Re: CALL for DISCUSSION: eXIf 20170115

On Wed, Feb 1, 2017 at 7:19 AM, Pavel Zlatovratskii <[email protected]> wrote:
> Using pngexif in real image processing encourage users to ignore iCCP,
> gAMA, pHYs and may be even size from IHDR in favour of EXIF data (e.g.
> if user have some tool to parse EXIF).

And such a tool (one that parses EXIF and produces the corresponding
PNG chunks) is potentially very useful.

Part of this is the argument about whether the chunk is safe-to-copy;
back at the start of January I started with safe-to-copy and then
changed it to unsafe-to-copy when I realized the potential for
accidental inclusion of conflicting information.  Cosmin argued
vociferously to make it safe-to-copy and, because at the start the
file *only* has EXIF *or* PNG specific information, I was ok with it
either way.

The other part is that code which reads colorspace information from a
PNG file has to be careful about potential conflicts.  For example
iCCP overrides both gAMA and cHRM.

This was discussed before and Cosmin argued that the EXIF data was
essentially historical (at least this is what I understood him to be
saying) and, on reflection, it seemed to me that this was
self-evident.  It's not a chicken and egg problem; we know the EXIF
came first, therefore if someone puts a gAMA chunk in there a
reasonable programmer will realize that, unlike iCCP, the gAMA chunk
is a modification not extra information.

This is expressed concisely in the current proposal; the programmer
only needs to read a couple of paragraphs.  I don't think we can do
better than that.

John Bowler

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot