Hi guys,
I am having a tough time with understanding where the diacritic should be placed on some of the Extended Latin glyphs. Namely:
- B with dor below 1E02 - under stem or centered?
- b with dot below 1E03 - under bowl or centered?
- D with cedilla, d with cedilla 1E10, 1E11 - 1. where is the cedilla placed 2. should I supply alternative corresponding glyphs with comma? If yes, how should they be coded?
- D with caron below, d with caron below 1E12 1E13 - where is the caron placed?
- h with dot above 1E23 - is the dot over the stem like an /i or over the bowl like in /b with dot above?
- h with dieresis 1E27 - same question as the previous
- H with cedilla, h with cedilla - same question as p.3
- p with dot above - same question as 5.
- Welsch letters: 1EFA to 1EFF - are they archaic or not at all?
- Are there some pecularities I need to know about designing pan-nigerian letters?
If I had to speculate, I would reason that most of the diacritics below the base letter should be centered, because nobody writing them out would have the time to precisely place the diacritic under the bowl. But I can't be aware of all local preferences. I am using UniBook and visiting various websites to get the needed information but get different results. e.g. I know hcircumflex should have the sign over the stem, and see it misplaced to the right in several sources. So I do not trust everything unconditionally.
I would kindly ask that we keep spam to a minimum because of deadlines. Only comment if you have precise info. Thanks!
Comments
André
André
2. Optically centred.
3. Despite character name, these are not usually cedilla form, but instead comma below. Optically centred.
4. Optically centred.
5. Either above the stem or optically centred. Not lowered as in ḃ.
6. Ditto.
7. Most examples I've seen attach the cedilla to the bottom of the left stem.
8. Optically centred.
9. Mediaeval letters, not in current use except specialist scholarship.
In general, the Unicode glyph chart for this character range is a pretty good reference for default forms.
My general approach to whether to optically centre or anchor above/below a stem is that if a letter has a single stem ascending or descending then I favour anchoring on the stem, but if the letter has multiple stems ascending or descending or has a stem+bowl etc. at the mark alignment height then I optically centre. Obviously connecting signs like cedilla and ogonek are a different matter, and generally connect to a stem in preference to a bowl, although this may vary by language.
What's the deal with
uni0186, uni018C, 018E, is 0190 different then a resized epsilon, is 0192 different than a florin, is 0194 used, is 0195 used, ditto 0196, 019C, 019E, 019F (and is the 019F different than Theta), are 01B5-01BF used at all?
Are there pecularities like the elonagted dot bellow that are not seen in the unicode table? Are there glyphs that are positively never ever used?
Ditto for Yogh (021C, 021D)
Which languages use hcaron? Is the diacritic placed similary like in hcircumflex - does it need an .alt version where it is centered horizontally?
Thanks in advance!
0192 unfortunately is used for both the florin and the lowercase of 0191 which don't normally look the same. A font supporting African languages would need to have a localized form of 0192 and just hope that users don't plan on discussing prices in guilders. For African languages it should be based on lowercase f, whereas the florin symbol is typically slightly smaller and often based on an italic f.
Used by whom? All of these are used, though some are only used in historical contexts (e.g. wynn, hwair) The unicode charts generally provide some info on where characters are used.
019F is based on Latin uppercase O. Theta may or may not be the same depending on the font.
Again, you'll have to be more specific. Yogh takes a number of different forms, in some cases looking more ezh-like and in other cases more three-like, depending on the font. It is used in middle English, though some normalized texts will replace it with gh or y depending on context.
André
I was asking if most of these glyphs are used outside of historical contexts, like the two with the bar - it seemed to me these were earlier versions of proposed glyphs that have stayed in Unicode without being actually used.
Even with the best of intent I do not think it would be possible to cover all the glyphs necessary for African languages, since most of them, as far as I can gather, are still in development. I can find no extensive reliable information about localizations and exceptions. Me and my colleague agreed to leave the uni0234-024D from L. Extended-B range and the IPA Extensions unfilled.