Document good and bad case studies in spec change proposals

[davelab6]

In the past, I must assume, a change has been proposed to the MOFF Spec that went very well - good idea, clear business case, submission text flew by without much discussion other than approval - all round exemplary.

And I must assume there is also "the worst case so far", and while it may (or, may not) be uncomfortable for some, in order to learn from the past, and not repeat past mistakes, it will be equally valuable to document this case study too.

I'd therefore like to suggest that the readme have links to some "trail heads" on the mpeg-otspec list archives where anyone interested can see these cases first hand.

Ultimately I'd like to work on a proper "missing manual" for MOFF AHG members, but these 2 case studies (and any interesting ones in between) will be the raw material for such a document, and everything is in the mail archive already, we just need the list of Subject lines.

[behdad]

A really bad case is the MERG table because it looks as if it works, but doesn't at all when you account for full complex positioning. It was done hastily, by MS only, to address what's a graphics issue in the font format, basically transferring responsibility to all font designers because the MS engineers were too lazy to find an engineering solution instead. A very common pattern at MS and one of the reasons they shouldn't be allowed to just shove things into MOFF the way they have been doing forever.

[davelab6]

OK, how was the MERG table proposed for inclusion in MOFF, and how was it accepted?

[behdad]

OK, how was the MERG table proposed for inclusion in MOFF, and how was it accepted?

It wasn't. MS just released it in MOT one day and it appeared in MOFF.

[davelab6]

MS just released it in MOT one day

Okay, I expect we can figure out when the first public date for this was.

https://docs.microsoft.com/en-us/typography/opentype/spec/merg is dated Aug 7th 2018.

and it appeared in MOFF.

It seems this is true: https://www.google.com/search?q=merg+https%3A%2F%2Flists.aau.at%2Fpipermail%2Fmpeg-otspec%2F turns up just 1 result, https://lists.aau.at/pipermail/mpeg-otspec/2018-September/000049.html - and the attached file isn't available for download.

(@vlevantovsky is that correct, that all files attached to emails sent to the list are not archived?)

So, it seems there was no proposal - which is odd, since there was a 'proposal' from MS for COLR (https://lists.aau.at/pipermail/mpeg-otspec/2013-July/000303.html) - although I say proposal in quotes since that was made after the format was announced (at the MS Build event.)

...Is what you say you want to do in https://lists.aau.at/pipermail/mpeg-otspec/2020-September/002388.html a 'best case'? ;p

[vlevantovsky]

The proposal was made and presented to this AHG for review and discussion back in 2016, as part of the group of proposals from MS, Apple and Adobe. See https://lists.aau.at/pipermail/mpeg-otspec/2016-October/000827.html The proposal from MS included adding the description of 'MERG' table, among other things.

[vlevantovsky]

As far as the attachments are concerned, Yahoo Groups that we used in the past offered file storage facilities that included both documents uploaded by group members and the attachments to emails (attached documents would be stored there with the download link being part of the email). When Yahoo decided to discontinue support for all user-generated content, they removed group files and kept only email archives. However, the proposals mentioned in my email from October 12, 2016 were official input contributions, and the originals are available from ISO (MPEG) document repository. I also have copies of them, so can easily share with all group members (as I've done in the past). There have been no objections raised to any part of those proposals, and the content has become the part of the then-current working draft of the future OFF 4th edition (published in 2019). There have been multiple opportunities to review that draft (starting with the very first version shared on the AHG list in December 2016 (https://lists.aau.at/pipermail/mpeg-otspec/2016-December/000836.html), and at all other stages of the document progression to its final published version.

[HinTak]

FWIW, I added generic acceptance of MERG as an opentype table to FontVal in Nov 2018 (just to stop FontVal warning it is unknown). So it "landed" quite a bit before then. There were multiple /regular drafts from @vlevantovsky . I try to keep up, but my willingness to spend time on FontVal is limited by resources. I suspect others feel the same way about reading draft documents ahead of GA. It is easy to criticize something as a dubious idea (I am not for or against MERG - just speaking in general terms) with a few years of hindsight.

[vlevantovsky]

I'd like to add that there have been significant efforts put in place by AHG members to review and contribute to the first working draft, see https://lists.aau.at/pipermail/mpeg-otspec/2017-January/subject.html There were updated draft (see https://lists.aau.at/pipermail/mpeg-otspec/2017-February/000881.html and the discussion that followed: https://lists.aau.at/pipermail/mpeg-otspec/2017-March/subject.html, https://lists.aau.at/pipermail/mpeg-otspec/2017-April/subject.html), before the decision was made that it was mature enough to be promoted to the Committee Draft stage (https://lists.aau.at/pipermail/mpeg-otspec/2017-April/000911.html), and many times ballot comments were discussed (see https://lists.aau.at/pipermail/mpeg-otspec/2017-July/000922.html and https://lists.aau.at/pipermail/mpeg-otspec/2017-August/000924.html)

[davelab6]

Thanks for the leads, vlad! However, since the attachments were deleted, and no longer accessible to newer members of the AHG like myself, would it be possible to make them available somewhere again? It's hard to piece together the timeline.

I also think it's relevant when the MSOT spec was updated, and moreover when the DWrite ship sailed.

[vlevantovsky]

It's hard to piece together the timeline.

Not hard at all! Every email in the archive linked in my previous comments has a date/time stamp! The documents that used to be attached to the emails on the AHG list are available form MPEG document registry and from ISO Livelink. The emails I referenced in my prior comments have document numbers, these are unique identifiers that allow any active SC29 member find and access them in the document registry.

[davelab6]

Okay, I'll join sc29.

Are MPEG document registry and ISO Livelink publicly available archives?

Obviously date stamps are there, but the contents in attachments isn't available as expected, and the fact I have to go anywhere else is what makes it hard..

[vlevantovsky]

Access to document archives is password protected, and available to active members only.

[PeterConstable]

Okay, I'll join sc29.

Members of SC29 are not individuals nor companies. They are national standards bodies. What individuals can do is to join a national standards body and participate in the national committee that corresponds to SC29 (if there is one in that country's standards body).

[vlevantovsky]

What individuals can do is to join a national standards body and participate in the national committee that corresponds to SC29

A minor clarification - it depends on national committee rules. In US, membership in INCITS is open to corporate entities only, an individual cannot join unless he/she incorporates as LLC or something similar. For corporate entities who are already members of INCITS L3 committee (a mirror of SC29 in US) - a primary member can nominate any number of associate members (i.e., company employees) to participate as experts.

[davelab6]

Thanks for the corrections!

What individuals can do is to join a national standards body and participate in the national committee

https://twitter.com/KhaledGhetas/status/1308816081222590470

Sadly it seems this is far from universal.

[behdad]

I do NOT feel safe to express my opinions in this forum.

[murata2makoto]

I studied the SC29 repository. Sadly, WG11 documents do not exist there.

[davelab6]

@PeterConstable please could you take a look at the private repo commit history for the markdown file behind https://docs.microsoft.com/en-us/typography/opentype/spec/merg and clarify the timeline for that piece?

I also wonder if you can determine when DWrite shipped support for it.

[davelab6]

@vlevantovsky please could you provide more details about how to retrieve the documents you referenced above? :)

1. 2017-02-21 w16626.zip
2. 2017-07-12 DRAFT-BallotComments-14496-22-CD.doc
3. 2017-08-03 DRAFT-BallotComments-14496-22-CD-final.doc

[PeterConstable]

The MicrosoftDocs/Typography repo was created in 2018. Prior to that, the OT spec was maintained on a different server as HTML files. You can see from the OT spec change log that it was added to the OT spec in version 1.8. I don’t know when it was first supported in DWrite, but it would have been prior to when I started working with the Window graphics team in spring 2014.

Peter

From: Dave Crossland notifications@github.com Sent: Wednesday, September 30, 2020 9:11 AM To: MPEGGroup/OpenFontFormat OpenFontFormat@noreply.github.com Cc: Peter Constable peter.constable@outlook.com; Mention mention@noreply.github.com Subject: Re: [MPEGGroup/OpenFontFormat] Document good and bad case studies in spec change proposals (#12)

@PeterConstablehttps://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FPeterConstable&data=02%7C01%7C%7C794f28a39c2b4aedb1ca08d8655b76e1%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637370790790883159&sdata=Qm00FiJ%2BwW7ZEi3as1xTGigjuJpsyN9OC%2FZ6d4w4gZw%3D&reserved=0 please could you take a look at the private repo commit history for the markdown file behind https://docs.microsoft.com/en-us/typography/opentype/spec/merghttps://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdocs.microsoft.com%2Fen-us%2Ftypography%2Fopentype%2Fspec%2Fmerg&data=02%7C01%7C%7C794f28a39c2b4aedb1ca08d8655b76e1%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637370790790883159&sdata=mrFe5t0viQOEOM0iKF3MZwUGO0F5yKXEfh3EbtLbmf4%3D&reserved=0 and clarify the timeline for that piece?

I also wonder if you can determine when DWrite shipped support for it.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FMPEGGroup%2FOpenFontFormat%2Fissues%2F12%23issuecomment-701491153&data=02%7C01%7C%7C794f28a39c2b4aedb1ca08d8655b76e1%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637370790790893152&sdata=MTlxgZPuB%2FMtvyA%2FvHTUEZhebpEjV%2B1elAj7kY2Mgao%3D&reserved=0, or unsubscribehttps://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAGLBYVO5YLPHJPE4UVJSE5TSINKCLANCNFSM4RRL3ULQ&data=02%7C01%7C%7C794f28a39c2b4aedb1ca08d8655b76e1%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637370790790893152&sdata=PT6LSyFw002k4IaVj%2FnUxRF%2FHWG%2BH9UuuILnWF0ElJg%3D&reserved=0.

[vlevantovsky]

please could you provide more details about how to retrieve the documents you referenced above? :)

2017-02-21 w16626.zip 2017-07-12 DRAFT-BallotComments-14496-22-CD.doc 2017-08-03 DRAFT-BallotComments-14496-22-CD-final.doc

I believe draft ballot comments shared on the AHG list would no longer be available (it was a draft after all) but the "official" submitted copies, as well as all output documents produced by MPEG since 1995 are still available from MPEG Document Management System (password protected).

[HinTak]

@vlevantovsky please could you provide more details about how to retrieve the documents you referenced above? :)

1. 2017-02-21 w16626.zip
2. 2017-07-12 DRAFT-BallotComments-14496-22-CD.doc
3. 2017-08-03 DRAFT-BallotComments-14496-22-CD-final.doc

I have the first 2 of the 3 in my hard drive. I had a habit of downloading and hoping to find time to read them, but rarely got round to...

[davelab6]

The overlap bit issue is a great case study, touching on TrueType, OpenType and WOFF2 format specs. Thread starts here

https://twitter.com/ArrowType/status/1311716316320018438

[murata2makoto]

I studied the SC29 repository. Sadly, WG11 documents do not exist there.

Correction. WG11 documents are in the directory "Archive folder (ISO/IEC JTC 1/SC 29/WG 11)" which is in turn in the directory "02. General committee documents".

[vlevantovsky]

Correction. WG11 documents are in the directory "Archive folder (ISO/IEC JTC 1/SC 29/WG 11)" which is in turn in the directory "02. General committee documents".

@murata2makoto Your investigative activities in this repo and communications on the AHG mail list are not covered by any of the AHG mandates, and your attempts to cast a shadow of doubt on the legitimacy of our work appear to be self-serving and disrespectful. This behavior is highly unprofessional, especially considering that you hold a position as SC34 officer!