Well, of course there were Zapf's fonts. I had these back in the mid 1990s, but don't think I kept them. I made up some of the needed characters in a different font for the University of New Mexico press to set a book. Things wound up where a U.N. representative looked the characters over & both said they were OK and pleaded for more fonts with the correct characters. But nobody ever took things further. I'm sure your work will be appreciated.

One thing I do remember was some wanted the underdot a dot, some wanted a bar. There was a compromise, where a bar with rounded corners was used. I think that got changed; I imagine the shapes of things are undergoing a certain standardization.

charles,

Thank you for the comment, you've given something to look in to.

Some of this information I'm looking for would be, for example, whether the Capital Latin Gamma (uni0194) should extend below the baseline (as I've seen it drawn both ways). I suppose the best method would be to go directly to the source.

You should maybe take a look (if you haven't already) to Nyala on Windows Vista/7. It's based on Sylfaen, a Microsoft multi-script font developed by Tiro Typeworks. If you want to learn a bit more about the project and design process, you could go here and download the PDF under Sylfaen.

Cheers,

JM Solé

Only the Ethiopic portion of Nyala is based on part of the original Sylfaen project: the Nyala Latin is a new design, and the Ethiopic was heavily reworked. Nayala's Latin subset targets languages of Ethiopia and Eritrea only.

Windows 7 ships with a pan-African UI font, Ebrima, for which I made the Latin subset based on the Segoe UI font. This might be a fairly good guide to basic letter shapes.

J.M Solé and John Hudson,

Thank you both! Sine I'm using a Windows 7 PC I'll definitely take a look at those.

Thanks for clarifying John. I remember reading the PDF but I didn't remember the specifics.

Cheers,

JM Solé

JamesT,

in order to define the language support for my own font, I did a research about alphabets using Latin script last year. I can send you some material about African languages, if you want. It is far from a comprehensive relation, but may prove useful.

Besides this, there are some good sources of information on the web:
PanAfrican
ScriptSource
Geonames
SIL International
Letter DB
Omniglot
Unicode Afrique (in French)

Some related Tipophile threads:
Eng and N with hook
K with hook and others

I also read Wikipedia pages for all documented African languages and also all pages for the base character letters, which link to unsuspect diacritical combinations. Note sometimes the English Wikipedia page for a given language may be not the more informative one.
.

John Hudson,

as usual, you kindly help us a lot with your deep knowledge. I was unaware about the pan-African support provided by Ebrima.

How could I get more info about the languages targeted by Ebrima? I saw the font have precomposed vowel+ogonek and vowel+cedilla to all vowels. I cannot find most of these combinations in the research I did and would like to know about the languages actually using them.

Besides this, some African combinations were not included in Ebrima, as Open O with dieresis (used in Dinka) or W with hook (used in Puguli). So I believe Ebrima development narrowed the language support according to some criteria to avoid the massive combinations one find in SIL fonts or African Serif.

Any info on this will be very welcome.

Igor, when I was researching African languages in the late 1990s, I spent a week at the African Studies library at Northwestern University. While there, I noted the instability of many orthographies, and inconsistency in the use of diacritics. Sometimes this inconsistency seemed to reflect either particular mechanisms of text production, i.e. what characters were available on a particular typewriter, and sometimes the influence of colonial languages or the native language of linguists/missionaries involved in alphabet development. So one of the things I noted was that nasalisation -- which is an optional feature of many African orthographies -- was sometimes indicated with a tilde above the vowel, sometimes with an ogonek, and sometimes with a cedilla. I believe the latter is really a variant of the ogonek resulting from the presence of the cedilla on French typewriters. I suspect as orthographies become stabilised, some of this variation is likely to disappear, but my recommendation to Microsoft for Ebrima was to support the variants I had documented in my earlier research for them.

The open o with dieresis should be supported in Ebrima with mark positioning for the combining dieresis character. Precomposed diacritics are only used in Ebrima if they were either encoded as such in Unicode, or involved merging of outline (as in the case of the letters with ogonek etc.). A huge number of additional diacritics are supportable with mark positioning.

The Puguli alphabet is pretty new, and I'm not sure it was even defined when I was conducting research on African languages. If they're using the w with hook, then that should be added to Ebrima.

John Hudson,

After studying Ebrima for a while, I've found the logic for the specific glyph inclusion very interesting and helpful in regards to directing my own research.

Regarding the Open O with dieresis (and the open E with dieresis), they do appear composed on my PC (both upper and lower case). The only issue is the Uppercase glyph uses the lowercase combining accent, resulting in a difficult-to-decode glyph.

I used to live right outside of Northwestern, if only I still had that resource nearby...

Precomposed diacritics are only used in Ebrima if they were either encoded as such in Unicode, or involved merging of outline (as in the case of the letters with ogonek etc.).

We generally follow the same principle in our SIL fonts. We used to include around 800 additional combinations primarily to provide better support for AAT, but we're removed those in the version that will come out next year.

As for the Capital Latin Gamma (uni0194), we choose to keep it on the baseline. That seems to be reasonably well-accepted. On an interesting historical note, the Practical Orthography for African Languages (POAL) from 1927/1930, which influenced many alphabets of West Africa, has it on the baseline, but with a flat bottom.

Gaultney,

Thank you for the response! That's very interesting about the Latin Gamma.

Would it make sense to have to have either stylistic alternates or localized forms with the different versions of the Capital Latin Gamma? I'm thinking of dealing with it in the same manner in which I'm handling the African Eng (which is based on the lowercase form of "N") and the Samí Eng (which is based on the Upper case form of "N").

John, thank you very much for the information. It's amazing how rich the study of alphabets could be.

Another question: in Ebrima the cedilla is linked to the first leg in H+cedilla and to the second one in A+ and uni0245+cedilla. I cannot find information about where the cedilla would be placed. Are there any known cultural or language preference on this issue?

Actually, the criteria used in Ebrima is very near the way I built my project. My early idea was to include all base+diacritic combination as precomposed glyphs, but after some research (and helpful threads in Typophile) this proved not good. So I added all Unicode precoposed glyphs plus all problematic combinations. By problematic I mean the situations where diacritic is merged with or touch the base letter, and also combinations using uni0332, as the underline must match the base width.

This gives an idea:

James, I sent you some material. Hope it helps.

Viktor, the source you did mention is the only I found about the Latin Capital Gamma while researching about African alphabets. Are there other examples of this glyph?

And, although the thread is about African characters, can you point any source about italic design of historic Cyrillic glyphs, like Yus or Ksi, or non-Slavic ones, like Kha? I'm trying to respect as much as possible historic characters and minority languages, but some Cyrillic letters are really hard to know about.

The Cyrillic set is going this way:

Hi,

From the data collected for the African Localization Network, I've seen nasalization of vowels indicated with cedilla, tilde above, tilde below or some nasal consonant after.

The cedilla is recommended in Cameroon since the 1970s/1980s so I doubt this will be replaced by the ogonek. In other countries some dictionaries use the cedilla to indicate some other feature (like advanced tongue root).

The tilde below are used at least in Sara-Ngambay in Tchad (I don't know if this was in the recent official alphabet, but at least it has been used for a few decades) or in some languages in Côte d'Ivoire (even if the -n is recommended there).

In Senegal, there's some nasalized consonant or semi-vowels indicated with a circumflex, but it's not generalized.