The Unicode Consortium Discussion Forum

The Unicode Consortium Discussion Forum

 Forum Home  Unicode Home Page Code Charts Technical Reports FAQ Pages 
 
It is currently Fri Aug 01, 2014 4:53 am

All times are UTC - 6 hours [ DST ]





Post new topic Reply to topic  [ 17 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Letterlike Symbols
PostPosted: Fri Jun 08, 2012 6:55 pm 
Offline

Joined: Tue Oct 11, 2011 4:34 pm
Posts: 19
This is the breakdown for the Other Symbols (So) in the Letterlike Symbols block.
To summarize, it seems MVO=U is a good answer for most of them, but
* U+2129 TURNED GREEK SMALL LETTER IOTA should be changed to MVO=R
* The apothecary measures (U+2108, U+2114, U+2125), inverted ohm (U+2127), and per sign (U+214C) might need more thought

East Asian Precomposed Degrees

Code:
℃   U+2103  DEGREE CELSIUS
℉   U+2109  DEGREE FAHRENHEIT


These are EAW=A. It probably make sense for them to stay upright; the decomposed °C can be used for sideways.

Communications and Addressing

Code:
№   U+2116  NUMERO SIGN
℡   U+2121  TELEPHONE SIGN
℻   U+213B  FACSIMILE SIGN


The first two in this list are EAW=A, and ℻ should be consistent with ℡.
Probably makes sense to stay upright.

Intellectual Property Rights

Code:
©   U+00A9  COPYRIGHT SIGN
®   U+00AE  REGISTERED SIGN
℗   U+2117  SOUND RECORDING COPYRIGHT
™   U+2122  TRADE MARK SIGN
℠   U+2120  SERVICE MARK


These should remain consistent. I don't have an opinion on whether they should have MVO=U or MVO=R.

Slashed Abbreviations

Code:
℀   U+2100  ACCOUNT OF
℁   U+2101  ADDRESSED TO THE SUBJECT
℅   U+2105  CARE OF
℆   U+2106  CADA UNA
⅍   U+214D  AKTIESELSKAB


These should remain consistent. I don't have an opinion on whether they should have MVO=U or MVO=R.

Apothecary Measures

Code:
℈   U+2108  SCRUPLE
℔   U+2114  L B BAR SYMBOL
℥   U+2125  OUNCE SIGN


These are used with together with ʒ (U+0292), which is MVO=R, but sometimes also with ♏ (U+264F), which is MVO=U.

Other Units

Code:
℧   U+2127  INVERTED OHM SIGN


This is the inverse of Ohm. Unicode recommends using omega for ohm, which is MVO=R, but there is also the letterlike symbol U+2126 OHM SIGN, which is MVO=U. Either option would make sense.

Musical Symbols

Code:
℣   U+2123  VERSICLE
℟   U+211F  RESPONSE


These are used in musical notation, so should be consistent with the musical symbols, which are MVO=U.

Architectural Symbols

Code:
℄   U+2104  CENTRE LINE SYMBOL
⅊   U+214A  PROPERTY LINE


Afaict, these are used in blueprints. Probably MVO=U makes sense.

Miscellaneous

Code:
℞   U+211E  PRESCRIPTION TAKE (indicates medical prescription)
℮   U+212E  ESTIMATED SYMBOL (used in EU packaging to label amounts)


I have no idea how to pick an orientation for these. Probably MVO=U is alright.

Code:
℩   U+2129  TURNED GREEK SMALL LETTER IOTA


The unicode code charts say that this is used with logic, so probably should be consistent with math symbols. I therefore recommend changing MVO to R.

Code:
⅌   U+214C  PER SIGN


This is an abbreviation for the word "per". It might make sense to keep it consistent with U+204A TIRONIAN SIGN ET, which is currently MVO=R; so either change this to R or make that one U?

Code:
℺   U+213A  ROTATED CAPITAL Q


This is used as a binding mark, and should be consistent with the symbols U+203B REFERENCE MARK, U+2720, and U+2767, which are MVO=U.

Code:
⅏   U+214F  SYMBOL FOR SAMARITAN SOURCE


This doesn't seem letterlike to me. Probably MVO=U is alright.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Sat Jun 09, 2012 7:51 am 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
There are some code points that are considered as symbols for Asians, but as letters for some scripts. Letterlike Symbols are one of the most obvious example, because they're "Symbols," but also "Letterlike."
The correct answer that suffices everyone does not exist. It's a trade-off, and it depends on where UTR#50 would focus on.
I prefer setting such ambiguous code points to U for reasons below:
  • This will make typesetting multi-lingual documents in vertical harder, but such usage would require markups anyway. They'd better set with fonts that are suitable for the script, or some characters such as dashes will not set properly. One can set orientation to sideways when they change fonts, so they wouldn't lose much by setting to U.
  • Ambiguous code points that are widely used in Asia as symbols are U. Having different orientation because a code point is less often used, or not defined in legacy encoding, doesn't look a consistent way to me.
  • Fonts tend to implement by blocks. Setting different policy by its origin is likely to give inconsistent results.
Either way we do, it makes some kind of consistency, and lose the other kind of consistency. I don't know which usage is lower than the other, but it's pretty clear that neither is high, which makes the discussion harder.
Since this is a discussion of trade-off rather than finding the correct answer, I would like careful discussions here.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Sat Jun 09, 2012 2:35 pm 
Offline

Joined: Sat Jan 14, 2012 4:10 am
Posts: 29
Most modern Japanese fonts have '©' U+00A9, '®' U+00AE, '™' U+2122 characters and these are usually set sideways in vertical writing applications. Modern Japanese fonts have 'vrt2' OpenType feature to rotate glyph for these characters.

InDesign does not use the 'vrt2' feature, but rotates these characters by default.
See the following PDF file created by InDesign for testing vertical orientations with popular Japanese fonts, Meiryo, Hiragino Mincho, Kozuka Mincho.
https://dl.dropbox.com/u/64715801/AdobeIndVert.pdf

I think this results are almost natural.
Probably there are some reason that 'ℓ', 'Å', '℃', '℉', '№', '℡' are upright (though I think it is not very good when they are used together with rotated digits).
Other characters are rotated, consistent with latin letters. They are not often used in Japanese vertical text and when upright orientation is needed for such characters (an example of upright ©), we can use the SVO mode or specify tate-chu-yoko or upright orientation explicitly, same as the case of Greek letters.

I don't want exactly same result with InDesign and I don't have strong opinion about these characters orientation, but I believe that the MVO should be good default for East Asian and Western mixed text composition while SVO is good for pure East Asian text.


Shinyu Murakami
Antenna House


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Sun Jun 10, 2012 2:24 am 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
I think ©®℗™℮℧ are really "ambiguous" in the real-world usages. We see not a few cases where these are set "upright" in vertical lines in Japanese. Still, as these are simply abbreviated symbols for the respective original English words, and these are used in today's English texts frequently. Therefore, these default posture should be "rotated" in vertical lines. But, I understand these are rather difficult examples.

Next,
℁℅℆

These three are pure English abbreviations, and should be "rotated" in vertical lines, though still I can imagine cases where ℅ may be set upright in vertical lines, for example, in a Japanese sentence like 「記号℅を気付の場合には用いる」("Use the ℅ symbol to mean 'care of'"), which teaches the reader how to write an address in English. But this is a rare, meta-level, self-referencing example, and can be ignored.

Just my thoughts. . .


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Mon Jun 11, 2012 9:52 pm 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
MurakamiShinyu wrote:
I believe that the MVO should be good default for East Asian and Western mixed text composition while SVO is good for pure East Asian text.

I don't think so. MVO is supposed to be used in CSS and in many applications as a default value, so it should target the majority use in East Asia.

East Asian and Western multi-lingual document is not the majority use case for vertical flow, because authors tend to set such documents in horizontal. Readers to switch directions between U and R is known to slow reading down, so documents that require a lot of R makes more sense to set in horizontal.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Tue Jun 12, 2012 7:21 pm 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
I don't think the source of the ambiguity of these characters consists in our reluctance to put higher priorities to their "Asian" usages than to the original Western usages.

Rather, I think that the source of, at least, some of the ambiguity is in that these characters cannot be naturally fit to "Asian" usage contexts accurately, and their typical usages in vertical lines are not yet stably, widely established.

This makes it possible that we tend to think that we should treat them as Western glyphs by default, though we all know that different "Asian", "Upright" usages in vertical lines for these characters are also widely used.

I don't think this approach viz. "If ambiguous, return to the original posture", works always. But it seems to work for the characters being discussed here.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Tue Jun 12, 2012 11:07 pm 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
I'm sorry, your English is a little too difficult for me to understand, and I don' think I got the point very well, so my reply might be pointless, apologize if that's the case.

If I understand correctly, I don't think I could tell you my thoughts correctly, so please allow me to try to rephrase.

I agree that the ambiguity doesn't change priority; I always put higher priority to Asian usage regardless of the ambiguity. If Asian usage and Latin usage doesn't conflict, it doesn't matter though.

Now the ambiguity came in. Our choice here is either to create a new mechanism to manage ambiguity, or make a trade-off. And I don't think making a new mechanism now is a good idea, because it will delay UTR#50 further and so many people are waiting for the spec to stabilize. The only option I'm seeing is to make a trade-off, and I hope all agree with this.

What I'm saying is, if we have to make a trade-off, we should use priority-based decisions, rather than discussing on one bad example, or one people took 5 samples and determine that it looks like it's widely used or not. It's very easy to find bad examples whichever we decide because we're making trade-off. And my priority for vertical flow is always on Asian usage.

I'm talking to developers, authors, publishers, printers, and so forth. Most developers prefer higher priority on Latin usage, and I'm aware of that. Authors, editors, publishers, and printers I'm talking to prefer higher priority on Asian usage. I'm seeing conflicting opinions here. I really wish if we were able to support both and make everyone happy, but at this point we can't due to technical problem and resource issues to resolve the problem in short term.

So this is a kind of triage-type question that who and what scenario do we really need to save at the first place. I suppose it's Asian usages, and authors.

It's very unfortunate for people who wants to use vertical flow in multi-lingual documents, and I feel sorry about that, but we're making this extensible so that we could support more scenario better in future.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Wed Jun 13, 2012 2:43 am 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
How can I express my idea better? I'll try to explain it some more.

For example, I don't deny that there are cases where the "care of" symbol may be set UPRIGHT in vertical lines.



I introduced one rare example of its being set UPRIGHT in vertical lines in my previous comment.

However, it is really ambiguous or unclear, whether this UPRIGHT usage can be regarded as the typical Japanese or Asian vertical usage of the character.

The stability of such a usage itself is unclear.

Because of this uncertainty, we cannot say that the character can be used in both two ways UPRIGHT and ROTATED, one of which is the "Asian" usage, and that the character's posture in vertical lines is "ambiguous", and that therefore we should put a higher priority to the "Asian" (upright) usage.

Such a case is doubly ambiguous.

Before we solve the first ambiguity or uncertainty question, we cannot move to the main question, can we?
If the first ambiguity remains unsolved, I think we should let it conform to the posture of the original language.

I myself don't know when we use the "care of" symbol, other than the rare case that I could imagine and introduce in one of my previous comments. Isn't the word "care of" a pure English word?

The same thing seems to apply to the following characters:
©®℗™℮℧


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Wed Jun 13, 2012 12:25 pm 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
I think I understand your points much better this time, thank you for your explanation.

I agree on your first point that these characters are ambiguous.

However, I don't agree with the conclusion that a character should follow its original language if ambiguous, especially if it's symbols and punctuation characters. I believe, if ambiguous, set it to upright.

We could decide based on how common a character is oriented, but I'm not a big fan of that method. It's not that hard to know zero or not, but "which is more common" is really hard to get the correct answer, and even if we could get it right, the answer is fragile; it could change in years. Question mark was imported only less than 100 years ago. TEL and No much less. We don't know which symbols are commonly used in 5 years.

If we could agree on not to rely on "how common," the only remaining question is whether a use of a symbol appearing within East Asian scripts should need markup, or a use of a symbol in combination of Arabic digits and Latin words should need markup.

I'd like to put higher priority on the former, because the more long Arabic digits and Latin words you have in a document, the more likely that the document is set in horizontal.

Setting the later kind of documents in vertical flow costs high, but its primary cost is reviewing and deciding orientation for each occurrence. Even a same word can be set differently in a single page (a sample here, find two "Growl",) so they can't get away from markup whatever we do.

On the other hand, if you set them to upright, it helps the former. Documents that consist of mostly East Asian scripts don't require much markup to set in vertical flow, so requiring them to review and markup symbols by when it was imported and by their origin doesn't make much sense to me.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Wed Jun 13, 2012 11:54 pm 
Offline

Joined: Sat Jan 14, 2012 4:10 am
Posts: 29
kojiishi wrote:
On the other hand, if you set them to upright, it helps the former. Documents that consist of mostly East Asian scripts don't require much markup to set in vertical flow, so requiring them to review and markup symbols by when it was imported and by their origin doesn't make much sense to me.

Unfortunately, MVO=U for "ambiguous" symbols (‰©®™ etc.) can't help much. Digits, Latin letters, Greek letters (e.g.,「β版」) are very often set upright but they are already MVO=R and many markups are required when we use the mixed mode as default. You have to specify 'Upright' to MVO=R characters and have to specify 'Rotate' to MVO=U characters.

If all "ambiguous" characters are MVO=R, the markup task will be simpler: you have to specify only 'Upright'.

Of course when we use the stacked mode (SVO) as default, we have to specify only 'Rotate' markup (e.g., using 'lang' attribute). This way can minimize the markup task for most East Asian vertical layout text.


Shinyu Murakami
Antenna House


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Thu Jun 14, 2012 2:37 am 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
You're still talking about technical documents, which I set lower priority on, because they're more unlikely to be set in vertical flow.

The majority of Asian documents that are set in vertical flow use Han digits rather than Arabic digits, so setting units and symbols to U makes sense without any markups.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Thu Jun 14, 2012 5:10 am 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
kojiishi wrote:
You're still talking about technical documents, which I set lower priority on, because they're more unlikely to be set in vertical flow.

The majority of Asian documents that are set in vertical flow use Han digits rather than Arabic digits, so setting units and symbols to U makes sense without any markups.


No, I don't think so. The correct way is to replace the unit symbol, often incorrectly inserted by someone or because it's just an early draft manuscript, with a set of katakana characters corresponding to the pronunciation of the unit name.

We should not compose characters and symbols in an arbitrary style, simply because the style is possible. We should use characters and symbols in correct ways.

I can agree some symbols are widely used in the UPRIGHT posture in vertical lines. But, I don't believe the "care of" symbol or the "per mill" symbol etc. should be UPRIGHT.
How can they be set UPRIGHT?

In the case of "per mill", I suspect the horizontal width can exceed the EM box boundary, depending on the typeface used, when set UPRIGHT in a vertical line.

Regards,


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Thu Jun 14, 2012 7:43 am 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
How do you resolve when the correctness is different among people? That's what I'm seeing in Japan.


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Thu Jun 14, 2012 5:45 pm 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
kojiishi wrote:
How do you resolve when the correctness is different among people? That's what I'm seeing in Japan.


I discussed it with the specific examples of the "per mill" and the "care of" symbols. My discussion was based on examples.

But you didn't answer any of the points that I wrote, viz. (1) that katakana characters, instead of unit symbols, should be used when using Chinese numbers in vertical lines, and (2) that the "per mill" glyph's width may exceed the EM body of the parent line, when set in the UPRIGHT posture.

It may be true that we cannot reach perfect consensus to all issues about all characters. But, I believe our experience and knowledge in Japanese typography can give "correct" answers to not a few of them.

I already pointed out the two real examples to which we already had "correct" answers, didn't I?

Please answer why do you believe that when using numbers in Japanese, the styles of "五〇%" and "50%" are more typical "East Asian" typesetting style in vertical lines than the style of "五〇パーセント"? Every serious typographer knows that the latter should be the "correct" style of typesetting these in vertical lines. (If you don't like the rewriting of the word "percent", you can use a ligature glyph, as long as your font supports it).

I think this applies to every unit symbol that has its origin in a Western word.
So, Western unit symbols should NOT be UPRIGHT by default in vertical lines, because the UPRIGHT posture is not genuinely "EAST ASIAN". (I agree many people may use it in the UPRIGHT posture. It may have been influenced by the ambiguity in handling JIS full-width characters. Yes, it is possible, but the style is NOT the most typically "East Asian", for the reasons mentioned above).


Top
 Profile  
 
 Post subject: Re: Letterlike Symbols
PostPosted: Thu Jun 14, 2012 5:55 pm 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
As the "care of" symbol is not a unit symbol, I need to add the following comments. But about the "care of" symbol, the answer is very simple. It should not be UPRIGHT by default in vertical lines, because it has no usages in Japanese. It's purely an abbreviation of English words.

So, which posture is the most typical "East Asian"?
The question doesn't apply to this character well. The answer should be "Nothing". It means it should be treated as a Western character.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 17 posts ]  Go to page 1, 2  Next

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 3 guests


Quick-mod tools:
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
Template made by DEVPPL.com