GlyphWiki logo


出典: フリーグリフデータベース『グリフウィキ(GlyphWiki)』



アラビア文字の作字ありがとうございます。UCSグリフにコピーしました。これからも頑張ってください。--kesuuko 2019年3月19日(火) 14:48


你個「唔要」字好正 xd. 香港人乎? -- chanhenryfaihang 19:09, 22 April 2015

美國出世 (^_^;) --umbreon126 2015年4月22日(水) 19:19

I see... in Hong Kong now? The government is adjusting the Big5/HKSCS glyphs, but they still looks very bad in my view, so see if I can find some other people to help. ^_^ chanhenryfaihang 19:26, 22 April 2015

I'm still in the US, I'm sorry orz (and even if I was in Hong Kong I wouldn't be sure what to do...) No one ever said font design is easy. umbreon126 2015年4月22日(水) 19:32

原來係咁 :馬騮掩眼: 係D筆形同部件設計得好「特殊」,就好似口字右下角要穿出,同個立字一點一撇須觸上下橫,搞到教科書同報紙用開用咗三十幾年嘅字體(即係蒙納宋同華康儷宋)都變咗「唔符合香港標準」 @_@ chanhenryfaihang 19:43, 22 April 2015

聽起係好麻煩(;´д`) umbreon126 2015年4月23日(木) 11:44


Hello, sorry to interrupt, I saw you have reverted the design to u97cb-g, but you mentioned that "@7 is closer to the original G source shape". However, it seems that we are looking at different version of CJK Unified Ideograph Code-charts? As I remember I had redesigned that glyph while overlaying the editor over the code-chart (Unicode Version 7.0) myself. --chanhenryfaihang 14:25, 22 April 2015

The ヰ on the bottom seemed a bit wide :/ (and I'm also referencing Unicode 7.0) umbreon126 2015年4月22日(水) 14:31 (although I should have looked a bit closer at the vertical bar of ヰ)

stretch function

Hello, I would like to know how the stretch function works, but I cannot read Japanese and Google Translate is not very helpful. Could you mind to explain to me? For example, what does the number 136:0 and 0:-80:0 in "99:136:0:61:0:196:200:u20094-07-var-001:0:-80:0" stand for? Thank you in advance! chanhenryfaihang 14:10, 20 April 2015

I'm still not confident that I understand it fully but I'll try (´∀`;) (i am also very bad at explaining things)
-sandbox:xx:xx:xx corresponds to the "ストレッチ境界" numbers on sandbox's talk page. 99:xx:yy:- controls how much effect the stretch has.

xxx and yyy control from which point on the x/y axis stretch affects a glyph
foo is a direction (U, D, L, R) (up/down/left/right) or two directions
On u529b-02, the "stretch point"(?) is at 95,100, and R means that in the Glyph Editor, +1 compresses everything to the right of (95,100) to the right (and as a result the tail of 力 becomes longer)
On u529b, 100,60(D) leads to everything below (100,60) being compressed downwards when +1 is pressed in the Glyph Editor
Sometimes two direction instructions are used in enclosure glyphs like -05

  • If a glyph doesn't have stretch set and I don't want to edit the talk page, I look up what I want in Special:Search (for example, I need 140,100), choose a glyph and adjust it in the Glyph Editor, and then copy the instructions ("99:214:0:0:0:200:200:u4e45-var-001:0:40:0" becomes "99:214:0:0:0:200:200:sandbox:0:40:0") ( ´∀`)
    • I tried figuring out how to directly adjust the numbers once, but I got nowhere; then I realized that I could do this

  • sandbox:0:0:0 is equal to ストレッチ境界:100,100(foo) when adjusting 99:xx:yy

  • or something like that (if I missed something don't hesitate to tell me) umbreon126 2015年4月20日(月) 14:35

  • I see, makes much more sense now to me! Thank you! chanhenryfaihang 15:33, 20 April 2015


Re: the H-glyph in Unicode chart for Big-5 characters (those marked HB1-xxx, HB2-xxxx) is not actually from Hong Kong. Due to certain issues in including H-source, the HB1-xxxx and HB2-xxxx have used the Taiwan glyph erroneously, which does not follow the HK guideline/HK standard. If you inspect carefully, you can find the HK glyph and Taiwan Glyph are completely exact copies (same coordinates, same positioning).

The proper regional glyph for HB1-xxxx and HB2-xxxx is scheduled to be ready this year. Currently only H0-xxxx is following Hong Kong standard. Not sure if the H-glyph should use the proper glyph as outlined in the industry guideline or the existing erroneous glyph? chanhenryfaihang

As I understand it, the -h suffix is for the H glyph in the charts. If it is true that the current Unicode glyphs are erroneous, then perhaps it is best to discuss the issue with 利用者:kamichi or at GlyphWiki:井戸端. umbreon126 2015年2月4日(水) 16:18

weibo-eateateat weibo-buybuybuy

These two "joke" characters are widely spread on Chinese social network i did add them to this wiki, and i thought it's not so improper since there are characters of gundam and heavy rain etc.

Why did you delete them?

Some screenshots of notable origins are here:

  • Please use the prefix "sosaku-" for not a real Chinese charatcter. "sosaku-" means "joke" (not exactly the same meanings). So "sosaku-weibo-eateateat", "sosaku-weibo-buybuybuy" are the better names. --kamichi 2015年2月17日(火) 20:09
    • Thanks for explaining. I have moved them to sosaku-xxx. farter 2015年2月18日(水) 01:07

Glyph shape for u5df1-t

The glyph shape for the taiwanese regional glyph u5df1-t should have the second stroke protrude from the first stroke, e.g. u5df1-t@3. This is required by the MOE glyph guideline and also reflected in the Unicode Charts. Ditto for u5df1-h.chanhenryfaihang 02:52, 2 April 2015
  • Please see the message I left on your talk page. umbreon126 2015年4月2日(木) 02:53

Which kind of parts should be used for "IDS類推字形"?

In the group:古壮字一覧, for example, is it reasonable to use u8fb6 for character which includes this part instead of u8fb6-g?And I can't understand why some characters are put in "电子方块壮字(SAWN外字)の字形" and some are put in "IDS類推字形" with G variants. I wonder if my editing on sawn-f0543 and u2ff1-u98ce-u5fd8 is right. Thank you! --keepout2010 2015年4月13日(月) 14:38

Creating a new glyph at -var- or -itaiji- is encouraged over replacing the old one. I am not familiar with how the Sawndip pages are structured. umbreon126 2015年4月13日(月) 14:59

    • For encoded characters, e.g. グループ:古壮字一覧-URO収録分 the J-shape should be in the first column, and the G-shape, if available, should be put under Gソースの字形. The IDS類推字形 should be blank.

    • Please note that some of the lines in the encoded characters/unencoded characters may be incorrect, i.e. one glyph may contain extra components. In such a case, it is better to look up the original sources and correct them.