bugFree UCS Outline Fonts - Bugs: bug #34687, Bad shape of Cyrillic letter Uk...

 
 

bug #34687: Bad shape of Cyrillic letter Uk (U+0478 & u+0479)

Submitted by:  Alessandro Ceschini <aleces>
Submitted on:  Sat 29 Oct 2011 07:58:41 PM UTC  
 
Category: individual character(s)Severity: 3 - Normal
Item Group: glyph(s) incorrectStatus: Won't fix
Privacy: PublicAssigned to: Steve White <Stevan_White>
Open/Closed: ClosedRelease: 2010-09-19

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 Jump to the original submission)

Fri 25 Nov 2011 08:56:16 AM UTC, comment #25:

Wait a minute, those samples are in so-called Church typeface, I can give you loads of monograph or digraph UK in Church typeface!

What I didn't manage to find was an UK in a civilian typeface.

The same is true for KSI, by the way.

Alessandro Ceschini <aleces>
Thu 24 Nov 2011 11:18:09 PM UTC, comment #24:

Alessandro,

You pointed out the main Cyrillic range Uk's are deprecated.
The ones in Cyrillic Extended-B have been implemented for a long time in Serif: U+A64A-B.

Uk appears several times in
1229 Treaty between Smolensk, Riga and Gotland
The shape is like a Greek gamma with very long curved arms.

You can search here:
Birchbark Documents from Novgorod
(Gramoty -- Titus Document: 109)
http://titus.uni-frankfurt.de/texte/etcs/slav/aruss/gramoty/gramo.htm

For a picture of thw whole document see
http://www.schaeken.nl/lu/research/online/editions/1229/

I found some of the uk's. See attached picture.

(file #24444)

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Tue 22 Nov 2011 09:01:23 PM UTC, comment #23:

Hi Steve

You'right about FreeSans, it doesn't contain UK, I must have clicked twice on FreeMono.

I know about monospace fonts, but I'm going to have that lot of trouble in order to convince you this IS INDEED a bad solution.

Just narrow down the two components, it's just the operation you did in FreeSerif but magnified and applied to CYRILLIC U as well.

Anyhow, this is just a mental masturbation, UK in Cyrillic is rarer than LONG S is in Latin, no one in this world is ever going to use it in a FreeMono font. I wonder why you included it in the first place. Usually this fonts are used in console, this is truly a no-go area for archaic characters.

On the other hand, implementing archaic characters are usually implemented in Sans fonts, like Arials etc... Maybe you should think about it, the OLD CHURCH SLAVONIC ONLINE http://www.utexas.edu/cola/centers/lrc/eieol/ocsol-0-X.html uses Mono fonts for his OCS lessons.

Bye bye

Alessandro Ceschini <aleces>
Tue 22 Nov 2011 08:04:12 PM UTC, comment #22:

Hi Alessandro,

The letter Uk has not been implemented in FreeSans. You are looking at a glyph from a different font.

As to FreeMono, I'm going to answer seriously, although I thought at first you were joking.

It could be you don't appreciate that monospace fonts are under different technical restrictions than the variable-spaced ones.

The letters of a monospace font must all have the same widths, or for many applications, it will fail completely. This causes an implementation problem for very wide letters. It is possible to let the contours of the glyphs stray out of the bounding box, but not a good idea -- it causes other problems.

The monospace font is of course very stylized in all alphabets, with the view toward certain mechanical typewriter faces. Of course no native speaker of Old Church Slavonic ever saw a typewriter.

The monospace Uk in FreeMono is my own invention, a solution of the problem of how to represent the letter in a way that fits with the restrictions on the font, but which is visually unambiguous and not ugly. Frankly, you're going to have a lot of trouble convincing me it was a bad solution.

Cheers!

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Tue 22 Nov 2011 04:19:49 PM UTC, comment #21:

Hi Steve

After many pieces of criticism, I must give you a thumb up for the way you handled that glyph IN FreeSerif, I like it very much, I couldn't even have spotted the O is narrower if I hadn't been forewarned. I really hope Malo is going to like that, for me its OK!

BUT, there's always a BUT :( in FreeMono and FreeSans the O and U look superimposed, is it an intended effect? I've never seen it before, are you sure it's allowed in Cyrillic typography?

Greetings

Alessandro Ceschini <aleces>
Fri 18 Nov 2011 09:59:01 AM UTC, comment #20:

Submitted a version with
-- у-shaped part the same as Cyrillic у
-- о-shaped part a narrow version of Cyrillic о

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Fri 11 Nov 2011 06:33:28 PM UTC, comment #19:

Hi Steve

As far as I'm concerned, I don't need a compromise, I don't use that glyph WITH FREE SERIF. I use other, specific fonts to typeset Church Slavonic, which have a standard o.

I was only arguing on principles, if you're OK with a narrower o, all that discussion crumbles.

That's it. I'm sorry if I've been a stalker, but I thought the bigger picture deserved more importance than a single user's wishes.

You tell me, this is your font.

Alessandro Ceschini <aleces>
Fri 11 Nov 2011 06:22:08 PM UTC, comment #18:

OK,

How about a compromise?

In the existing glyphs, the 'о' part is slightly narrower than the Cyrillic 'о'. It's enough to somewhat offset the width of the glyph visually, but not enough to be easily noticed.

Could you both agree to that?

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Fri 11 Nov 2011 05:24:28 PM UTC, comment #17:

Hi Steve

Of course you don't follow slavishly the shapes in Unicode samples, I was just pointing out to Malo that although an example of usage of a narrower o (but also narrower y) does exists, (Unicode was adduced as an example of this), the style of Free Serif is indeed, not slavishly modeled on what Unicode proposes.

Overall, it's just an idiotic syllogism I'd better have avoided. I'm REALLY sorry for that.

Now, back to the core issue, as far as Malo is concerned, he wants the o to be narrower because the recension of Church Slavonic he follows does use a narrower o, but I can rebut this by saying that what concerns me (a much older recension, called OLD Church Slavonic), definitely DOES NOT, and I really find it inappropriate.

Malo also says that, as far as I can understand, what I call UK CYRILLIC U+CYRILLIC O is ONIK for him, and the version with a narrower o is the IK. So what for me (and Unicode) are just ONE letters, for him is TWO. Also, he calls UK the digraph version, which has a different Unicode point and doesn't concern us here.

You understand that the nomenclature he utilizes differs from both what Unicode establishes and my references. As I said, the STAROSLAVJANSKI SLOVAR' (an established reference in the Church Slavonic world), displays an UK with a normal o, and no narrow o.

All in all, and I'm ready to apologize if my impression turns out to be false, Malo has very customized requirements, and somehow doesn't grasp that if the font is tailored according to his wishes, it will eventually impose his particular needs to other users that don't have those particular needs.

If he doesn't agree with the Unicode nomenclature (not glyphs, I'm talking about code points), I think he should resort to private areas encoding in a custom made font.

Regards

Alessandro Ceschini <aleces>
Fri 11 Nov 2011 04:20:33 PM UTC, comment #16:

Alessandro,

We do not slavishly follow the shapes in the Unicode samples.
Did you somehow think we did? If so, where did you get that impression? If not, what's your point?

Malo,

If we understand you correctly to mean the 'о' part of UK must be narrower than the Cyrillic 'о', please provide some outside evidence, where it is explicitly stated that this is important to the shape of the letter, rather than a stylistic choice.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Thu 10 Nov 2011 02:40:20 PM UTC, comment #15:

And what I said about Unicode is that although the shape they propose for UK entails a narrower CYRILLIC O (but also a narrower CYRILLIC U), the "flavor", I don't know how to call it, the style, is different than Free Serif. I brought forward the example of the different shape of CYRILLIC OMEGA U+460/U+461 in the Unicode specimen to make it clear that you don't have to NECESSARILY follow what Unicode proposes. Indeed you did not, as your CYRILLIC OMEGA does look very different than THEIR CYRILLIC OMEGA.

And that's only an example, there are scores of other minor differences between FreeSerif and the font used in the Unicode specimen.

The bottom line was that you have to respect the style of a font, you cannot just change a glyph because it suits your need, you have into account consistency: if we bow to Unicode's authority in the case of UK, why not in the case of OMEGA? And what about ZHE? Unicode's ZHE is slightly different, will you change that too? I don't think so!

I hope you understood the point I'm trying to make here.

Alessandro Ceschini <aleces>
Thu 10 Nov 2011 02:24:46 PM UTC, comment #14:

Hi Steve
Yes, as I told you we both agree that the second component of UK should be like the Latin y (that is, identical to CYRILLIC U).

I'm sorry if I took too much of an adversarial stance, but the issue of the shape of the first component of UK does remain.

I want it just like the CYRILLIC O.

Malo wants it narrower.

That's THE issue.

Alessandro Ceschini <aleces>
Thu 10 Nov 2011 02:10:21 PM UTC, comment #13:

The original question concerned the shape of the у-shaped part of UK. Currently in FreeSerif, it is styled quite differently from у. As near as I can tell, you both agreee it should be like the у. Am I wrong?

Now you're arguing about the shape of the o part. Near as I can tell, you both agree it should be the same shape as the Cyrillic o. Am I wrong???

I am willing to make the UK from the other two letters of Cyrillic.

As to your other argument: it strongly resembles "talking past one another". A case where in fact, on the points that are important to you both, you really agreee completely, but have taken adversarial stances, and therefore interpret the other person's points as being in disagreement with your own.

What are the issues here, and what is their order of importance?

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Thu 10 Nov 2011 11:22:01 AM UTC, comment #12:

Hi Malo,

OK, Unicode specimen stands by your wishes: unicode.org/charts/PDF/U0400.pdf shows CYRILLIC UK with a narrower CYRILLIC O (actually also the second component is narrower than CYRILLIC U) but then if we want to follow that specimen, we've got to change the shape of CYRILLIC OMEGA U+460/U+461, which as you can see is very different from the one adopted by UCS Fonts.

The bottom line is there's got to be a consistency in fonts, the "flavor" has got to be consistent throughout the whole glyphs inventory. It cannot be a matter of personal preferences, we cannot hijack a font because it doesn't suit our personal needs.

My proposal:
UK as just CYRILLIC O + CYRILLIC U as a standard glyph, narrower shape in salt/hist (or character variant?).
And maybe the same thing for OMEGA, since we're at it.

Regards

Alessandro Ceschini <aleces>
Thu 10 Nov 2011 08:15:55 AM UTC, comment #11:

>>> 2) Malo wants the first component to be a narrower CYRILLIC O, I'll agree to that only if Malo can demonstrate such a shape was in actual use in civil script, for the example of the Church script is of no use here, UCS Fonts use standard, civilian Cyrillic glyphs.


It's not correct approach, we don't make a limitation to a language range for a font. As early as I denoted, we used modern font to show old cyrillic texts. And shapes of the glyphs shell follow rules according to the specific glyphs, and its history. Thus, in the 'oy' glyph o must be narrower than others.

Malo Skrylevo <malo_skr>
Project Member
Wed 09 Nov 2011 01:28:11 PM UTC, comment #10:

Hi Steve
I'm really sorry for having hijacked this topic, of course this is no place for history discussions BUT:
1) We (I and Malo) do agree that the second component of U+0478/9 should be identical to CYRILLIC LETTER U, and that was the motive why I started this topic, that indeed it is different in FreeSerif, and that's bad.
2) Malo wants the first component to be a narrower CYRILLIC O, I'll agree to that only if Malo can demonstrate such a shape was in actual use in civil script, for the example of the Church script is of no use here, UCS Fonts use standard, civilian Cyrillic glyphs. Otherwise you'd better stick to what I call, a compatibility, or the most neutral form, that is "identical in form to a Latin o and y next to one another".
Regards

Alessandro Ceschini <aleces>
Wed 09 Nov 2011 01:07:53 PM UTC, comment #9:

Guys,

It sounds as if your disagreement is about history. (It seems you two might have a lot of interesting stuff to discuss!)

It sounds as if you agree that U+0478-9 should be identical in form to a Latin o and y next to one another.

Is this correct? If so I will make that the case.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Tue 08 Nov 2011 11:16:25 AM UTC, comment #8:

Hi Malo,
Listen I don't want to start a flame here, but I'll provide you with my reference: STAROSLAVJANSKIJ SLOVAR' pp. 56-57, you'll find that the letter both as a digraph and as a monograph is called ukŭ (to be precise we must not omit the hard jer or, jerŭ). There is no trace of what you call onik or ik. But these could be later innovations added in later centuries, I only deal with OLD Church Slavonic (till the XI century), not with the rest.
BUT there is no way you can convince me that what is modern Russian JA did originate from IOTIFIED A. As a matter of fact, Southern Slavic sounds JA (represented by IOTIFIED A) and nasal E (represented by LITTLE YUS) coalesced very early so that both letters were used interchangeably until Peter the Great decided to drop IOTIFIED A and adopt a cursive version of LITTLE YUS as the only letter for J+A. I've been taught this at the university and you can easily find that in books. You want the proof? In Russian what was Common Slavic pamętĭ became память in Russian BUT памет in Serbian. Wikipedia confirms my version: http://en.wikipedia.org/wiki/%D0%AF#History.
Last but not least, you cannot use pre-18th century texts written in Church script as a reference for a civilian font. And since the civil script (not an alphabet as you seem to imply, Church and civil scripts are two "styles" of the same Cyrillic alphabet, like Blackletter and Antiqua are of Latin.) aimed at standardizing letter shapes, the shape of U+0478/U+0479, however you wish to call it, should be the sum of CYRILLIC O and CYRILLIC U. With archaic spellings, I always meant post-18th century, civil script archaic spellings, I mean, like JAT, which is well attested in 19th century's Serbian and Russian texts.
I hope I've made my point clear.
Cordially

Alessandro Ceschini <aleces>
Tue 08 Nov 2011 06:17:08 AM UTC, comment #7:

Hello Alessandro!

Please note. We call the letters as follows: onik is correct name of U+478/9, "uk" is U+A64A/B, and "ik" is U+443/23.

>>> Listen Unicode has made lots of blunders as far as OCS is concerned,


Yes, I guess. At the last discussion we agreed (we are UCS/HIP text writers) to use some letters like "i:", "jad" etc. with var selector. But we still can't use it to replace "onik", as we still haven't "uzkij on" in the Unicode table.

>>> they wanted to force us U+044F CYRILLIC SMALL LETTER YA for the IOTIFIED A, when everybody (who was aware of the history of the Cyrillic alphabet), knew U+044F actually originated from a variant shape of U+0467 CYRILLIC SMALL LETTER LITTLE YUS.


U+044F was originated even not from "jus_malyj" (U+467), but from "Az" (U+430/10). Because in middle church slavonic texts (15-17 cent.) we have shape of capital "Az" (U+410) quite similar to modern "Ya"(U+42F) char. But in the functional mean your statment is correct.

>>> I also want to state it clearly I have no problem with a narrower CYRILLIC O, but the shape of U ought to be identical to U+0443.


I'm agree. That the "u", and the "ik" in the "onik" should be the same. I'll fix that.

>>>But, last but not least, if you do want to typeset Church Slavonic, this is not the font for you: you must use fonts which use the archaic, uncial shapes reserved for Church Slavonic.


Some of our text is required to be typed with the modern shapes (for example article captions). So we require FreeFont at least for this.

>>> So, I think it's safe for FreeSerif to stick with a "compatibility" shape, provided that no one is going to use it for the purpose of typesetting Church Slavonic. Neither does UK appear in archaic spellings of Serbian (as, say, YAT does) and as far as I know, nor of Russian.


I think we should look in to texts which were written before Peter's I alphabeth change. if it will be possible, I'll look into them.

Malo Skrylevo <malo_skr>
Project Member
Sat 05 Nov 2011 12:31:16 PM UTC, comment #6:

Hello Malo!

Listen Unicode has made lots of blunders as far as OCS is concerned, they wanted to force us U+044F CYRILLIC SMALL LETTER YA for the IOTIFIED A, when everybody (who was aware of the history of the Cyrillic alphabet), knew U+044F actually originated from a variant shape of U+0467 CYRILLIC SMALL LETTER LITTLE YUS.
Also, of course no letter U+0443 SMALL LETTER U existed among the archaic Cyrillic letters: UK was the only grapheme for U, so talking about a ligature or a digraph of CYRILLIC O and U, is, from a historical point of view, an aberration, the reverse is true: the second part of the grapheme slowly began to have a life of its own, ultimately severing its umbilical cord with O.
I also want to state it clearly I have no problem with a narrower CYRILLIC O, but the shape of U ought to be identical to U+0443.
But, last but not least, if you do want to typeset Church Slavonic, this is not the font for you: you must use fonts which use the archaic, uncial shapes reserved for Church Slavonic. So, I think it's safe for FreeSerif to stick with a "compatibility" shape, provided that no one is going to use it for the purpose of typesetting Church Slavonic. Neither does UK appear in archaic spellings of Serbian (as, say, YAT does) and as far as I know, nor of Russian.

Alessandro Ceschini <aleces>
Sat 05 Nov 2011 11:52:32 AM UTC, comment #5:

Hello, Alessandro

>>> It was originally a digraph of the Cyrillic letters O (О о) and U (У у)[dubious – discuss]


This wikipadia line has small error, as well as, the following unicode declaration (Unicode 6.0):

CYRILLIC CAPITAL LETTER UK
• may be rendered as either monograph or digraph form
• for “digraph uk” the preferred spelling is 041E О 0443 у
• for “monograph uk” the preferred character is A64A Ꙋ
→ 0423 У cyrillic capital letter u

The error is consist in that, the cyrillic digraph uk "oy" includes not simple "on" letter - 'o'(U+043E), but non-included in Unicode so called "uzkij on" (narrow o) char. So neither we, (in our texts) nor each other can replace digraph uk "oy" to pair "o + y".

And also, please, note that the Unicode doc has no "MUST" word, only preferred.

Malo Skrylevo <malo_skr>
Project Member
Wed 02 Nov 2011 02:07:15 PM UTC, comment #4:

I did some researches in the Internet and I actually found out that this character is deprecated: http://en.wikipedia.org/wiki/Uk_%28Cyrillic%29#Representation_on_computers

Quote: To resolve this ambiguity, Unicode 5.1 has deprecated the use of the original code points, introduced the new U+A64A and A64B, CYRILLIC CAPITAL/SMALL LETTER MONOGRAPH UK (Ꙋ ꙋ), and recommends composing the digraph with two individual characters ‹о›+‹у›

That means that I'm right :) You must drop that glyph, replace it with one that is just CYRILLIC O + CYRILLIC U, and if you happen to be interested in implementing archaic Cyrillic letters, implement U+A64A and U+A64B, or while you are at it, the whole range of Cyrillic Extended B.

Alessandro Ceschini <aleces>
Sat 29 Oct 2011 10:51:56 PM UTC, comment #3:

OK.

I'll review the information, and try to remember why this glyph was changed.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Sat 29 Oct 2011 09:21:55 PM UTC, comment #2:

No, that's only used in Church Slavonic, as you say, and very archaic spelling of other Cyrillic-based languages, but there are also reasons of consistency. If you change Cyrillic UK, you also have to change Cyrillic U to a more archaic-looking glyph.
And that would not be good: FreeSerif is a standard Serified font that uses the so-called "Civil" script, that is the modern shape of the Cyrillic Alphabet. There are plenty of fonts which support the older, Medieval style of Cyrillic, and these should be used for typesetting Church Slavonic.
I highly recommend you switch back to the previous glyph, as virtually no font with modern (i.e. standard) Cyrillic glyphs make use of it.
To make that understandable even to those who're not really into Cyrillic typography, you surely know the glyphs we use nowadays in Latin are a Renaissance revival of Carolingian minuscule (aka Antiqua), overriding the Blackletter face which was in use during the Low Middle Ages. Now, having a Medieval glyph for Cyrillic UK and a modern one for Cyrillic U is like having an Antiqua shape and a Blackletter one coexisting in the same font.
Not really consistent and beyond any doubt ugly.

Alessandro Ceschini <aleces>
Sat 29 Oct 2011 09:02:27 PM UTC, comment #1:

Hi Alessandro,

That would certainly have made it easier on me, and I know that's how I first implemented them, just following the Unicode samples.

But this font is used by a guy who is writing in Old Church Slavonic. I think he wanted these forms. Have a look in the SVN logs -- I know this happened in the past 3 years.

Is there a specific Serbian for which these forms are wrong?

Cheers!

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Sat 29 Oct 2011 07:58:41 PM UTC, original submission:

These letters, respectively capital and small, should look just like Cyrillic O plus Cyrillic U, or, if you prefer, Latin O plus Latin Y.

Alessandro Ceschini <aleces>

 

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

Attach File(s):
   
   
Comment:
   

Attached Files
file #24444:  1229-uk.jpg added by Stevan_White (37KiB - image/jpeg - uk from 1229 Treaty between Smolensk, Riga and Gotland)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by malo_skr (Posted a comment)
  • -unavailable- added by Stevan_White (Posted a comment)
  •  

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

    Only logged-in users can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 8 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Thu 03 May 2012 06:04:25 PM UTCStevan_WhiteStatusFix posted=>Won't fix
      Open/ClosedOpen=>Closed
    Fri 25 Nov 2011 02:12:19 PM UTCalecesCarbon-CopyRemoved 85679=>-
    Thu 24 Nov 2011 11:18:52 PM UTCStevan_WhiteSummaryBad shape of Cyrillic U+0478 & u+0479=>Bad shape of Cyrillic letter Uk (U+0478 & u+0479)
    Thu 24 Nov 2011 11:18:09 PM UTCStevan_WhiteAttached File-=>Added 1229-uk.jpg, #24444
    Fri 18 Nov 2011 09:59:01 AM UTCStevan_WhiteStatusNeed info=>Fix posted
    Sat 29 Oct 2011 10:51:56 PM UTCStevan_WhiteStatusNone=>Need info
      Assigned toNone=>Stevan_White

    Back to the top


    Powered by Savane 3.1-cleanup1