Re: DRAFT: eXIf 2017-03-02

John Bowler <[email protected]> Fri, 3 Mar 2017 07:50:00 -0800
Newsgroups gmane.comp.graphics.png.general
Message-ID <CAP7U39-gUrb6ACZuU6B==zBa4+F-H_svBtPXkAje2XCD51x_vg@mail.gmail.com>
On Fri, Mar 3, 2017 at 4:28 AM, Glenn Randers-Pehrson <[email protected]> wrote:
> So I suppose we should allow an empty eXIf chunk.

I see no reason to duplicate parts of the EXIF specification inside
the PNG specification, and I see every reason *not* to do so:

I don't see anything in the CIPA spec which permits the first IFD to
be omitted, or identifies it by the presence of EXIF or GPS offset
tags.  Maybe it is in there somewhere; the thing is 186 pages long and
the translation from Japanese is clearly wrong.  Consider "0th" (i.e.
"first") and 1st (should be "second"); see page 17.

I also don't see why I shouldn't just put the EXIF tags into IFD[0];
the level of indirection is of no benefit in a single-image format.
Maybe CIPA2016 intends to mandate that they appear there, but I can't
be sure and, anyway, CIPA may change the spec.

I also don't see any reason why IFD[0] must contain either EXIF or GPS
offset tags; it seems perfectly possible for both to be empty.  Take a
look at Tables 17 and 18.  I know it says that the EXIF offset is
mandatory, but if you don't put "Copyright holder" into IFD[0] then
all the other tags in Table 17 are part of TIFF, not EXIF.

-- 
John Bowler <[email protected]>
+1 (541) 450-9885
PO BOX 3151
KERBY OR 97531-3151
USA

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