Re: [PNG] Meeting minutes - Mar 31st, 2025

Quick update:

It looks like parallel decoding already has both a testable implementation
and wide adoption through Apple's iDOT.
Test implementation: https://github.com/bvibber/mtpng
iDOT:
https://www.hackerfactor.com/blog/index.php?/archives/895-Connecting-the-iDOTs.html
and https://github.com/w3c/png/issues/54#issuecomment-1654886603

I suggest we read the GitHub issues on this topic (#60
<https://github.com/w3c/png/issues/60>, #54
<https://github.com/w3c/png/issues/54>, and #41
<https://github.com/w3c/png/issues/41>) and we reconsider it at our next
meeting.
Specifically, #54 <https://github.com/w3c/png/issues/54> has 3 separate
proposals in it.

If I get to it, I'll try to implement [one/all of them] as a libpng fork.

On Mon, Mar 31, 2025 at 12:08 PM Chris Blume (ProgramMax) <
programmax@gmail.com> wrote:

> Attached is a PDF of the meeting minutes from Mar 31st, 2025.
>
> Executive summary:
>
>    - Alternative compression methods -- We are open to the idea. But we
>    want to be mindful about it. Rather than blindly drop other compression
>    methods into the standard and wait (hope) for adoption, we want to
>    experiment and gather user interest.
>       - Unlikely to make it into Fourth Edition due to experiment & user
>       adoption time. Possibly by Fifth Edition.
>    - Negative cHRM values -- General agreement. We expect old decoders to
>    ignore the chunk when it believes there is an error due to the bit being
>    set. This would result in the image still being displayed but colors being
>    off. That is the same behavior as an old decoder encountering a cICP chunk.
>    - Deflate restart markers -- We are open to the idea. But we need a
>    reference implementation and user demand (IE that implementation being
>    used) before we add it to the spec.
>    - There were other topics for Fourth Edition, but we ran out of time.
>    We'll resume those next meeting.
>
> Our next meeting is Apr 14th, 2025.
> See you then!
>