bugFree UCS Outline Fonts - Bugs: bug #51561, Misunderstanding of UTN #41 in re...


bug #51561: Misunderstanding of UTN #41 in re canonical reordering of diacritics

Submitted by:  David Corbett <dscorbett>
Submitted on:  Mon 24 Jul 2017 03:16:15 PM UTC  
Category:  individual character(s) Severity:  3 - Normal
Item Group:  letter positioning poor Status:  Fix posted
Privacy:  Public Assigned to:  Steve White <Stevan_White>
Open/Closed:  Open Release:  development

Add a New Comment(Rich Markup)

You are not logged in

Please log in, so followups can be emailed to you.


( Jump to the original submission)

Wed 26 Jul 2017 12:09:52 AM UTC, comment #10:

Hi David,

I'm not going to fix the issue of positioning the titlo with the typicon dots right away. Once I realized it was the right-dot and not the usual dot, and that there were other dots involved as well, I decided this wasn't a project for the short term. It's postponed, and put in a separate bug report.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Tue 25 Jul 2017 06:04:02 PM UTC, comment #9:

This is better but there is still a problem. In the following example, the dot-above-right is centered, but it should be offset to the right:

data:text/html;charset=utf-8,<div style="font: 300px FreeSerif">%D0%B6%D2%83%CD%98</div>

David Corbett <dscorbett>
Tue 25 Jul 2017 08:05:15 AM UTC, comment #8:

I've made a separate report for the Typicon project.
bug #51569: Cyrillic: Typicon symbols for Old Church Slavonic

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Mon 24 Jul 2017 10:53:02 PM UTC, comment #7:

I have heard that blanket statement before. As stated, it's obviously false. I would agree with it if it were qualified by "sometimes".
It seems to me that the rule of thumb ought to be to encode conceptually identical things.

The entry for that letter still says: Latin transliterations of the Southern Min dialects of Chinese.
That sign is placed over certain Latin letters in a peculiar way, and it means a certain thing.

Do the dots really look the same? Well the left-dot in the standard looks more like a rectangle than a circle.
I really doubt the wisdom of this kind of re-use. Ah well, not for me to decide.

What is for me to decide is whether I can implement this tipicon notation.

It will require a new set of anchors, on all (? I don't see a specification in the proposal) letters used in Church Slavoic, one for the left and the other for the right. That's going to be messy, besides a lot of work.

And again, it's for this one notational system (one of several, it seems), for a purpose (printing of religious texts) that is really beyond the purview of this project.

So I'm going to drop it.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Mon 24 Jul 2017 10:22:03 PM UTC, comment #6:

Unicode unifies common-script diacritics that look the same, regardless of usage. For example, U+0308 is a diaeresis, an umlaut, a double derivative mark, and a Syriac seyame. There is therefore no problem with using U+0358 in both Taiwanese romanization and typicon notation. It is simply a generic dot diacritic without any prescribed meaning.

David Corbett <dscorbett>
Mon 24 Jul 2017 09:08:47 PM UTC, comment #5:


OK Now I see. My example used completely the wrong character.
I can confirm that the re-ordering does occur, as described, with U+0358.

But then... I looked more carefully at this U+0358 -- an above-right dot used to transcribe certain East Asian languages. What is that doing in this discussion? Well I looked through the documents, and I still don't get it. This stuff is described in
Proposal to Encode some Additional Symbols used in
Church Slavonic Text (Revision 2)
They propose another code point be used,

1DF8) ◌ Combining Dot Above Left

(Whose inclusion into Unicode had snuck past me.)
And something about a reference from that to U+0358

But... I'm very reluctant to use it for this purpose. That character is explicity for something else, and its use here looks different from that purpose, and... I don't understand this connection with 1DF8.

AND it's just an annotation scheme for litugical texts invented in the 20th century. That's getting damned specialized.

Right now, I'm not liking this whole thing, for several reasons, first being what appearst to be a dubius re-use of a character.

If you're willing to shed light on it, please do.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Mon 24 Jul 2017 08:35:19 PM UTC, comment #4:

The reordering is indeed done in the shaping layer. The only thing the font needs to do is to provide mark-to-mark attachment lookups for U+0358 and U+0483 in both orders.

David Corbett <dscorbett>
Mon 24 Jul 2017 06:37:25 PM UTC, comment #3:


Well I balled it all up. Long story. See latest commit.

However, I stand by my statement that the marks are not reordere as described in the TN, in the remark above the figure about "normalization". My understanding is, this is something done by the font shaping layer, not by the font.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Mon 24 Jul 2017 04:02:48 PM UTC, comment #2:

No, the faces do not all have anchors to support both orderings of U+0358 and U+0483.

In FreeSerif, for example, dotabvright.cmb has one anchor (far-right-dot) and titlo.cmb has two (abovemk and cyr-above). To support both orderings they need to have an anchor in common.

David Corbett <dscorbett>
Mon 24 Jul 2017 03:37:16 PM UTC, comment #1:

Hi David,

I agree there is some misunderstanding.

The all faces have anchors to support both orderings of these characters.

The problem, as near as I can tell, is that, the font shaper doesn't perform the re-ordering -- on my system anyway.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Mon 24 Jul 2017 03:16:15 PM UTC, original submission:

The attachment Church-Slavonic.html of bug #51539 has a section demonstrating canonical reordering with and without CGJ. However, it does not demonstrate the issue mentioned in Unicode Technical Note #41. The note refers to U+0358 COMBINING DOT ABOVE RIGHT and U+0483 COMBINING CYRILLIC TITLO, which have different combining character classes and therefore are reordered; the attachment uses U+0307 COMBINING DOT ABOVE, which has the same class as U+0483 and is therefore not reordered.

For full conformance to UTN #41, the fonts should support combinations of U+0358 and U+0483 in both orders.

Currently, the sequence U+0307 U+0483 shifts the titlo to the left. It shouldn’t. The titlo should be centered.

David Corbett <dscorbett>


(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach Files:

No files currently attached


Depends on the following items: None found

Items that depend on this one: None found


Carbon-Copy List
  • -email is unavailable- added by Stevan_White (Posted a comment)
  • -email is unavailable- added by dscorbett (Submitted the item)

    Do you think this task is very important?
    If so, you can add your encouragement to it.
    This task has 0 encouragements so far.

    Only logged-in users can vote.




    Follow 3 latest changes.

    Date Changed by Updated Field Previous Value => Replaced by
    2017-07-25 Stevan_White StatusNeed info => Fix posted
    2017-07-24 Stevan_White StatusNone => Need info
        Assigned toNone => Stevan_White

    Back to the top

    Powered by Savane 3.4