED handling from pngconf.h
fukasawa e60969
  Corrected the width limit calculation in png_check_IHDR().
fukasawa e60969
  Removed user limits from pngfix. Also pass NULL pointers to
fukasawa e60969
    png_read_row to skip the unnecessary row de-interlace stuff.
fukasawa e60969
  Added testing of png_set_packing() to pngvalid.c
fukasawa e60969
  Regenerated configure scripts in the *.tar distributions with libtool-2.4.4
fukasawa e60969
  Implement previously untested cases of libpng transforms in pngvalid.c
fukasawa e60969
  Fixed byte order in png_do_read_filler() with 16-bit input. Previously
fukasawa e60969
    the high and low bytes of the filler, from png_set_filler() or from
fukasawa e60969
    png_set_add_alpha(), were read in the wrong order.
fukasawa e60969
  Made the check for out-of-range values in png_set_tRNS() detect
fukasawa e60969
    values that are exactly 2^bit_depth, and work on 16-bit platforms.
fukasawa e60969
  Merged some parts of libpng-1.6.17beta01 and libpng-1.7.0beta47.
fukasawa e60969
  Added #ifndef __COVERITY__ where needed in png.c, pngrutil.c and
fukasawa e60969
    pngset.c to avoid warnings about dead code.
fukasawa e60969
  Added "& 0xff" to many instances of expressions that are typecast
fukasawa e60969
    to (png_byte), to avoid Coverity warnings.
fukasawa e60969
fukasawa e60969
Version 1.6.17beta02 [February 7, 2015]
fukasawa e60969
  Work around one more Coverity-scan dead-code warning.
fukasawa e60969
  Do not build png_product2() when it is unused.
fukasawa e60969
fukasawa e60969
Version 1.6.17beta03 [February 17, 2015]
fukasawa e60969
  Display user limits in the output from pngtest.
fukasawa e60969
  Eliminated the PNG_SAFE_LIMITS macro and restored the 1-million-column
fukasawa e60969
    and 1-million-row default limits in pnglibconf.dfa, that can be reset
fukasawa e60969
    by the user at build time or run time.  This provides a more robust
fukasawa e60969
    defense against DOS and as-yet undiscovered overflows.
fukasawa e60969
fukasawa e60969
Version 1.6.17beta04 [February 21, 2015]
fukasawa e60969
  Added PNG_WRITE_CUSTOMIZE_COMPRESSION_SUPPORTED macro, on by default.
fukasawa e60969
  Allow user to call png_get_IHDR() with NULL arguments (Reuben Hawkins).
fukasawa e60969
  Rebuilt configure scripts with automake-1.15 and libtool-2.4.6
fukasawa e60969
fukasawa e60969
Version 1.6.17beta05 [February 25, 2015]
fukasawa e60969
  Restored compiling of png_reciprocal2 with PNG_NO_16BIT.
fukasawa e60969
fukasawa e60969
Version 1.6.17beta06 [February 27, 2015]
fukasawa e60969
  Moved png_set_filter() prototype into a PNG_WRITE_SUPPORTED block
fukasawa e60969
    of png.h.
fukasawa e60969
  Avoid runtime checks when converting integer to png_byte with
fukasawa e60969
    Visual Studio (Sergey Kosarevsky)
fukasawa e60969
fukasawa e60969
Version 1.6.17rc01 [March 4, 2015]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.17rc02 [March 9, 2015]
fukasawa e60969
  Removed some comments that the configure script did not handle
fukasawa e60969
    properly from scripts/pnglibconf.dfa and pnglibconf.h.prebuilt.
fukasawa e60969
  Free the unknown_chunks structure even when it contains no data.
fukasawa e60969
fukasawa e60969
Version 1.6.17rc03 [March 12, 2015]
fukasawa e60969
  Updated CMakeLists.txt to add OSX framework, change YES/NO to ON/OFF
fukasawa e60969
    for consistency, and remove some useless tests (Alexey Petruchik).
fukasawa e60969
fukasawa e60969
Version 1.6.17rc04 [March 16, 2015]
fukasawa e60969
  Remove pnglibconf.h, pnglibconf.c, and pnglibconf.out instead of
fukasawa e60969
    pnglibconf.* in "make clean" (Cosmin).
fukasawa e60969
  Fix bug in calculation of maxbits, in png_write_sBIT, introduced
fukasawa e60969
    in libpng-1.6.17beta01 (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.17rc05 [March 21, 2015]
fukasawa e60969
  Define PNG_FILTER_* and PNG_FILTER_VALUE_* in png.h even when WRITE
fukasawa e60969
    is not supported (John Bowler).  This fixes an error introduced in
fukasawa e60969
    libpng-1.6.17beta06.
fukasawa e60969
  Reverted "& 0xff" additions of version 1.6.17beta01. Libpng passes
fukasawa e60969
    the Coverity scan without them.
fukasawa e60969
fukasawa e60969
Version 1.6.17rc06 [March 23, 2015]
fukasawa e60969
  Remove pnglibconf.dfn and pnglibconf.pre with "make clean".
fukasawa e60969
  Reformatted some "&0xff" instances to "& 0xff".
fukasawa e60969
  Fixed simplified 8-bit-linear to sRGB alpha. The calculated alpha
fukasawa e60969
    value was wrong.  It's not clear if this affected the final stored
fukasawa e60969
    value; in the obvious code path the upper and lower 8-bits of the
fukasawa e60969
    alpha value were identical and the alpha was truncated to 8-bits
fukasawa e60969
    rather than dividing by 257 (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.17 [March 26, 2015]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.18beta01 [April 1, 2015]
fukasawa e60969
  Removed PNG_SET_CHUNK_[CACHE|MALLOC]_LIMIT_SUPPORTED macros.  They
fukasawa e60969
    have been combined with PNG_SET_USER_LIMITS_SUPPORTED (resolves
fukasawa e60969
    bug report by Andrew Church).
fukasawa e60969
  Fixed rgb_to_gray checks and added tRNS checks to pngvalid.c.  This
fukasawa e60969
    fixes some arithmetic errors that caused some tests to fail on
fukasawa e60969
    some 32-bit platforms (Bug reports by Peter Breitenlohner [i686]
fukasawa e60969
    and Petr Gajdos [i586]).
fukasawa e60969
fukasawa e60969
Version 1.6.18beta02 [April 26, 2015]
fukasawa e60969
  Suppressed some warnings from the Borland C++ 5.5.1/5.82 compiler
fukasawa e60969
    (Bug report by Viktor Szakats).
fukasawa e60969
fukasawa e60969
Version 1.6.18beta03 [May 6, 2015]
fukasawa e60969
  Replaced "unexpected" with an integer (0xabadca11) in pngset.c
fukasawa e60969
    where a long was expected, to avoid a compiler warning when PNG_DEBUG > 1.
fukasawa e60969
  Added contrib/examples/simpleover.c, to demonstrate how to handle
fukasawa e60969
    alpha compositing of multiple images, using the "simplified API"
fukasawa e60969
    and an example PNG generation tool, contrib/examples/genpng.c
fukasawa e60969
    (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.18beta04 [May 20, 2015]
fukasawa e60969
  PNG_RELEASE_BUILD replaces tests where the code depended on the build base
fukasawa e60969
    type and can be defined on the command line, allowing testing in beta
fukasawa e60969
    builds (John Bowler).
fukasawa e60969
  Avoid Coverity issue 80858 (REVERSE NULL) in pngtest.c PNG_DEBUG builds.
fukasawa e60969
  Avoid a harmless potential integer overflow in png_XYZ_from_xy() (Bug
fukasawa e60969
    report from Christopher Ferris).
fukasawa e60969
fukasawa e60969
Version 1.6.18beta05 [May 31, 2015]
fukasawa e60969
  Backport filter selection code from libpng-1.7.0beta51, to combine
fukasawa e60969
    sub_row, up_row, avg_row, and paeth_row into try_row and tst_row.
fukasawa e60969
  Changed png_voidcast(), etc., to voidcast(), etc., in contrib/tools/pngfix.c
fukasawa e60969
    to avoid confusion with the libpng private macros.
fukasawa e60969
  Fixed old cut&paste bug in the weighted filter selection code in
fukasawa e60969
    pngwutil.c, introduced in libpng-0.95, March 1997.
fukasawa e60969
fukasawa e60969
Version 1.6.18beta06 [June 1, 2015]
fukasawa e60969
  Removed WRITE_WEIGHTED_FILTERED code, to save a few kbytes of the
fukasawa e60969
    compiled library size. It never worked properly and as far as we can
fukasawa e60969
    tell, no one uses it. The png_set_filter_heuristics() and
fukasawa e60969
    png_set_filter_heuristics_fixed() APIs are retained but deprecated
fukasawa e60969
    and do nothing.
fukasawa e60969
fukasawa e60969
Version 1.6.18beta07 [June 6, 2015]
fukasawa e60969
  Removed non-working progressive reader 'skip' function. This
fukasawa e60969
    function has apparently never been used. It was implemented
fukasawa e60969
    to support back-door modification of png_struct in libpng-1.4.x
fukasawa e60969
    but (because it does nothing and cannot do anything) was apparently
fukasawa e60969
    never tested (John Bowler).
fukasawa e60969
  Fixed cexcept.h in which GCC 5 now reports that one of the auto
fukasawa e60969
    variables in the Try macro needs to be volatile to prevent value
fukasawa e60969
    being lost over the setjmp (John Bowler).
fukasawa e60969
  Fixed NO_WRITE_FILTER and -Wconversion build breaks (John Bowler).
fukasawa e60969
  Fix g++ build breaks (John Bowler).
fukasawa e60969
  Quieted some Coverity issues in pngfix.c, png-fix-itxt.c, pngvalid.c,
fukasawa e60969
    pngstest.c, and pngimage.c. Most seem harmless, but png-fix-itxt
fukasawa e60969
    would only work with iTXt chunks with length 255 or less.
fukasawa e60969
  Added #ifdef's to contrib/examples programs so people don't try
fukasawa e60969
    to compile them without the minimum required support enabled
fukasawa e60969
    (suggested by Flavio Medeiros).
fukasawa e60969
fukasawa e60969
Version 1.6.18beta08 [June 30, 2015]
fukasawa e60969
  Eliminated the final two Coverity defects (insecure temporary file
fukasawa e60969
    handling in contrib/libtests/pngstest.c; possible overflow of
fukasawa e60969
    unsigned char in contrib/tools/png-fix-itxt.c). To use the "secure"
fukasawa e60969
    file handling, define PNG_USE_MKSTEMP, otherwise "tmpfile()" will
fukasawa e60969
    be used.
fukasawa e60969
  Removed some unused WEIGHTED_FILTER macros from png.h and pngstruct.h
fukasawa e60969
fukasawa e60969
Version 1.6.18beta09 [July 5, 2015]
fukasawa e60969
  Removed some useless typecasts from contrib/tools/png-fix-itxt.c
fukasawa e60969
  Fixed a new signed-unsigned comparison in pngrtran.c (Max Stepin).
fukasawa e60969
  Replaced arbitrary use of 'extern' with #define PNG_LINKAGE_*.  To
fukasawa e60969
    preserve API compatibility, the new defines all default to "extern"
fukasawa e60969
    (requested by Jan Nijtmans).
fukasawa e60969
fukasawa e60969
Version 1.6.18rc01 [July 9, 2015]
fukasawa e60969
  Belatedly added Mans Rullgard and James Yu to the list of Contributing
fukasawa e60969
    Authors.
fukasawa e60969
fukasawa e60969
Version 1.6.18rc02 [July 12, 2015]
fukasawa e60969
  Restored unused FILTER_HEURISTIC macros removed at libpng-1.6.18beta08
fukasawa e60969
    to png.h to avoid compatibility warnings.
fukasawa e60969
fukasawa e60969
Version 1.6.18rc03 [July 15, 2015]
fukasawa e60969
  Minor changes to the man page
fukasawa e60969
fukasawa e60969
Version 1.6.18 [July 23, 2015]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.19beta01 [July 30, 2015]
fukasawa e60969
  Updated obsolete information about the simplified API macros in the
fukasawa e60969
    manual pages (Bug report by Arc Riley).
fukasawa e60969
  Avoid potentially dereferencing NULL info_ptr in png_info_init_3().
fukasawa e60969
  Rearranged png.h to put the major sections in the same order as
fukasawa e60969
    in libpng17.
fukasawa e60969
  Eliminated unused PNG_COST_SHIFT, PNG_WEIGHT_SHIFT, PNG_COST_FACTOR, and
fukasawa e60969
    PNG_WEIGHT_FACTOR macros.
fukasawa e60969
  Suppressed some warnings from the Borland C++ 5.5.1/5.82 compiler
fukasawa e60969
    (Bug report by Viktor Szakats).  Several warnings remain and are
fukasawa e60969
    unavoidable, where we test for overflow.
fukasawa e60969
  Fixed potential leak of png_pixels in contrib/pngminus/pnm2png.c
fukasawa e60969
  Fixed uninitialized variable in contrib/gregbook/rpng2-x.c
fukasawa e60969
fukasawa e60969
Version 1.6.19beta02 [August 19, 2015]
fukasawa e60969
  Moved config.h.in~ from the "libpng_autotools_files" list to the
fukasawa e60969
    "libpng_autotools_extra" list in autogen.sh because it was causing a
fukasawa e60969
    false positive for missing files (bug report by Robert C. Seacord).
fukasawa e60969
  Removed unreachable "break" statements in png.c, pngread.c, and pngrtran.c
fukasawa e60969
    to suppress clang warnings (Bug report by Viktor Szakats).
fukasawa e60969
  Fixed some bad links in the man page.
fukasawa e60969
  Changed "n bit" to "n-bit" in comments.
fukasawa e60969
  Added signed/unsigned 16-bit safety net. This removes the dubious
fukasawa e60969
    0x8000 flag definitions on 16-bit systems. They aren't supported
fukasawa e60969
    yet the defs *probably* work, however it seems much safer to do this
fukasawa e60969
    and be advised if anyone, contrary to advice, is building libpng 1.6
fukasawa e60969
    on a 16-bit system. It also adds back various switch default clauses
fukasawa e60969
    for GCC; GCC errors out if they are not present (with an appropriately
fukasawa e60969
    high level of warnings).
fukasawa e60969
  Safely convert num_bytes to a png_byte in png_set_sig_bytes() (Robert
fukasawa e60969
    Seacord).
fukasawa e60969
  Fixed the recently reported 1's complement security issue by replacing
fukasawa e60969
    the value that is illegal in the PNG spec, in both signed and unsigned
fukasawa e60969
    values, with 0. Illegal unsigned values (anything greater than or equal
fukasawa e60969
    to  0x80000000) can still pass through, but since these are not illegal
fukasawa e60969
    in ANSI-C (unlike 0x80000000 in the signed case) the checking that
fukasawa e60969
    occurs later can catch them (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.19beta03 [September 26, 2015]
fukasawa e60969
  Fixed png_save_int_32 when int is not 2's complement (John Bowler).
fukasawa e60969
  Updated libpng16 with all the recent test changes from libpng17,
fukasawa e60969
    including changes to pngvalid.c to ensure that the original,
fukasawa e60969
    distributed, version of contrib/visupng/cexcept.h can be used
fukasawa e60969
    (John Bowler).
fukasawa e60969
  pngvalid contains the correction to the use of SAVE/STORE_
fukasawa e60969
    UNKNOWN_CHUNKS; a bug revealed by changes in libpng 1.7. More
fukasawa e60969
    tests contain the --strict option to detect warnings and the
fukasawa e60969
    pngvalid-standard test has been corrected so that it does not
fukasawa e60969
    turn on progressive-read. There is a separate test which does
fukasawa e60969
    that. (John Bowler)
fukasawa e60969
  Also made some signed/unsigned fixes.
fukasawa e60969
  Make pngstest error limits version specific. Splitting the machine
fukasawa e60969
    generated error structs out to a file allows the values to be updated
fukasawa e60969
    without changing pngstest.c itself. Since libpng 1.6 and 1.7 have
fukasawa e60969
    slightly different error limits this simplifies maintenance. The
fukasawa e60969
    makepngs.sh script has also been updated to more accurately reflect
fukasawa e60969
    current problems in libpng 1.7 (John Bowler).
fukasawa e60969
  Incorporated new test PNG files into make check.  tests/pngstest-*
fukasawa e60969
    are changed so that the new test files are divided into 8 groups by
fukasawa e60969
    gamma and alpha channel.  These tests have considerably better code
fukasawa e60969
    and pixel-value coverage than contrib/pngsuite; however,coverage is
fukasawa e60969
    still incomplete (John Bowler).
fukasawa e60969
  Removed the '--strict' in 1.6 because of the double-gamma-correction
fukasawa e60969
    warning, updated pngstest-errors.h for the errors detected with the
fukasawa e60969
    new contrib/testspngs PNG test files (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.19beta04 [October 15, 2015]
fukasawa e60969
  Worked around rgb-to-gray issues in libpng 1.6.  The previous
fukasawa e60969
    attempts to ignore the errors in the code aren't quite enough to
fukasawa e60969
    deal with the 'channel selection' encoding added to libpng 1.7; abort.
fukasawa e60969
    pngvalid.c is changed to drop this encoding in prior versions.
fukasawa e60969
  Fixed 'pow' macros in pngvalid.c. It is legal for 'pow' to be a
fukasawa e60969
    macro, therefore the argument list cannot contain preprocessing
fukasawa e60969
    directives.  Make sure pow is a function where this happens. This is
fukasawa e60969
    a minimal safe fix, the issue only arises in non-performance-critical
fukasawa e60969
    code (bug report by Curtis Leach, fix by John Bowler).
fukasawa e60969
  Added sPLT support to pngtest.c
fukasawa e60969
fukasawa e60969
Version 1.6.19rc01 [October 23, 2015]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.19rc02 [October 31, 2015]
fukasawa e60969
  Prevent setting or writing over-length PLTE chunk (Cosmin Truta).
fukasawa e60969
  Silently truncate over-length PLTE chunk while reading.
fukasawa e60969
  Libpng incorrectly calculated the output rowbytes when the application
fukasawa e60969
    decreased either the number of channels or the bit depth (or both) in
fukasawa e60969
    a user transform.  This was safe; libpng overallocated buffer space
fukasawa e60969
   (potentially by quite a lot; up to 4 times the amount required) but,
fukasawa e60969
   from 1.5.4 on, resulted in a png_error (John Bowler).
fukasawa e60969
fukasawa e60969
Version 1.6.19rc03 [November 3, 2015]
fukasawa e60969
  Fixed some inconsequential cut-and-paste typos in png_set_cHRM_XYZ_fixed().
fukasawa e60969
  Clarified COPYRIGHT information to state explicitly that versions
fukasawa e60969
    are derived from previous versions.
fukasawa e60969
  Removed much of the long list of previous versions from png.h and
fukasawa e60969
    libpng.3.
fukasawa e60969
fukasawa e60969
Version 1.6.19rc04 [November 5, 2015]
fukasawa e60969
  Fixed new bug with CRC error after reading an over-length palette
fukasawa e60969
    (bug report by Cosmin Truta) (CVE-2015-8126).
fukasawa e60969
fukasawa e60969
Version 1.6.19 [November 12, 2015]
fukasawa e60969
  Cleaned up coding style in png_handle_PLTE().
fukasawa e60969
fukasawa e60969
Version 1.6.20beta01 [November 20, 2015]
fukasawa e60969
  Avoid potential pointer overflow/underflow in png_handle_sPLT() and
fukasawa e60969
    png_handle_pCAL() (Bug report by John Regehr).
fukasawa e60969
fukasawa e60969
Version 1.6.20beta02 [November 23, 2015]
fukasawa e60969
  Fixed incorrect implementation of png_set_PLTE() that uses png_ptr
fukasawa e60969
    not info_ptr, that left png_set_PLTE() open to the CVE-2015-8126
fukasawa e60969
    vulnerability.  Fixes CVE-2015-8472.
fukasawa e60969
fukasawa e60969
Version 1.6.20beta03 [November 24, 2015]
fukasawa e60969
  Backported tests from libpng-1.7.0beta69.
fukasawa e60969
fukasawa e60969
Version 1.6.20rc01 [November 26, 2015]
fukasawa e60969
  Fixed an error in handling of bad zlib CMINFO field in pngfix, found by
fukasawa e60969
    American Fuzzy Lop, reported by Brian Carpenter.  inflate() doesn't
fukasawa e60969
    immediately fault a bad CMINFO field; instead a 'too far back' error
fukasawa e60969
    happens later (at least some times).  pngfix failed to limit CMINFO to
fukasawa e60969
    the allowed values but then assumed that window_bits was in range,
fukasawa e60969
    triggering an assert. The bug is mostly harmless; the PNG file cannot
fukasawa e60969
    be fixed.
fukasawa e60969
fukasawa e60969
Version 1.6.20rc02 [November 29, 2015]
fukasawa e60969
  In libpng 1.6 zlib initialization was changed to use the window size
fukasawa e60969
    in the zlib stream, not a fixed value. This causes some invalid images,
fukasawa e60969
    where CINFO is too large, to display 'correctly' if the rest of the
fukasawa e60969
    data is valid.  This provides a workaround for zlib versions where the
fukasawa e60969
    error arises (ones that support the API change to use the window size
fukasawa e60969
    in the stream).
fukasawa e60969
fukasawa e60969
Version 1.6.20 [December 3, 2015]
fukasawa e60969
  No changes.
fukasawa e60969
fukasawa e60969
Version 1.6.21beta01 [December 11, 2015]
fukasawa e60969
  Fixed syntax "$(command)" in tests/pngstest that some shells other than
fukasawa e60969
    bash could not parse (Bug report by Nelson Beebe). Use `command` instead.
fukasawa e60969
fukasawa e60969
Version 1.6.21beta02 [December 14, 2015]
fukasawa e60969
  Moved png_check_keyword() from pngwutil.c to pngset.c
fukasawa e60969
  Removed LE/BE dependencies in pngvalid, to 'fix' the current problem
fukasawa e60969
    in the BigEndian tests by not testing it, making the BE code the same 
fukasawa e60969
    as the LE version.
fukasawa e60969
  Fixes to pngvalid for various reduced build configurations (eliminate unused
fukasawa e60969
    statics) and a fix for the case in rgb_to_gray when the digitize option
fukasawa e60969
    reduces graylo to 0, producing a large error.
fukasawa e60969
fukasawa e60969
Version 1.6.21beta03 [December 18, 2015]
fukasawa e60969
  Widened the 'limit' check on the internally calculated error limits in
fukasawa e60969
    the 'DIGITIZE' case (the code used prior to 1.7 for rgb_to_gray error
fukasawa e60969
    checks) and changed the check to only operate in non-release builds
fukasawa e60969
    (base build type not RC or RELEASE.)
fukasawa e60969
  Fixed undefined behavior in pngvalid.c, undefined because
fukasawa e60969
    (png_byte) << shift is undefined if it changes the signed bit
fukasawa e60969
    (because png_byte is promoted to int). The libpng exported functions
fukasawa e60969
    png_get_uint_32 and png_get_uint_16 handle this. (Bug reported by
fukasawa e60969
    David Drysdale as a result of reports from UBSAN in clang 3.8).
fukasawa e60969
  This changes pngvalid to use BE random numbers; this used to produce
fukasawa e60969
    errors but these should not be fixed as a result of the previous changes.
fukasawa e60969
fukasawa e60969
Version 1.6.21rc01 [January 4, 2016]
fukasawa e60969
  In projects/vstudio, combined readme.txt and WARNING into README.txt
fukasawa e60969
fukasawa e60969
Version 1.6.21rc02 [January 7, 2016]
fukasawa e60969
  Relocated assert() in contrib/tools/pngfix.c, bug found by American
fukasawa e60969
    Fuzzy Lop, reported by Brian Carpenter.
fukasawa e60969
  Marked 'limit' UNUSED in transform_range_check().  This only affects
fukasawa e60969
    release builds.
fukasawa e60969
fukasawa e60969
Version 1.6.21 [January 15, 2016]
fukasawa e60969
  Worked around a false-positive Coverity issue in pngvalid.c.
fukasawa e60969
fukasawa e60969
Send comments/corrections/commendations to png-mng-implement at lists.sf.net
fukasawa e60969
(subscription required; visit
fukasawa e60969
https://lists.sourceforge.net/lists/listinfo/png-mng-implement
fukasawa e60969
to subscribe)
fukasawa e60969
or to glennrp at users.sourceforge.net
fukasawa e60969
fukasawa e60969
Glenn R-P
fukasawa e60969
#endif