correction.

on a variety of images, limited documentation (in the header file.)

The logical shift fix for Microsoft Visual C is required by other compilers,
so this enables that fix for all compilers when using compile-time constants.
Under MSYS 'byte' is a name declared in a system header file, so we
changed the name of a local variable to avoid the warnings that result.

These fixes attend to most of the errors revealed in pngvalid, however doing
the gamma work twice results in inaccuracies that can't be easily fixed.
There is now a warning in the code if this is going to happen.

The 16-bit cases still use "L" as we don't have a 16-bit test system.

and fixed spelling (truely -> truly) in comments.

being applied.

pngvalid still fails two tests, problem with row buffer.

If this is not enabled, png_set_strip_16() and png_do_scale_16_to_8()
aren't built.

png_set_scale_16_to_8 from pngpriv.h

    1.5.3beta06
  Also removed a spurious (totally unused and confusing) member from png_info.
  The palette expand optimization prevented expansion to an intermediate RGBA
    form if tRNS was present but alpha was marked to be stripped; this exposed
    a check for tRNS in png_do_expand_palette() which is inconsistent with the
    code elsewhere in libpng.

by disabling PNG_CHOP_16_TO_8_SUPPORTED

called.

Fixed some compiliation problems with scaling options.

to get the same (inaccurate) output as libpng-1.5.2 and earlier.

    wrong (high by one) 25% of the time. Dividing by 257 with rounding is
    wrong in 128 out of 65536 cases. Getting the right answer all the time
    without division is easy.

 This problem is present in 1.5.2; if png_set_background is called with
 need_expand false and the matching 16 bit color libpng erroneously just treats
 it as an 8-bit color because of where png_do_expand_16 is in the transform
 list.  This simple fix reduces the supplied colour to 8 bits, so it gets
 smashed, but this is better than the current behavior.