Closed Bug 88579 Opened 24 years ago Closed 21 years ago

Bad default Traditional Chinese font (MS Windows)

Categories

(Core :: Internationalization, defect)

x86
Windows 98
defect
Not set
normal

Tracking

()

RESOLVED FIXED

People

(Reporter: lawrenceho, Assigned: jshin1987)

References

Details

(Keywords: fixed-aviary1.0, fixed1.7.5, intl)

Attachments

(9 files, 1 obsolete file)

From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; rv:0.9.2) Gecko/20010628 BuildID: 2001062815 Mozilla chose an older font (???) as default when newer, better font (??? ?) is installed by default on, IIRC, MS Windows 95 and above.
ylong, could you check this?
I checked it on 07-02 Branch build, the tradition chinese default font set to "MingLiu". Lawrence Ho: I can not read the font name in your description, could you please give the font names by changing the charset? Thanks!
Switching QA contact to ylong@netscape.com.
QA Contact: andreasb → ylong
I've checked my Win98SE setup, both the old version and the new version of the font is stored in the same file, MINGLIU.TTC. I wonder if Mozilla can be configured to use the new version as the default instead of the ugly old version. I'm not sure how to make Chinese chars appear correctly in the bug report, let me try here: (Charset: Big5) Old version: 細明體 New version: 新細明體
Status: UNCONFIRMED → NEW
Ever confirmed: true
The current default is set to the old font. http://lxr.mozilla.org/seamonkey/source/modules/libpref/src/win/winpref.js Is the new one available for all version of Windows? Shanjian, how can I specify fallback? We could put the new one as the default then put the old one to the fallback list.
Keywords: intl
I think we need to tune the default font list in modules/libpref/src/win/winpref.js see http://lxr.mozilla.org/seamonkey/source/modules/libpref/src/win/winpref.js We may need to add additional fall back. 98 pref("font.name.serif.zh-TW", "細明體"); // "MingLiU" 99 pref("font.name.sans-serif.zh-TW", "細明體"); // "MingLiU" 100 pref("font.name.monospace.zh-TW", "細明體"); // "MingLiU" 101 pref("font.name.serif.zh-TW.1", "MingLiU"); 102 pref("font.name.sans-serif.zh-TW.1", "MingLiU"); 103 pref("font.name.monospace.zh-TW.1", "MingLiU"); we probably should change 98 pref("font.name.serif.zh-TW", "新細明體"); // "MingLiU" 99 pref("font.name.sans-serif.zh-TW", "新細明體"); // "MingLiU" 100 pref("font.name.monospace.zh-TW", "新細明體"); // "MingLiU" and add pref("font.name.serif.zh-TW.2", "細明體"); pref("font.name.sans-serif.zh-TW.2", "細明體"); pref("font.name.monospace.zh-TW.2", "細明體"); In this way, we will first try to use 新細明體, and then MingLiU (In English system) and them 細明體 Another question, should we map serif, sans-serif and monospace all the the same font? Can some one compose a list of font which available on Traditional Chinese System I would like the see something like the following List of font available on base installation of Traditional Chinese Win95 List of font available on base installation of Traditional Chinese Win98 List of font available on base installation of Traditional Chinese WinME List of font available on base installation of Traditional Chinese WinNT4 List of font available on base installation of Traditional Chinese Win2K List of font available on base installation of Traditional Chinese WinXP (optional) Suggest Serif font Suggest San-Serif font Suggest monospace font Suggest Fantacy font Suggest cursive font Cursive font should map to something like 草書 行書 should we map some font to 宋體 黑體?
Target Milestone: --- → mozilla0.9.3
Status: NEW → ASSIGNED
Target Milestone: mozilla0.9.3 → ---
Target Milestone: --- → mozilla0.9.4
I check the Win2K again, they have 3 type faces 標楷體 新細明體 細明體 I cannot tell which one is better than the other. In what sense 新細明體 is better than 細明體 ? Anyway, move to m0.9.5
Target Milestone: mozilla0.9.4 → mozilla0.9.5
move to 0.9.6
Target Milestone: mozilla0.9.5 → mozilla0.9.6
Need info, could anyone attach screen shots of available fonts (sample view by control panel), so we can compare them? Do IQA have Chinese Windows with those fonts installed?
Attached image MinLiu - old
Attached image PMiuLiu - new
I am not sure which is better. Anybody has prefference?
Do we really want to change it? Otherwise, I will move this to future.
Move to future.
Target Milestone: mozilla0.9.6 → Future
On build 2002051006, English version, on English Windows XP: Font size for Traditional Chinese is set to 16 pixels. This is too small compared to other Windows applications. A solution is to change it to 18 pixel by default.
remove future. Should we change to 18 on rtm?
Target Milestone: Future → ---
Reassign to shanjian, cc to rchen.
Assignee: nhotta → shanjian
Status: ASSIGNED → NEW
accepting.
Status: NEW → ASSIGNED
新細明體 is a lousy name, because it's not really "new" in anyway. The english font name PMingLiu describe it better. When you choose PMingLiu it comes with "proportional" english font. And when you choose 細明體 or MingLiu it comes with monospace english font. So it's clear we should choose 新細明體/PMingLiu first if possible. Of course in the monospace column, we should leave it as "細明體/MingLiu". And forget 標楷體 or any other fonts. In standard english win2k the only traditional chinese font we have is MingLiu.ttc(PMingLiu & MingLiu). By the way, why anyone says MingLiu 16 is too small? Actually MingLiu 16 is even a bit larger then serif 16, and IE default is also 16, I don't think it should change to 18.
Is any one working on this bug? I'd suggest this page: http://www.trigeminal.com/samples/font_choices.html After checking for several Chinese and non-Chinese windows configurations, I think we can change file in CVS: mozilla/modules/libpref/src/init/all.js from (line 910) pref("font.name.serif.zh-TW", "細明體"); // "MingLiU" pref("font.name.sans-serif.zh-TW", "細明體"); // "MingLiU" pref("font.name.monospace.zh-TW", "細明體"); // "MingLiU" pref("font.name-list.serif.zh-TW", "MingLiU, 細明體"); pref("font.name-list.sans-serif.zh-TW", "MingLiU, 細明體"); pref("font.name-list.monospace.zh-TW", "MingLiU, 細明體"); To pref("font.name.serif.zh-TW", "新細明體"); // "PMingLiU" pref("font.name.sans-serif.zh-TW", "新細明體"); // "PMingLiU" pref("font.name.monospace.zh-TW", "細明體"); // "MingLiU" pref("font.name-list.serif.zh-TW", "PMingLiU, 新細明體, MingLiU, 細明體"); pref("font.name-list.sans-serif.zh-TW", "PMingLiU, 新細明體, MingLiU, 細明體"); pref("font.name-list.monospace.zh-TW", "MingLiU, 細明體"); Because only MingLiU is included in most computer with the ability to display Chinese (that's why I list it in fallback of font name list) while PMingLiu is included in modern Windows (2000+).
smontagu/jshin: can you review the proposal in comment #20?
Wow, it's over 3 years old... Hung-te, thanks for your info. If TC Windows is like KO/JA window, I'm rather sure it comes with more TC fonts than non-TC Windows' have. Given that, I like to do better than what you suggested in comment #20. Could you give us the list of TC fonts that come with TC Windows and which generic fonts they correspond to (roughly)?
Assignee: shanjian → jshin
Status: ASSIGNED → NEW
OK. Basically, from Win3.1 to WinXP there're only 3 fonts appeared in standard TC Windows in history. Ref: http://www.microsoft.com/typography/fonts/winxp.htm EnglishName ChineseName MingLiU 細明體 (MonoSpace, like Courier) PMingLiU 新細明體 (Proportional, like Courier New) DFKai-SB 標楷體 (not good for general screen display without anti-alias) DFKai-SB is usually used for printing out. In Windows 3.1, there were only MingLiU. After Win 95(to XP), we have PMingLiU and MingLiU in all TC Windows. DFKai-SB is also available, but please do not use it for screen output. And if you install Traditional Chinese Font support by IE or Win2000/XP (to display Chinese on a non-TC Windows), you'll always have MingLiU and PMingLiU (sometimes you'll also get DFKai-SB) Some 3rd vendor party software to make non-TC Windows display Chinese will install "MingLiU"and "PMingLiU", while some old software have just "MingLiU". Are these information enough? In summary, MingLiU is the oldest font in TC. And PMingLiU is a modern proportional replacement. A standard TC Windows has no any other TC fonts installed. So please just use PMingLiU and MingLiU. I'll also include a screenshot of DFKai-SB later, then all fonts can be found in attachments.
DFKai-SB (aka. kaiu.ttf) screenshot. Do NOT use this for screen display. Because this is very ugly when displaying in small size, specially when without anti-alias or cleartype enabled.
(In reply to comment #22) > which generic fonts they correspond to (roughly)? serif: PMingLiU sans-serif: PMingLiU mono: MingLiU unfortunately, we don't have official sans-serif font in TC windows. All the 3 fonts listed above are actually serif (even the DFKai-SB is also serif). So you can use PMingLiu for both serif and sans-serif.
one more comment. In fact, I'm not sure about whether we should set a real Chinese font for that or not. When Chinese characters and English letters are displayed in same page, IE will use English font settings to render English words and Chinese font settings to render Chinese words. (The concept of "Fontset") However, Mozilla try to render both English and Chinese characters with same font -- that is, a Chinese font in our case. Many people consider such result "worse" than IE because English letters are rendered better with English fonts. And in that case they can keep real "sans" and "sans-serif" for English display within a page with both English and Chinese. So, another idea is to leave these values untouched (or same to English settings) and let Windows use its fontset ability to pick up correct fonts (in our proposed settings, all fonts -- serif type, sans-serif type, and mono -- are using MingLiU/PMingLiU for Chinese characters (these 2 fonts differ mostly to English letters. They do look same for Chinese characters) Even if I set fonts of TC to English fonts, my TC Windows still renders Chinese pages correctly without errors. (tested on Windows 2000 and XP). So what do you think?
Let me explain in more detail. As I described above, for Windows 95 to XP, we have only 2 good Chinese fonts: MingLiU (Chinese name: 細明體) and PMingLiU (Chinese name: 新細明體). Unfortunately, they are almost similiar. So we have only one set of Chinese font. Chinese fonts usually contain glyphs for both Chinese and English (or, ASCII). And for MingLiU and PMingLiU, their glyphs for Chinese are almost same (because there's no "proportional" Chinese). So the real difference of these 2 fonts are only in their English glyphs. MingLiU has a 'serif monospace' fontface while PMingLiU is 'serif proportional'. As you can see, we don't have any official Chinese font with 'sans-serif' in its English glyphs. (Other commercial fonts do have that, but none in standard Windows systems) So if we want to honor "proportional" and "monospace", we can set the fonts to: serif = PMingLiU sans-serif = PMingLiU monospace = MingLiU However, English glyphs in PMingLiU and MingLiU are not good enough in comparision to real English fonts like Times New Roman or Tahoma. I'll attach screenshot later. They really look twisted in small fontsize. So if we look at Internet Explorer, it tries to render English glyphs in the font set for English, and only Chinese glyphs used the Chinese settings. On the contrary, Mozilla will use English glyphs "in a Chinese font" (which is ugly) in a page with both Chinese and English at the same time. As a result, most people consider Internet Explorer as "more beautiful" than Mozilla. Setting Chinese font settings to PMingLiU only honors the "proportional" parts but still "ugly" to some people. The real solution is "fontset". However, Windows is already ready for fontsets. We can set "Chinese font" entry in Mozilla settings to an "English font", and Windows will render correct Chinese glyphs in Chinese fonts (which is appraently PMingLiU/MingLiU) and English glyphs in that "Chinese font" entry. That is, if I set serif font = "Times New Roman", in a page with both Chinese and English, Mozilla will render English letters by Times and Chinese glyphs still in (P)MingLiU. This makes the best output for users. I asked serveral people to try the new settings, and almost everyone I asked think this style is better. So, I think we can set Chinese font settings to "use" exactly the same settings of English fonts (Times, Arial, Courier). And will this cause any problems? I've tested in the following environments: (1) All Traditional Chinese Windows : from 95/98/2000/XP/2003: OK (2) English (or Non-TC) Windows with unicode + Chinese fonts installed: 2000/XP/2003: OK (3) English (or Non-TC) Windows WITHOUT unicode system + Chinese fonts/system: (eg, English Win98+ Twinbridge CJK system) Unfortunately, I couldn't find anyone with the last (3) environment. Can anyone with a system like that report will setting all your Chinese fonts to English fonts result in any problems?
(In reply to comment #26) I agree with LIN's opinion. You might just set the Chinese fonts as English, ex: serif -> Times New Roman, sans-serif -> Arial, etc. But I suggest set the monospace font to MingLIU (²Ó©úÅé), for keeping both Chinese and English displayed with the same width (1 Chinese character occupy 2 English letters space). If we set the monospace to Courier New, the Chinese will display in MingLIU and the English in Courier New, and the widths of these two fonts are different.
The left one is using serif="Times New roman" while the right one is serif="PMingLiU". As you can see, both of them can display Chinese correctly. However, English words (especially those marked by red rectangles) are better in leftside. The letters looks like glued too close.
(In reply to comment #28) > But I suggest set the monospace font to MingLIU (²Ó©úÅé), for keeping both > If we set the monospace to Courier New, the Chinese will display in MingLIU and > the English in Courier New, and the widths of these two fonts are different. Well, what I mentioned was "Courier", not "Courier New" :P However... You are right. To keep the character width correct, we need MingLiU. So let's fix the settings: serif="Times New Roman" (Or just follow whatever English used) sans-serif="Arial" (Or whatever English used) monospace="MingLiU" These settings should work!
Attached file testcase
simple testcase for showing the rendering result with different font families in various languages.
Thank you for your help and comments. What you guys wrote is a well-known problem (see bug 227815 comment #2 and bug 227815 comment #3). I don't like glyphs for Latin letters in Korean fonts, either (I'm a Korean). The Windows font-linking does some magic, but what you really have to test is to run mozilla on Win 2k/XP with the default locale set to non-TC and in presence of SC, Japanese and Korean fonts as well as TC fonts. In addition, let me ask you a question. What 'font-family' setting do TC home pages usually have in their stylesheets? In case of Korean pages, the majority of them have only Korean Windows core fonts (Gulim, Batang, Gulimche, Batangche, etc) listed so that I kept the Mozilla's default in line with them. _If_ most TC home pages have font-family set to 'PMingLiu and MingLiu', I guess we'd better let Mozilla do the same because individual users can always change them to whatever they like.
CC'ing Masayuki Nakano who is an active Japanese member here, might know how the CJK fonts should be mapped as default.
(In reply to comment #32) > In addition, let me ask you a question. What 'font-family' setting do TC home > pages usually have in their stylesheets? In case of Korean pages, the majority Because many designers know that most TC users have only MingLiU, they do not write any Chinese fonts in font family (except for those generated by Frontpage, converted by word, or newbie designers) Many people only write font-family just for overriding English fonts. However if they do specified font-family as (P)MingLiU, Firefox and IE will both display them in (P)MingLiU way, so that's not a problem. But for those pages without specifying font-families, most people want to see English letters rendered in English font. They "think" this is the correct result (maybe because IE can do that). So, I still think setting English fonts is better than keeping MingLiU.
testcase, rendered with FF1.0PR on TC Win XP, with TC font settings set to serif=times,sans-serif=arial,mono=mingliu
testcase rendered with FF0.9.3 on an English XP with default locale set to English and no any settings were changed. The Traditional Chinese font is set to serif=mingliu,sans-serif=mingliu,mono=mingliu (The old, or say current, settings of Firefox)
testcase rendered with FF0.9.3 on an English XP with default locale set to English and settings were changed as we proposed here. The Traditional Chinese font is set to serif="times new roman",sans-serif=arial,mono=mingliu (The new, or say proposed/not-commited-yet, settings of Firefox)
(In reply to comment #32) > The Windows font-linking does some magic, but what you really have to test is to > run mozilla on Win 2k/XP with the default locale set to non-TC and in presence > of SC, Japanese and Korean fonts as well as TC fonts. I've attached 3 screenshots for comparison. And according to the screenshots, I believe that the new settings has exactly same effect to non-TC Windows just like that to TC Windows.
Sorry, I don't read all discussion. Because I don't have enough time for reading it. I know a problem for "English font" + "Japanese font". See this page's screenshots. http://www.mozilla.gr.jp/standards/webtips/webtips0031.html (written in Japanese) http://www.mozilla.gr.jp/standards/webtips/screenshots/webtips0031-TestCase.png In this case, the symbols of "Circle" and "Disc" in "Helvetica" and "Arial" should not be used in the document written in Japanese. Because the height of symboles is same x-height. However, the height of Japanese characters are same font-size, not x-height. I think that a document should be rendered by single font. If a document rendered by some fonts, same problem may be reproduced.
(In reply to comment #39) > I know a problem for "English font" + "Japanese font". I've heard of this. But maybe it's not a real "bug" to some pages. It's not easy to decide which should be the "default" action. > In this case, the symbols of "Circle" and "Disc" in "Helvetica" and "Arial" > should not be used in the document written in Japanese. > I think that a document should be rendered by single font. > If a document rendered by some fonts, same problem may be reproduced. In my opinion, if a web designers want to have those symbols displayed in full width/height, he can always specify page font to a DBCS font. So what we need to do is to decide which should be the default action for most pages. Apparently, there're more pages with only Chinese/English than with special symbols. And they do look nicer with Engilsh fonts. And for Chinese, because the English glyph rendered by Chinese fonts is REALLY bad (hard to recognize in small size), most pages (usually caused by viewed with IE) are designed with thinking English type fonts as their default values. In my opinion, I really do wish this change can be made. I've seen so many reports complaining that pages are rendered "poorly" "ugly". The old settings - MingLiU - does not respect "serif" settings, bug the newer PMingLiU still results in so-called "ugly" output. Only setting to English fonts can change all of these.
Any more comments? Because new L10N policy disallows us to make changes in localized builds, we really hope this can be commited. In fact, some zh-HK users wants to have same changes, too
The question is which one to put at the top of the list, a Latin font or PMingLiU. It's surprising that even MS Windows comes with so poor TC fonts (when it comes to glyphs for Latin letters) that no Taiwanese wants to use it for rendering Latin letters. The situation is rather different from that of Japanese and Korean users. Some of them prefer Latin fonts (like myself), but most people want to use Japanese and Korean fonts even for Latin letters. Anyway, I've just tested the setting in this patch I'm uploading **in presence** of SC, J, and K fonts and our font-list mechanism turned out to be robust enough to avoid 'ransom note-style' rendering.
(In reply to comment #42) > patch (Latin font at the top of the list) Thanks, tha patch seems OK. > PMingLiU. It's surprising that even MS Windows comes with so poor TC fonts > (when it comes to glyphs for Latin letters) that no Taiwanese wants to use it > for rendering Latin letters. The situation is rather different from that of > Japanese and Korean users. Some of them prefer Latin fonts (like myself), but > most people want to use Japanese and Korean fonts even for Latin letters. I think that was mainly because (1) we don't have real sans-serif fonts while Japanese and Korean do. And (2) the propotional font (PMingLiU) has strange spacing so that using PMingLiU will render glyphs too close to each other, (3) TC windows has almost one type of usable Chinese font(MingLiu/PMingLiu) so most programs set font to English (let system pick up default Chinese font) and most people are already used to that. The English font in MingLiU is not so bad but it is not propotional (monospace) but changing to PMingLiU is somehow even worse in visual results. That's why we need to change to English fonts, and it helps Mozilla having similiar reneder results in comparison to other browsers.
This comment is in UTF-8. I propose similar settings for zh-HK users as they have basically the same issue and do some minor cleanup(reorder the english name fonts on the front, cleaning whitespaces, etc). -pref("font.name.serif.zh-TW", "細明體"); // "MingLiU" -pref("font.name.sans-serif.zh-TW", "細明體"); // "MingLiU" -pref("font.name.monospace.zh-TW", "細明體"); // "MingLiU" -pref("font.name-list.serif.zh-TW", "MingLiU, 細明體"); -pref("font.name-list.sans-serif.zh-TW", "MingLiU, 細明體"); -pref("font.name-list.monospace.zh-TW", "MingLiU, 細明體"); +// Per Taiwanese users' demand. They don't want to use TC fonts for +// rendering Latin letters. (bug 88579) +pref("font.name.serif.zh-TW", "Times New Roman"); +pref("font.name.sans-serif.zh-TW", "Arial"); +pref("font.name.monospace.zh-TW", "細明體"); // MingLiU +pref("font.name-list.serif.zh-TW", "PMingLiu, 新細明體, MingLiU, 細明體"); +pref("font.name-list.sans-serif.zh-TW", "PMingLiU, 新細明體, MingLiU, 細明體"); +pref("font.name-list.monospace.zh-TW", "MingLiU, 細明體"); // hkscsm3u.ttf (HKSCS-2001) : http://www.microsoft.com/hk/hkscs -pref("font.name.serif.zh-HK", "細明體_HKSCS"); -pref("font.name.sans-serif.zh-HK", "細明體_HKSCS"); -pref("font.name.monospace.zh-HK", "細明體_HKSCS"); -pref("font.name-list.serif.zh-HK", "MingLiu_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); -pref("font.name-list.sans-serif.zh-HK", "MingLiU_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); -pref("font.name-list.monospace.zh-HK", "MingLiU_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); +pref("font.name.serif.zh-HK", "Times New Roman"); +pref("font.name.sans-serif.zh-HK", "Arial"); +pref("font.name.monospace.zh-HK", "細明體_HKSCS"); +pref("font.name-list.serif.zh-HK", "PMingLiu, 新細明體, MingLiu_HKSCS, 細明體 _HKSCS, Ming(for ISO10646), MingLiU, 細明體"); +pref("font.name-list.sans-serif.zh-HK", "PMingLiU, 新細明體, MingLiu_HKSCS, 細 明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); +pref("font.name-list.monospace.zh-HK", "MingLiU_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體");
oops. This should be in UTF-8. Sorry for spam. -pref("font.name.serif.zh-TW", "細明體"); // "MingLiU" -pref("font.name.sans-serif.zh-TW", "細明體"); // "MingLiU" -pref("font.name.monospace.zh-TW", "細明體"); // "MingLiU" -pref("font.name-list.serif.zh-TW", "MingLiU, 細明體"); -pref("font.name-list.sans-serif.zh-TW", "MingLiU, 細明體"); -pref("font.name-list.monospace.zh-TW", "MingLiU, 細明體"); +// Per Taiwanese users' demand. They don't want to use TC fonts for +// rendering Latin letters. (bug 88579) +pref("font.name.serif.zh-TW", "Times New Roman"); +pref("font.name.sans-serif.zh-TW", "Arial"); +pref("font.name.monospace.zh-TW", "細明體"); // MingLiU +pref("font.name-list.serif.zh-TW", "PMingLiu, 新細明體, MingLiU, 細明體"); +pref("font.name-list.sans-serif.zh-TW", "PMingLiU, 新細明體, MingLiU, 細明體"); +pref("font.name-list.monospace.zh-TW", "MingLiU, 細明體"); // hkscsm3u.ttf (HKSCS-2001) : http://www.microsoft.com/hk/hkscs -pref("font.name.serif.zh-HK", "細明體_HKSCS"); -pref("font.name.sans-serif.zh-HK", "細明體_HKSCS"); -pref("font.name.monospace.zh-HK", "細明體_HKSCS"); -pref("font.name-list.serif.zh-HK", "MingLiu_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); -pref("font.name-list.sans-serif.zh-HK", "MingLiU_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); -pref("font.name-list.monospace.zh-HK", "MingLiU_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); +pref("font.name.serif.zh-HK", "Times New Roman"); +pref("font.name.sans-serif.zh-HK", "Arial"); +pref("font.name.monospace.zh-HK", "細明體_HKSCS"); +pref("font.name-list.serif.zh-HK", "PMingLiu, 新細明體, MingLiu_HKSCS, 細明體 _HKSCS, Ming(for ISO10646), MingLiU, 細明體"); +pref("font.name-list.sans-serif.zh-HK", "PMingLiU, 新細明體, MingLiu_HKSCS, 細 明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體"); +pref("font.name-list.monospace.zh-HK", "MingLiU_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體");
(In reply to comment #43) > oops. This should be in UTF-8. Sorry for spam. Next time, you want to upload a patch, please, attach it instead of adding as a comment. The following is not acceptable. I'd be glad to make changes if there's PMingLiU_HKSCSS that can be used in place of PMingLiU. However, apparently, there isn't. > +pref("font.name-list.serif.zh-HK", "PMingLiu, 新細明體, MingLiu_HKSCS, 細明體 > _HKSCS, Ming(for ISO10646), MingLiU, 細明體");
(In reply to comment #46) > (In reply to comment #43) > The following is not acceptable. I'd be glad to make changes if there's > PMingLiU_HKSCSS that can be used in place of PMingLiU. However, apparently, > there isn't. > > > +pref("font.name-list.serif.zh-HK", "PMingLiu, 新細明體, MingLiu_HKSCS, 細明體 > > _HKSCS, Ming(for ISO10646), MingLiU, 細明體"); Actually, once glyphs for Latin letters and numbers are taken from fonts like 'Times New Roman' and 'Arial', it doesn't matter whether glyphs for Chinese characters are taken from PMingLiU or MingLiU so that for zh-HK, we can use pref("font.name-list.serif.zh-HK", "MingLiu_HKSCS, 細明體_HKSCS, Ming(for ISO10646), MingLiU, 細明體");
Added the change for zh-HK as well.
Attachment #161488 - Attachment is obsolete: true
Comment on attachment 161643 [details] [diff] [review] patch 2 (changes fonts for zh-HK as well) rbs, Taiwanese and Hong Kong users want to make the default what is kinda optional for Korean and Japanese users (recall bug 227815).
Attachment #161643 - Flags: superreview?(bsmedberg)
Attachment #161643 - Flags: review?(rbs)
Comment on attachment 161643 [details] [diff] [review] patch 2 (changes fonts for zh-HK as well) r=rbs > Taiwanese and Hong Kong users I suggest you folks document our font name-list mechanism more to that audience. It is quite neat and robust for these situations (and it has a hidden Unicode-range detect on Win32), and emulates the fontset in a controllable way, and should save you from waiting 3 years for your requests...
Attachment #161643 - Flags: review?(rbs) → review+
(In reply to comment #50) > > Taiwanese and Hong Kong users > I suggest you folks document our font name-list mechanism more to that > audience. It is quite neat and robust for these situations (and it has a hidden > Unicode-range detect on Win32), and emulates the fontset in a controllable way, > and should save you from waiting 3 years for your requests... Sorry I don't understand... Does that mean we have to change the patch? Or this patch will be commited recently? The status is still NEW...
(In reply to comment #51) > Sorry I don't understand... Does that mean we have to change the patch? > Or this patch will be commited recently? The status is still NEW... Pls, rest assured that it'll get committed as it is when I get sr. What rbs meant is that you could have kinda solved the problem on your own without waiting for this bug to be fixed by documenting the feature more widely to those who prefer Latin fonts to TC fonts for rendering Latin letters. Of course, we're fixing this bug because to some people (probably a lot of people), tweaking something like this is rocket-science so that we have to make it work out of the box.
Status: NEW → ASSIGNED
Comment on attachment 161643 [details] [diff] [review] patch 2 (changes fonts for zh-HK as well) I'm not a superreviewer, nor especially competent at i18n font issues, but since jshin/smontagu appear to be ok with it, I'll go with it. r=bsmedberg (sr=rbs)
Attachment #161643 - Flags: superreview?(bsmedberg) → superreview+
sorry for the mistake and thanks for r. fix checked into the trunk
Status: ASSIGNED → RESOLVED
Closed: 21 years ago
Resolution: --- → FIXED
Attachment #161643 - Flags: approval-aviary?
Comment on attachment 161643 [details] [diff] [review] patch 2 (changes fonts for zh-HK as well) a=asa for aviary checkin. Is this something we want on the 1.7 branch too?
Attachment #161643 - Flags: approval-aviary? → approval-aviary+
Comment on attachment 161643 [details] [diff] [review] patch 2 (changes fonts for zh-HK as well) If this goes into aviary it should go into 1.7 as well. But I'm curious -- do all versions of Windows have "Times New Roman" and "Arial" fonts that contain Chinese glyphs or treat the names as fontsets, or is that only true on newer versions of Windows? If not, would it be better to use the other font as a fallback, second in the list?
Attachment #161643 - Flags: approval1.7.x+
(In reply to comment #56) > But I'm curious -- do all versions of Windows have "Times New Roman" and > "Arial" fonts that contain Chinese glyphs or They don't have Chinese glyphs at all in any version of Windows.Gfx:Win has its own controlled mechanism of font-sets. It looks for glyphs in fonts listed in font.name-list.* if they're not found in the font listed for font.name.*.For Chinese characters, the first font available on the system among the listed in 'font.name-list' will be used while for Latin letters, 'New Times ROman' or 'Arial' will be used. We're making this change because that's exactly what TC users want. Basically, they want a UA-style entry like this: .serif { font-family: Arial, PMingLiu, MingLiu, serif}
Checked in on aviary and 1.7
See Also: → 1946357
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: