Re: [PNG] Cancel May 16th, 2022 meeting?

Ah, okay. I think I follow what you're saying now. Let me see if I can
summarize it.

We can embed iccMAX inside the v4 profile, providing backward compatibility.
The down side is the file contains data for two profiles.

During the transition period, people would want to use both profiles
anyway. So this is fine.

Once someone decides they are past the tipping point where the v4 data is
not worth the extra space / bandwidth (reminder: transferred before pixel
data) they may want an iCCM chunk.


If I understand correctly, that means we might not need to specify the iCCM
chunk right away. It is only useful once iccMAX is well adopted, right?

On Mon, May 16, 2022 at 10:22 AM Leonard Rosenthol <lrosenth@adobe.com>
wrote:

> Unversioned ICC is **not** recommended – which I believe we called out in
> that embedding doc (which I co-authored).
>
>
>
> If we want to support both v4 and v5 profiles natively, then yes, we would
> want two separate chunks – iCCP and iCCM.  Alternatively, we only continue
> to support v4 and anyone wanting to use iccMAX embeds it in the v4.
>
>
>
> Leonard
>
>
>
> *From: *Chris Blume (ProgramMax) <programmax@gmail.com>
> *Date: *Monday, May 16, 2022 at 10:19 AM
> *To: *Leonard Rosenthol <lrosenth@adobe.com>
> *Cc: *Pierre-Anthony Lemieux <pal@sandflow.com>, public-png@w3.org <
> public-png@w3.org>
> *Subject: *Re: [PNG] Cancel May 16th, 2022 meeting?
>
> There is an ICC Technical Note about how to embed like that:
> https://www.color.org/technotes/Embedding_an_ICC.2_profile_in_an_ICC.1_profile.pdf
> <https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.color.org%2Ftechnotes%2FEmbedding_an_ICC.2_profile_in_an_ICC.1_profile.pdf&data=05%7C01%7Clrosenth%40adobe.com%7Cd8af81b3e8ec410a958308da374705de%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C637883075557864151%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=I8c4aqjqqQs%2B20%2B%2BSQp2plidCl78SNrJ%2BvB%2BkEleLQo%3D&reserved=0>
> It sounds like it allows a safe fallback into v4 when the decoder is not
> aware of iccMAX.
>
> This means it already incorporates a transition period option.
> (A PNG could contain both iCCP and iCCM chunks until enough viewers
> support iCCM. Similarly, the ICC profile could contain the v4 profile and
> the embedded iccMAX profile until enough viewers support iccMAX.)
>
>
> According to another ICC Technical Note (
> https://www.color.org/technotes/ICC-Technote-ProfileEmbedding.pdf
> <https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.color.org%2Ftechnotes%2FICC-Technote-ProfileEmbedding.pdf&data=05%7C01%7Clrosenth%40adobe.com%7Cd8af81b3e8ec410a958308da374705de%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C637883075557864151%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=pgeJ%2FppQK8nCGhAQbSDZpet%2BCjWJ02HfwvYkDLcubD0%3D&reserved=0>),
> the formats TIFF, TIFF-EP, and WebP contain a versionless ICC spec. We were
> already considering the iCCP chunk allowing both v2 and v4. It sounds like
> allowing iccMAX into iCCP might work for us AND not be unheard of.
>
>
>
> On Mon, May 16, 2022 at 8:24 AM Leonard Rosenthol <lrosenth@adobe.com>
> wrote:
>
> I think having two separate chunks for ICC color seems like a really bad
> idea, especially since there is a standard way to embed iccMAX data into a
> v4 profile.
>
>
>
> Leonard
>
>
>
> *From: *Chris Blume (ProgramMax) <programmax@gmail.com>
> *Date: *Monday, May 16, 2022 at 8:15 AM
> *To: *Pierre-Anthony Lemieux <pal@sandflow.com>
> *Cc: *public-png@w3.org <public-png@w3.org>
> *Subject: *Re: [PNG] Cancel May 16th, 2022 meeting?
>
> Hello Pierre-Anthony,
>
> For iCCN, Chris Seeger has a patch out:
> https://github.com/w3c/PNG-spec/pull/117
> <https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fw3c%2FPNG-spec%2Fpull%2F117&data=05%7C01%7Clrosenth%40adobe.com%7Cd8af81b3e8ec410a958308da374705de%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C637883075557864151%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=gK%2FoFCZAmy%2B7FikI7hr0tbWYdhR5yjloRtNkdwfzFU0%3D&reserved=0>
> I'm helping him squash it, as per a review comment. (We ran into a problem
> where Git really does not want to squash commits.)
> Additionally, I need to go back through Color on the Web Community Group
> history and convince myself that iCCM (iccMAX) is indeed an option for us.
> That will end up changing the iCCN patch.
>
> So right now, I think I am the holdup. I'll try to unblock.
>
>
> I do not think all "3rd edition" or even "2nd edition errata" issues are
> blocking. For example, under "3rd edition" is an issue "Update figures".
> That issue wants to both modernize some dated figure designs and fix text
> clipping. (I looked into it and do not see a way in SVG to prevent that
> text clipping.) The figures are usable as-is so this isn't a blocking
> issue. This issue is more of a "nice to have".
>
> I think the only work left blocking the 3rd edition draft is the iCCN/iCCM
> and ReSpec error cleanup.
>
>
> Once I unblock the iCCN/iCCM patch, I'll update the issues with a more
> clear blocking/nice-to-have indicator.
>
>
>
> On Mon, May 16, 2022 at 12:35 AM Pierre-Anthony Lemieux <pal@sandflow.com>
> wrote:
>
> Hi Chris,
>
> Nothing urgent from me.
>
> What is blocking https://github.com/w3c/PNG-spec/issues/95
> <https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fw3c%2FPNG-spec%2Fissues%2F95&data=05%7C01%7Clrosenth%40adobe.com%7Cd8af81b3e8ec410a958308da374705de%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C637883075557864151%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=G%2FdjbZMak2NBMkIfxJnYZ%2FPhXODjtc3DwauL%2FQU7Bf8%3D&reserved=0>
> ?
>
> Are all issues scheduled against "3rd edition" milestone [1] blocking?
>
> [1] https://github.com/w3c/PNG-spec/milestone/2
> <https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fw3c%2FPNG-spec%2Fmilestone%2F2&data=05%7C01%7Clrosenth%40adobe.com%7Cd8af81b3e8ec410a958308da374705de%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C637883075557864151%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=cAtWIChEUW9X4pZl0IwbXjsw7YLqFNcVsVtFNH4pddk%3D&reserved=0>
>
> Best,
>
> -- Pierre
>
> On Fri, May 13, 2022 at 8:45 AM Chris Blume (ProgramMax)
> <programmax@gmail.com> wrote:
> >
> > Hello everyone,
> >
> > It seems there are no new issues that need group discussion and thought.
> > Unless someone has an idea for discussion, let's cancel the May 16th,
> 2022 meeting.
>
>