Re: CALL for DISCUSSION: eXIf 20170115

Hi,

Sorry for the MIA, was super-busy...

I cannot read the CFD text at simplesystems.org, is the site down?
Will try again tomorrow. I am slowly catching up with the CFD emails.

I have half-implemented the EXIF extraction in the existing,
custom-made TIFF-to-PNG converter inside OptiPNG. I will ping you
again when it's fully done.

One thing that I was planning to do is implement, experimentally, both
the plain raw uncompressed form (temporarily called "uxIf") and the
deflate-compressed form (temporarily called "zxIf"), and report on the
implementation effort.

On the topic of me catching up with the discussion, I still can't find
the point where the deflate encoding was dropped. Last time I was
participating, Glenn made a compelling point in its favor, we also
established that deflate-stored uncompressed chunks (including ICCP
and EXIF) are trivially doable, John demanded an actual example with 3
function calls, and I eventually beat that with 2 calls max (and I
posted sample code on this list). So can someone please summarize it
for me?

In my implementation so far, I confirm Glenn's findings, cutting the
EXIF chunk size in about half with compression.

The work that we're doing right here is in direct competition with
TIFF, which, for now, is the well-established, industry-standard image
format. Every piece of PNG metadata of non-trivial size (if
compressible) has built-in compression by design, and I would really
appreciate understanding the reason why we are now giving up this
practice and competitive advantage.

Sincerely,
Cosmin

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