[MPEG-OTSPEC] Introducing breaking changes into the spec (was: RE: [EXTERNAL] Proposal to deprecate derived search values)

[Peter Constable]

Let me clarify what I meant:

Wrt parallel activities regarding (1) docs on already-implemented tech, (2) incremental improvements to tech, or (3) a “step change” improved format, I wasn’t at all saying that there should not be parallel activity. By “one step at a time”, all I was saying was that, before technical discussion of #3 (step-change improved format) happens in this list, I think discussion it should be preceded by discussion of business goals, etc. for such work.

Formally, SC29 doesn’t have any approved project for working on a new format, and so technical discussion of any new format is not formally in scope for the AHG. In terms of formal SC29 process, technical activity on new formats would need to be preceded by approval of a New Work Item Proposal (NWIP) by SC29. A NWIP is primarily about business goals and project plan. Now discussion on preparing a NWIP could certainly potentially happen here (I don’t think the scope given to the AHG by SC29/WGx precludes that). But IMO I think it would be better for that to happen elsewhere.

I’ll also clarify my comments as they pertain to the FTCG, which are based on the FTCG draft charter<https://github.com/font-text/font-text.github.io/pull/1/files?short_path=dd2f8af#diff-dd2f8af7de66f4d3122bd6571fb0aa09>:


“The Font and Text community group gathers individuals and organizations interested in discussing and developing specifications and implementations for technologies such as shaping and layout which operate on and at the interface between text encoding and font formats.

“Under the present charter, the Community Group will determine a structure and process under which such specifications may be created by which the greatest number of interested parties may be able to participate. Until such a structure and process is determined, no technical development of specifications or implementations will be undertaken by this Community Group and no technical contributions will be accepted.”
My understanding of this has been there was interest in discussing various potential areas of activity, and where/how each of those areas of activity _might be_ pursued. The draft charter suggests to me there’s not much more that would be in scope beyond discussion of business justification, cost-benefit analysis, high-level goals, potential roadmaps, or other high-level exploration for possible areas of activity.

I certainly did not mean that any technical design discussion should happen within FTCG since that is explicitly out of scope in the draft charter. And while the FTCG might discuss where and how technical activities might be pursued, it certainly wouldn’t be able to _determine_ that technical activities will be pursued in any particular context, because it has no power of decision over other contexts.

I also did not intend to suggest that FTCG would _necessarily_ lead to some new activity related to font formats, or that any technical activity related to font formats should end up happening somewhere other than the AHG, or that conclusions regarding business goals in regard font formats should happen anywhere other than the AHG. I wasn’t even necessarily assuming that FTCG would discuss font formats at all: nothing in the draft charter requires (or precludes) that. I was only suggesting that FTCG would be a better place to start preliminary discussion related to step-change improved formats that might eventually lead to a NWIP.



Peter

From: mpeg-otspec <mpeg-otspec-bounces at lists.aau.at> On Behalf Of Dave Crossland
Sent: Monday, September 14, 2020 11:32 AM
To: mpeg-otspec <mpeg-otspec at lists.aau.at>
Subject: Re: [MPEG-OTSPEC] Introducing breaking changes into the spec (was: RE: [EXTERNAL] Proposal to deprecate derived search values)


Hmm. I've heard from some that TFCG *isn't* expected to charter font format spec work, but rather only the shaping spec work, which is adjacent and interrelated. I've heard from others that the TFCG *is* expected to spin off several CGs/WGs, at least one on font format specs.

That work is already happening on 3 levels: (1) authoring docs on the already implemented tech, (2) authoring incremental improvements to that tech, and (3) authoring a step-change improved format.

Peter, suggesting a serial approach ("one step at a time"), seems to me counter to what I see as generally agreed and already widely enacted, which is a parallel approach: These 3 tracks have been happening, in parallel, for years and will continue to happen _somewhere_.

It's only question of where.

Rod and Peter, while I didn't see your suggestions to use the FTCG mailing list or other fora instead of the AHG repo as attempts to start a font war, lol, I don't think _any_ FTCG venue is yet chartered for _any_ technical discussion, and as I said above, its not clear that FTCG fora will have font formats in scope.

Peter and Vlad, you have special positions as the editors of the MSOT and OFF specs. You've both posted responses in the last few days that I myself did interpret as trying to shut down discussion and progress that is newly happening in the AHG space around the 3rd track. So I'm glad to hear from Vlad that this was a misunderstanding, and this 3rd track is welcome within the AHG. However, what seems to still be missing is that collaborative authoring environment; the https://github.com/MPEGGroup/OpenFontFormat/<https://eur05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FMPEGGroup%2FOpenFontFormat%2F&data=02%7C01%7C%7C405bf39a40744eb7a41f08d858dc8cd8%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637357051558768104&sdata=DQZnsyIu9KVjLyhyCg%2B2n6Yf%2B01G1ylh%2FZLMomoQnco%3D&reserved=0> repo is "issue only" and no collaborative authoring of files is expected there, at present.

But perhaps if we follow Rod's suggestion to start posting all desirable change ideas as issues on the repo, and tag (or put into milestones) each issue to clearly mark which track it is on, then another tag can be added to indicate the idea is ready to move to authoring, and Pull Requests with drafts can be linked to the issues and commented on line by line before merged into meaningful "trunk" branches. (Perhaps Peter this is more what you mean by "one step at a time"? :)

What happened so far this year, to my eyes, is that ISO/MPEG had a window of opportunity to lead in providing a space for the font format spec discussions _and_ collaborative authoring, and it seems that window is now closing (but not fully closed) and they have ceded that opportunity to W3C, as there is no collaborative authoring space allowed here.

If it does turn out to be the case, that collaborative authoring isn't possible in the AHG repo, and the window closes, then what I expect will happen is that the "real work" on all 3 tracks will happen at W3C, and what will happen in the MPEG space is that "finished" change proposals will be posted here, and this will become (remain?) a 'mere' forum to voice objections that weren't heard upstream, and with the formal ISO WG as a final backstop forum for objections.

...This would be surprising to me, because I had heard earlier this year that this was the best forum available today for font format work, which I'd understood to mean all 3 tracks. But if that's how it is, that's how it is :)  And I don't see how this should be controversial, since that is pretty much what has been happening for many years, where the "real work" happens elsewhere (like private Unicode lists) and then goes into the MSOT spec, and finally then change proposals are posted here.

Really, my interest is to clarify and clearly document the whole thing; I don't mind too much where the "real work" happens, as long as it is effective, and everyone knows where to go.

It seems worth being explicit here that I as Google have already commissioned work on (3) from Black Foundry and Just van Rossum as part of the RoboCJK project, which is happening on that GitHub repo. Just like the COLR work, I don't think this location is ideal, not at all. Similarly, Simon is working on all 3 tracks at an impressive pace of his own initiative, not funded by Google, on commontype.org<https://eur05.safelinks.protection.outlook.com/?url=http%3A%2F%2Fcommontype.org%2F&data=02%7C01%7C%7C405bf39a40744eb7a41f08d858dc8cd8%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637357051558768104&sdata=A3B1maPo2%2Be6Utq8%2FcaCEHs7YOBBVumYq4Po8s9I4Yk%3D&reserved=0> and its associated Github repos. I can imagine that Simon's project may move to an MPEG repo, or a FTCG repo - but that's up to him.

While there are separate adjacent areas of interest and different costs and benefits to different organizational homes, there is value in common procedures. A set of repos scattered in org-less or single-vendor spaces, a set of repos within GitHub.com/font-text, and issues-only repos for Microsoft OpenType & Related Specs ("MSOT"?) and ISO/MPEG OFF ("MOFF"?) is the current state of affairs.

I'd like to see them converge a bit more, and de-duplicate as much as possible.

But in no scenario do I see any font wars emerging. The font wars were of a time when font formats were encumbered by proprietary licensing regimes, and I've heard nothing in 2020 that indicates any one expects any contributions to any upcoming formats to not be intended for extremely wide implementation. I think the business case for that is clear to everyone here... If any vendor chooses to delay implementation of a format that they are completely free to implement, that's not a war :) And its also the current state of affairs for OFF: large parts of OFF are not widely implemented by major vendors today.

Cheers
Dave
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.aau.at/pipermail/mpeg-otspec/attachments/20200914/320f3136/attachment-0001.html>