[MPEG-OTSPEC] Draft AHG report

[Hin-Tak Leung]

 Hi Vlad. Thanks for the update.
Sorry perhaps I am getting a bit confused about the actual present paragraphs in the report document, versa people background explaining / claiming the rationale and benefits of said proposed changes in the zoom calls (which are mostly absent from the doc). The report itself without the zoom context is largely fine as is now.
I think I am generally skeptical of people proposing new features based on either of these direction of claims:
- new is better than old/alt executed poorly. Old/alt executed poorly is just poor, and no indication of new being better. Size-claims are mostly in this category, when you look at how poorly executed the old/alternative is, in the comparison.
- new is magical without any complications from discarding years of learned wisdom tuning and fine-tuning old. That would be claims about hinting being "just seem to work" ... 


    On Sunday, 15 October 2023 at 20:32:54 BST, Vladimir Levantovsky <vladimir.levantovsky at gmail.com> wrote:  
 
 Thank you Hin-Tak for your review and quick response, really appreciate it.Regarding proposed changes:- "braking" -> "breaking" type fixed;- portions of the report related to composite glyph size benefits are modified by introducing specific style references and editing the language of those claims.(See attached edited draft report.)
I am not sure I understand your comment about "The other size claim ..." - what portion of the AHG report you are referring to? The August AHG meeting summary (see topic 4: Combining cubic and quadratic Bézier curves in glyphs) describes a part of the discussion where we talked about mixing quadratic and cubic outlines, and it does include a statement that "mixed versions showed better file sizes in the initial experiments". This statement is based on what was presented during the meeting based on the information available at the time, and the text of the report presented for AHG review does not mention anything regarding the possibility of sharing data between old / new glyph tables. 
Thanks again,Vlad


On Sun, Oct 15, 2023 at 6:26 PM Hin-Tak Leung <htl10 at users.sourceforge.net> wrote:

 Hi Vlad. Thanks for the update.
Firstly, a small correction - "braking changes" surely that is "breaking changes"?
I am reading claims of size-benefits of proposed changes quite carefully, and quite skeptical of them in general.
The claims about composite usage in hangul and CJK is really limited to optimal/stylistic usage scenarios with Hei/San styles, really, and perhaps put quite severe constraints on both the style (not plural) in heavily composite-oriented design based approach, and also the actual day-to-day design work-flow. Worst case: you see styles coming out looking like those from B-movies with Fu Manchu as the villain from the 60's. That reminds me of the work of a famous artist called Xu Bing - one of the things he is famous for, is writing English phrases/slogans in chinese-calligraphic-style composite glyphs. What it sounds like is the reverse - losing all the calligraphic tradition of the far east and recommending/forcing hangul/CJK into a San style.
The other size claim is about the possibility of sharing data between old glyf and new GLY2(?). otf tables have alignment and  checksum requirements, so that would mean contiguous-ness of the old subset, and possibly padding (and also a 1 to 3(?) byte padding break between old subset and new) complications. Some of the details are certainly behind closed-source implementations, but I believe there are non-overlapping-ness checks in various font checking schemes / tools, both stand-alone and built into OSes too: ie valid tables are taken as non-overlapping, and overlapping tables indicate broken fonts. That will need to change/specified/excepted if glyf and GLYX/GLY2(?) is to share content.
I can continue my rant on size claims regarding SVG vs COLRv1, and quadratic vs cubic too. Poorly written SVG is poor, and qudratics not using implicit on-curves points are larger. Claims about being better than poor SVG and better than poor quadratics, size-wise, are meaningless.
So I'd prefer to see those size-related claims removed, if not heavily qualified/caveated.
    On Sunday, 15 October 2023 at 11:06:57 BST, Vladimir Levantovsky <vladimir.levantovsky at gmail.com> wrote:  
 
 Dear all,

Please see attached the text of the AHG report I plan to present tomorrow during the opening SC29/WG meeting Plenary. This report is based on the AHG meeting summaries that were previously distributed on this list.If you have any comments, please send them by the end of the day today.
Thank you,Vladimir
_______________________________________________
mpeg-otspec mailing list
mpeg-otspec at lists.aau.at
https://lists.aau.at/mailman/listinfo/mpeg-otspec
  
  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.aau.at/pipermail/mpeg-otspec/attachments/20231016/9967e99d/attachment-0001.html>