Changes to Libpng from version 1.2.43 to 1.4.1 (February 25, 2010) 

Libpng-1.4.1 was released by the PNG Development Group on February
25, 2010.  Important new features include support for the iTXt chunk
and a function for limiting the amount of memory that a possibly
malicious compressed chunk can consume.  Some long-deprecated functions
and features have been removed.  A new private header file that is
not visible to applications has been created, to improve our ability
to maintain binary compatibility among future libpng versions.

Specifically,

1. Private libpng prototypes and macro definitions were moved from
png.h and pngconf.h into a new pngpriv.h header file that is not
visible to applications.

2. Some obsolete/deprecated macros and functions have been removed.

   a. We eliminated the obsolete, unused pnggccrd.c and pngvcrd.c files
      and related makefiles.

   b. We removed the typecasted NULL definitions such as

          #define png_voidp_NULL    (png_voidp)NULL

      If you used these in your application, just use NULL instead.

   c. We eliminated the PNG_1_0_X and PNG_1_2_X macros.

   d. We eliminated the PNG_LEGACY_SUPPORTED macro.

   e. We removed many WIN32_WCE #ifdefs.

   f. We removed the functions png_read_init(info_ptr),
      png_write_init(info_ptr), png_info_init(info_ptr),
      png_read_destroy(), and png_write_destroy().
      They have been deprecated since libpng-0.95.

   g. We removed the png_permit_empty_plte() function. It has been
      deprecated since libpng-1.0.9.  Use png_permit_mng_features() instead.

   h. We removed the obsolete stub functions png_get_mmx_flagmask(),
      png_set_mmx_thresholds(), png_get_asm_flags(),
      png_get_mmx_bitdepth_threshold(), png_get_mmx_rowbytes_threshold(),
      png_set_asm_flags(), and png_mmx_supported()

   i. We removed the obsolete png_check_sig(), png_memcpy_check(), and
      png_memset_check() functions.  Instead use !png_sig_cmp(),
      png_memcpy(), and png_memset(), respectively, i.e.,

      replace
        png_check_sig(buf, 8)
      with
        png_sig_cmp(buf, 0, 8) == 0

      replace
        png_memcpy_check(png_ptr, dest, src, length)
      with
        png_memcpy(dest, src, length)

      and replace
        png_memset_check(png_ptr, dest, value, length)
      with
        png_memset(dest, value, length)

   j. We removed support for global arrays to improve thread safety.

   k. We removed support for numbered error messages by default, since we
      never got around to actually numbering the error messages. The function
      png_set_strip_error_numbers() was removed from the library by default.

   l. The png_zalloc() and png_zfree() functions are no longer exported.
      The png_zalloc() function no longer zeroes out the memory that it
      allocates.

   m. The function png_set_gray_1_2_4_to_8() was removed. It has been
      deprecated since libpng-1.0.18 and 1.2.9, when it was replaced with
      png_set_expand_gray_1_2_4_to_8() because the former function also
      expanded palette images.

3. Several function prototypes were changed.

   a. We changed the prototype for png_malloc() from
      png_malloc(png_structp png_ptr, png_uint_32 size) to
      png_malloc(png_structp png_ptr, png_alloc_size_t size)

   b. We changed the prototypes of png_get_compression_buffer_size() and
      png_set_compression_buffer_size() to work with png_size_t instead of
      png_uint_32.

4. We added some new features:

   a. We enabled iTXt chunk support by default.  This changes the png_text
      structure by adding the itxt_length, lang, and lang_key members.

   b. We added the functions png_set_benign_errors(), png_benign_error(),
      and png_chunk_benign_error().  With these, you can choose to treat
      certain normally fatal errors as nonfatal.

   c. Support for setting the maximum number of ancillary chunks that the
      application will store and the amount of memory that the application
      will allocate for reading chunks was added, as a security measure.
      The functions png_set_chunk_cache_max(), png_get_chunk_cache_max()
      png_set_chunk_malloc_max(), and png_get_chunk_malloc_max()
      were added to the library. Also, in libpng-1.4.1, support was
      added for setting default limits via macros PNG_USER_CHUNK_CACHE_MAX
      and PNG_USER_CHUNK_MALLOC_MAX. These are set to 0, meaning unlimited,
      in the libpng distribution.

   d. We added the png_calloc() function and used it in place of
      of "png_malloc(); png_memset();" except in the case in png_read_png()
      where the array consists of pointers; in this case a "for" loop is used
      after the png_malloc() to set the pointers to NULL, to give robust.
      behavior in case the application runs out of memory part-way through
      the process.

   e. We implemented support for I/O states by adding png_ptr member io_state
      and functions png_get_io_chunk_name() and png_get_io_state() in pngget.c

   f. We added PNG_TRANSFORM_GRAY_TO_RGB to the available high-level
      input transforms.

5. Other changes

   a. We changed the names of the png_struct and info_struct members "trans"
      and "trans_values" to "trans_alpha" and "trans_color", respectively.

   b. Checking for and reporting of errors in the IHDR chunk is more thorough.

   c. We removed the trailing '.' from the warning and error messages.

   d. Direct access to png_ptr->jmpbuf has been deprecated since libpng
      version 1.0.6, and libpng now generates a warning about it.

      To avoid such warnings, change
        setjmp(png_ptr->jmpbuf)

      to
        setjmp(png_jmpbuf(png_ptr))

   e. In libpng-1.4.1, and backported to libpng-1.2.43 and 1.0.53, the
      png_decompress_chunk() function is much more efficient in terms of
      CPU time and memory allocated.


6. Building and linking with the libraries

    The configure script and most of the makefiles will build

    libpng10   (libpng-1.0.52, provided for legacy 1.0.x applications)
    libpng12   (libpng-1.2.42, stable version, for recent legacy 1.2.x apps)
    libpng14   (libpng-1.4.0, the current version)
    libpng     (whichever of the above was built last!)

    You should build them in order
       1) libpng10 (if you must; you probably don't need libpng10)
       2) libpng12 (overwrites the libpng created in step 1)
       3) libpng14 (overwrites the libpng created in step 2)

    Optionally, if you aren't ready to switch, you can build them
    in order
       1) libpng14
       2) libpng12 (this will overwrite the libpng created in step 1)

    If you simply want to link with the last version built, use
    -lpng -lz when building your application.  If you need a specific
    version, use, e.g., -lpng12 -lz.  Most applications can probably
    simply use -lpng -lz.

Send comments to png-mng-implement at lists.sf.net (subscription required;
visit https://lists.sourceforge.net/lists/listinfo/png-mng-implement
to subscribe) or to glennrp at users.sourceforge.net