The Unicode Consortium Discussion Forum

The Unicode Consortium Discussion Forum

 Forum Home  Unicode Home Page Code Charts Technical Reports FAQ Pages 
 
It is currently Mon Sep 01, 2014 6:27 pm

All times are UTC - 6 hours [ DST ]





Post new topic Reply to topic  [ 29 posts ]  Go to page Previous  1, 2
Author Message
 Post subject: Re: Fundamental questions
PostPosted: Sat Jun 16, 2012 10:20 am 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
MurakamiShinyu wrote:
Greek and Cyrillic letters and "ambiguous" symbols are U.

I don't think anybody needs Greek and Cyrillic be U, including publishers, as far as I heard. What the target user?


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Sat Jun 16, 2012 5:33 pm 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
(I think the explanation of my previous comment is not sufficient in avoiding misinterpretations, I should add some more).

Though I wrote there are no serious typographers who use full-width Arabic number glyphs for multiple digits, and full-width numbers can be used only for stand-alone single digit numbers (used as numbers or item numbers). This applies to both writing directions. And, as far as full-width Arabic numbers are used, their posture is ALWAYS UPRIGHT, regardless of the writing direction.

If so, what will happen to the "precisely half-width" Arabic number glyphs used for multiple numbers? In the horizontal writing mode, they are naturally UPRIGHT without any transformation. In the vertical mode, if the Arabic numbers are encoded in the original full-width codes, the posture can remain UPRIGHT, but still you need to apply TCY to the "precisely half-width" Arabic number digits. So, depending on the leading between lines, the maximum number of digits can be limited to 2 or 3.

I don't oppose using proportional Arabic numbers (−90 degrees ROTATED) in vertical lines, if appropriate, and in this case, any number of digits can be set.

But I mean that the usage of composing UPRIGHT Arabic numbers, full-width or half-width, in vertical lines are very limited, and it's not the typical Japanese or "East Asian" usage, though there are some methods for doing it, as I itemized above.

The most widely accepted standard is converting them into Chinese numbers.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Sat Jun 16, 2012 9:29 pm 
Offline

Joined: Sat Jan 14, 2012 4:10 am
Posts: 29
kojiishi wrote:
MurakamiShinyu wrote:
Greek and Cyrillic letters and "ambiguous" symbols are U.

I don't think anybody needs Greek and Cyrillic be U, including publishers, as far as I heard. What the target user?

If nobody needs such spec the CVO was a bad idea and please forget it.

By the way, is there anybody who needs ‰, ‱, ©, ®, ™, ℀, ℅, ℧, ℮ etc. be U by default, in publishers?
I heard that publishers can accept any default orientation for ambiguous characters because they can specify U or R to those characters explicitly.

Greek and Cyrillic letters are typical ambiguous characters in Japanese text processing. They (especially Greek) are often used as symbols and most Japanese fonts have only full-width glyphs for those characters and they are set UPRIGHT by default in today's vertical writing softwares including InDesign and MS Word.

If you have the policy "ambiguous characters should be U", it is very natural that Greek and Cyrillic letters are U.

In my opinion, as repeated, if Greek and Cyrillic letters are R, other ambiguous characters should also be R. This policy will make the Mixed Vertical Orientation spec more consistent and understandable, it will be good for Japanese publishers and also good for multilingual world.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Sun Jun 17, 2012 9:10 am 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
tyamamot wrote:
emuller wrote:
Quote:
<verticalText>
2012年6月<tcy>15</tcy>日
</verticalText>

YYYY: 2012 /*2, 0, 1, 2 fullwidth variant */
MM: 6 /* fullwidth variant */
DD: 15 /* ASCII */


or DD in fullwidth characters + half width styling.

Isn't it what people do today? And in horizontal as well, minus the tcy?


No. There are no serious Japanese typographers who set multiple Arabic number digits with their full-width glyphs, in whatever method the shapes and widths are implemented. Only junk, third-rate typographers incorrectly, casually, happen to make this error, due to their ignorance. Or, I have to admit that our users may make the same mistake, and that it may be due to the imperfection of our user education.


Japanese newspapers adopt the writing method which sets each Arabic number digit upright for calendar years or large numbers.

Typically, "Handbook for Editors & Writers" published from K.K.Kyodo News recommends such writing.

http://blog.cas-ub.com/wp-content/uploads/2012/06/shinbun.png (added on 18/06/2012)

BTW: Today, I found that each digit was set upright in the phrase "1秒の299792458分の1” in Japan Economic Journal. :-)

tyamamot wrote:
It is possible that a stand-alone single-digit full-width number may be used, if (1) used as a symbolic stand-alone number for each item in an itemized list, or (2) used purely as a single digit number to indicate a month or date, etc.


There are another usage of Arabic number digit in vertical writing printed books such as:

(3) Page numbers in table of contents. Page number of each chapter and/or section is set with TCY even if it has more than three digits.
(4) Chapter and/or section numbers in TOC or main contents.
(5) Reference of chapter and/or section numbers within main contents.
(6) Reference of figure number and/or table number within main contents. In vertical writing book, caption of figure or table will be shown as horizontal line and, at the same time, figure number and/or table number are shown with Arabic number digits. So, references will be in Arabic number digits within vertical line.
(7) Page number citation within main contents. Page number in footer/header is normally set Arabic number digits, and it will often be 3 digits. As for citation in main contents, some books convert it to Kanji number digits, some books set it Arabic number digits with TCY even if it has three digits.

tyamamot wrote:
Still, it is clear that mixing the full-width and proportional number glyph styles is NOT elegant at all. So, most serious typographers don't want to use the easiest solution, and prefer using genuinely proportional numbers only. It is obvious this needs more careful spacing between characters before and after the number(s).


It seems to me that this paragraph mentions to horizontal writing. When Arabic number digit appears in vertical line, more than 90% is upright[*1]. In case of upright, there is no spacing between characters before and after the number(s) [*2]. In case of sideways, Arabic number digit is in the sequence of Latin alphabets where English formatting rule applies.

[*1] There is a typical example of frequencies: http://blog.cas-ub.com/?p=1885 (I am sorry that this is Japanese.)
[*2] http://www.w3.org/TR/jlreq/#method_for_setting_fullwidth_latin_letters_and_european_numerals


tyamamot wrote:
Anyway, what I can I recommend to you is, at least, is, in Japanese typographic contexts, not to compose multiple digits with full-width Arabic number glyphs. Also, in vertical lines, most numbers should be rewritten or trans-coded into their Chinese equivalent characters,


This is out of the scope of UTR#50.

tyamamot wrote:
with few exceptions such as cases where TCY is applied. But the TCY usage for one or two-digit Arabic numbers is also just one of many expedient tricks, and not a stably established Japanese typographic convention.


TCY is quite frequently used in Japanese vertical writing books, newspapers and magazines. Reading this paragraph, it is clear that he has not checked any actual printed matters seriously.

tyamamot wrote:
In most serious cases, the first principle of setting numbers in Japanese vertical lines is that they should be set in Chinese number characters, and when a unit name is appended, it should be written in katakana characters.


This is a traditional writing style.

There is another style; Arabic number digits in vertical writing, whose main user is newspaper industry. Next, such style is used in magazines as well as in some percentage of vertical writing books.

Here is a example of Japanese personal computing magazine, Nikkei PC 21.

http://blog.cas-ub.com/wp-content/uploads/2012/02/NPC21-2012-3-66.png

Regards,

Tokushige Kobayashi
Antenna House, Inc.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Mon Jun 18, 2012 12:10 am 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
Quote:
Japanese newspapers adopt the writing method which sets each Arabic number digit upright for calendar years or large numbers.
Typically, "Handbook for Editors & Writers" published from K.K.Kyodo News recommends such writing.
BTW: Today, I found that each digit was set upright in the phrase "1秒の299792458分の1” in Japan Economic Journal. :-)

I don't say the UPRIGHT posture cannot be used. TCY is also possible. But it's just possible. It's not the typical Japanese usage. No one stops you to set Arabic numbers upright in vertical lines, though.
Quote:
There are another usage of Arabic number digit in vertical writing printed books such as:
(3) Page numbers in table of contents. Page number of each chapter and/or section is set with TCY even if it has more than three digits.

If the stylistic consistency is required, you should use Chinese characters. But I don't prohibit you to use UPRIGHT Arabic numbers in vertical lines. One-digit full-width Arabic number glyphs and half-width number glyphs have design inconsistency, which is not desirable in many cases, though it's possible.
Quote:
(4) Chapter and/or section numbers in TOC or main contents.
(5) Reference of chapter and/or section numbers within main contents.
(6) Reference of figure number and/or table number within main contents. In vertical writing book, caption of figure or table will be shown as horizontal line and, at the same time, figure number and/or table number are shown with Arabic number digits. So, references will be in Arabic number digits within vertical line.
(7) Page number citation within main contents. Page number in footer/header is normally set Arabic number digits, and it will often be 3 digits. As for citation in main contents, some books convert it to Kanji number digits, some books set it Arabic number digits with TCY even if it has three digits.

The same thing applies to all these. I don't prohibit you from using the style. But these sare all peripheral cases.
Quote:
It seems to me that this paragraph mentions to horizontal writing. When Arabic number digit appears in vertical line, more than 90% is upright[*1]. In case of upright, there is no spacing between characters before and after the number(s) [*2]. In case of sideways, Arabic number digit is in the sequence of Latin alphabets where English formatting rule applies.

Right. I meant the horizontal usage.
Quote:
Quote:
Anyway, what I can I recommend to you is, at least, is, in Japanese typographic contexts, not to compose multiple digits with full-width Arabic number glyphs. Also, in vertical lines, most numbers should be rewritten or trans-coded into their Chinese equivalent characters,

This is out of the scope of UTR#50.

Yes. So, I meant that we could not make any decisions.
Quote:
Quote:
TCY is quite frequently used in Japanese vertical writing books, newspapers and magazines. Reading this paragraph, it is clear that he has not checked any actual printed matters seriously.

I dont't mean that TCY is not used frequently. TCY is used frequently, as you wrote. But just that. TCY is not a generally applicable method, because it is constrained by the leading between lines.

Quote:
Quote:
In most serious cases, the first principle of setting numbers in Japanese vertical lines is that they should be set in Chinese number characters, and when a unit name is appended, it should be written in katakana characters.

This is a traditional writing style.
There is another style; Arabic number digits in vertical writing, whose main user is newspaper industry. Next, such style is used in magazines as well as in some percentage of vertical writing books.

I totally agree with you. TCY and full-width Arabic numbers should be possible. No one must prohibit them. About this, no difference of opinion we have.

Also, already, many Arabic numbers and symbols are doubly encoded, for the differentiation of Full-width and Proportional glyphs. So, we don't need to this cuss these doubly encoded characters, because either posture can be selectable.

The main things about which we need to make decisions are other symbols and marks.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Tue Jun 19, 2012 5:18 pm 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
tyamamot wrote:
The only exception may be that in order to achieve the above-mentioned "full-width and purely half-width" numbers on InDesign. :-)

You first use the full-width numbers (encoded as full-width numbers), and only to the affected multiple-digits numbers, you apply the GSUB feature to replace the glyphs with the corresponding "precisely half-width glyphs". In this case, the full width coded characters are used only as an entry point for the glyph substitution to realize the "precisely half-width" number glyphs. As mentioned above, pure full-width Arabic number characters may be used only for one-digit numbers.


This is the special feature of InDesign. it is very interesting how InDesign converts such multiple Arabic numbers encoded as full-width variant into XHTML markup for EPUB and/or Web.

InDesign treat the following vertical text without problem.

<verticalText>
2012年6月<tcy class='replace to half-width by GSUB'>15</tcy>日
</verticalText>

YYYY: 2012 /*2, 0, 1, 2 fullwidth variant */
MM: 6 /* fullwidth variant */
DD: 15 /* fullwidth variant */

But, browser and EPUB reader will not support "class='replace to half-width by GSUB'". Does InDesign clear tcy, or change codepoint ?

Regards,

Tokushige Kobayashi


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Tue Jun 19, 2012 5:43 pm 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
tyamamot wrote:
I totally agree with you. TCY and full-width Arabic numbers should be possible. No one must prohibit them. About this, no difference of opinion we have.


We agree. But unfortunately, this is not an important point. Important point is:

(1) In vertical writing text, Arabic numbers are popular in newspapers, magazines and even in books. And tcy is also used frequently.
(2) Arabic numbers are upright with or without tcy more than about 90% and sideways less than about 10% in frequency. Especially, 100% upright in newspaper.

tyamamot wrote:
Also, already, many Arabic numbers and symbols are doubly encoded, for the differentiation of Full-width and Proportional glyphs. So, we don't need to this cuss these doubly encoded characters, because either posture can be selectable.


This brings big incompatibility between horizontal text and vertical text. In horizontal text, we normally use ASCII codepoint for Arabic number. In vertical text and if MVO=R for ASCII codepoint we must use its full-width variant. As a result, MVO=R for ASCII codepoint will be quite inconvenient for changing horizontal writing to vertical writing.

Jumping to a conclusion, I am wondering MVO may not be necessary.

Regards,

Tokushige Kobayashi
Antenna House, Inc.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Wed Jun 20, 2012 12:40 am 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
TKobayashi wrote:
But, browser and EPUB reader will not support "class='replace to half-width by GSUB'". Does InDesign clear tcy, or change codepoint ?

I guess it's 'hwid' feature in OpenType, and author can enable the feature using font-feature-settings in CSS3 Fonts. I don't know of any UA that implements this and vertical flow as of today, but the spec is available.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Wed Jun 20, 2012 12:46 am 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
TKobayashi wrote:
Jumping to a conclusion, I am wondering MVO may not be necessary.

MVO is the real request from real authors and vendors. It's just technically difficult, and we must make some compromises to reach consensus because discussion is against where no single correct answer exists as you pointed out before.

Nobody here wants to compromise as of now, that's a big issue, but it doesn't necessarily mean that the request was invalid.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Wed Jun 20, 2012 9:25 am 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
Quote:
This is the special feature of InDesign.

No. a wide range of GSUB features are defined by the OpenType font format specification, and as the CSS standard enahnces its GSUB features in the future, more glyph substitution features will be available and controllable from CSS.
Quote:
In vertiical writing text, Arabic numbers are popular in newspapers, magazines and even in books. And tcy is also used frequently.

As I have repeated for many times. I don't deny that Arabic numbers as well as TCY numbers are used in Japanese publications. But it remaind to be an expedient solution, because the style may result in bad, optically uneven spacing between characters, and the TCY method has limitations constrained by leadings. Because use of Arabic numbers and roman glyphs designed within the EM box type body themselves are just a historical compromise in Japanese typography. They can be used, if necessary. But they should not be used in book typography, though sometimes you need to use the style.

Also, newspaper typesetting in Japan has very many special limitations and conditions, which may not apply to the general usages of Japanese type.

Even if you specify your type size to be 9-pt in the text composition stage, for example, you won't see the same result (type size), when the newspaper is printed out. Also, its composition rules are constrainted by various conditions related to the distribution of adviertisement areas. Newspaper is a special field.

Quote:
(2) Arabic numbers are upright with or without tcy more than about 90% and sideways less than about 10% in frequency. Especially, 100% upright in newspaper.


I agree that TCY or whatever style should be selectable for vertical numbers. I agree that use of Chinese numbers is not the only method.

But the basic start point of usages of type should be based on traditional book typography.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Tue Jun 26, 2012 6:25 pm 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
fantasai wrote:
So, I think Koji does not agree, but here are my thoughts on how to categorize:

For SVO

  • If it's at all sensible to put as U, then the codepoint should be U.
  • If it's only sensible to put as R, then the codepoint should be R.

Most characters, it's sensible to make them U. Some characters (mainly dashes and enclosing punctuation like brackets) it never makes sense to make them U, so they should be R.



In my understanding, SVO tries to define the minimum set of characters which shall be R in vertical line.

I prefer to select SVO, in case with the addition of Devanagari or such characters to R.

MVO tends to be ambiguous and it will bring confusion. Moreover the double standard for ASCII characters and their Full-width variant is bringing confusion among professional and non-professional writers already. I am considering MVO may not be necessary for Japanese vertical writing. Though more investigation will be necessary if it is possible to set any instance of Japanese vertical writing easily by using SVO + CSS writing mode.

Regards,

Tokushige Kobayashi


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Wed Jun 27, 2012 1:23 am 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
Making most characters in Japanese vertical lines UPRIGHT is not sensible at all. It's sheer destruction of Japanese typography and it will help making it difficult for us to improve Japanese typography in the world wide web.

It is true that newspaper articles tend to use upright Arabic numbers, because they wanted to simplify and regulate the usages of characters, numbers and symbols, as far as their main articles are concerned, for efficiency. Also, they use 2,000 or so Joyo-kanji Chinese characters only, with exceptions. However, this does not apply to Japanese publishing and typography in general. In newspapers, only vertical lines are used for main texts, but in Japanese typography in general, both directions can be used. Also, there are many newspaper-oriented restrictions about use of type other than the writing direction, such as limitation to type sizes and use of vertically compressed type, etc.

However, if you look at a serial newspaper novel, you can find many instances of Chinese numbers, even on a newspaper page, because it is the most authentic method of composing numbers in vertical lines. Writers of such newspaper novels need not use the old "newspaper-only" typographic styles. Also, whether classified or unclassified, look at advertisement on newspaper pages. You will find no advertisers stick to the narrow "newspaper-only" restrictive usages of type.

We should try to liberate type users from the old narrow alleys of the Japanese newspaper industry where there were many technical and economical limitations and constraints. It's time to provide them with a widened, expanded scope of more sophisticated styles of Japanese typography, instead of pushing them and their readers into a dead end of typographic expressions that might have existed in the past and had reasons.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Fri Jul 13, 2012 10:44 pm 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
fantasai wrote:
So, I think Koji does not agree, but here are my thoughts on how to categorize:

For SVO

  • If it's at all sensible to put as U, then the codepoint should be U.
  • If it's only sensible to put as R, then the codepoint should be R.

Most characters, it's sensible to make them U. Some characters (mainly dashes and enclosing punctuation like brackets) it never makes sense to make them U, so they should be R.


Such characters as U+2329, 232A, 2768 .. 2775, 3000, FE4D .. FE4F are SVO=R MVO=U in http://unicode.org/reports/tr50/tr50-5.Orientation.html. If it never makes sense to make them U in SVO, why it makes sense U in MVO ?

Regards,

Tokushige Kobayashi
Antenna House, Inc.


Top
 Profile  
 
 Post subject: Re: Fundamental questions
PostPosted: Fri Jul 13, 2012 10:52 pm 
Offline

Joined: Wed Dec 07, 2011 3:01 am
Posts: 71
TKobayashi wrote:
Such characters as U+2329, 232A, 2768 .. 2775, 3000, FE4D .. FE4F are SVO=R MVO=U in http://unicode.org/reports/tr50/tr50-5.Orientation.html. If it never makes sense to make them U in SVO, why it makes sense U in MVO ?


Thank you for reporting. The code points you listed are all fixed to MVO=R. Any code points where SVO=R and MVO=U are likely to be bug.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 29 posts ]  Go to page Previous  1, 2

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


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