Edited by author 09-02-2001 03:28 AM
Introduction...
This is an experimental e-mail list/bulletin board for discussion of how to facilitate use of the complete Hausa alphabet (including upper and lower case special characters: "hooked" b, d, k & y) in computer & web software, and how to set up keyboards to enable people to easily compose text in Hausa.

Voici une liste de E-mail/tableau d'affichage expérimental pour discuter les sujets de comment faciliter l'utilisation de l'alphabet complet de Hausa (y compris les caractères spéciaux majuscules et minuscules: b, d, k et y "crochus") dans les logiciels d'ordinateur et du Web, et de comment faire installation des claviers pour permettre à des personnes de composer aisément le texte en langue Hausa.

Don Osborn
Bisharat.net

First questions ...
I am prompted to set up this discussion board/list by a question from Sambo Markus of Fantsuam Foundation about setting up a keyboard for Hausa. As I understand it, this is in the context of extending their IT for development efforts into rural areas in parts of Nigeria where Hausa is the main language.

Having some background in the use of specialized characters on computers for African languages (esp. Fulfulde and Bambara), and with recent experience working with Hausa in the Lucida Sans Unicode font in Word 2000 and on the "Unicode" (utf-8) Hausa pages at http://www.bisharat.net, I have also come up against the keyboard issue. It's not too much trouble for someone with some computer literacy to "Insert, Symbol" and choose the needed character from a chart in Word, or to insert the &#...; codes in HTML, but how to streamline the process? And more importantly, how to make novice-friendly interfaces for computing in langages like Hausa that use extended character sets?

For those using recent MS products, at least, the task is made somewhat easier by the fact they use Unicode (as I've been told) and have at least one extended Unicode font in the standard font reperatoire. Key macros can be set up, if not a "new" keyboard, to allow easier entering of specialized characters. But what about non-MS products (Macs, Linux, ...)? ...

What about alternatives like font packages from Linguists Software or Unitype? ...

And beyond that what about the possibility of standardization of keyboard setups so that every computer doesn't have a different way of entering, say, the implosive ("hooked") d?

Don Osborn
Bisharat.net
bisharat@kabissa.org

1. AfroRoman, available for Windows and Macintosh, solves the keyboard and font problems for Hausa and many other African languages. You can buy it at http://www.worldlanguage.com/

2. Macintosh was Unicode compliant before Windows, as was Linux, though Linux development is largely a matter of getting some Linux geek interested in doing it. Macintosh is more expensive than Windows, but for those who work with different languages and scripts it has definite advantages.

3. Unicode even has characters for the special signs of Hausa Ajami. Check out http://www.unicode.org/ for more information.

4. This discussion probably belongs on the Hausa mailing list, or at least should be mirrored there. In fact, a lot of ground related to this has already been covered there. Information is available at http://www.humnet.ucla.edu/humnet/aflang/H...es/mailingList.html or write me at <philips@cc.hirosaki-u.ac.jp> if you have problems.

5. Another important place for information on Hausa is the UCLA Hausa site at http://www.humnet.ucla.edu/humnet/aflang/Hausa/hausa.html.

6. Ultimately I think Hausa and other major African languages will have to be entered by Apple and Microsoft into their standard language kits. This will be a matter of the size of the market. If there is a demand for such languages, the software companies will supply it. Just remember that demand size is measured in currency units, not numbers of people using a language.

7. My recent book _Spurious Arabic_ may be of interest if you want to know more about the Romanization of Hausa and its use in colonial administration. It is available from the African Studies Center at the University of Wisconsin, Madison.

John Edward Philips
http://human.cc.hirosaki-u.ac.jp/philips/

Edited by author 09-03-2001 03:51 AM
John, Thank you for your detailed and helpful reply. Here are a few quick follow-ups:

1. I am aware of AfroRoman from http://www.linguistsoftware.com, and found Unitype's Global Writer 98 on the Worldlanguage site (both are on the Bisharat.net links page). There are two issues that such packages raise in my mind: One obviously is the cost, which while perhaps not exhorbitant in our budgets, would likely be an issue for a small project running a number of computers in telecenters.

The other issue is the proliferation of mutually incompatible software packages. For instance, there are at least 4 separate national language font systems for Windows locally here in Niamey alone (INDRAP 98, Nigerienne, SIL Doulos, & ADD), not counting another one for Macs (PanKwa). Hence the interest in Unicode compliant fonts.* Are the AfroRoman and Global Writer fonts Unicode compliant? (I found no indication either way.) If not, it might not be so "sustainable," as we still say in development, to introduce them into an already confusing IT environment.

2. Macs have had many advantages (I recall a Pular instructional text that Sonja Fagerberg-Diallo did on Mac in 1985-6, with the "hooked" letter glyphs and all). But as I understand it they supply no Unicode fonts, as Windows does now. In any event, and for better or worse, I don't think Macs are an option for IT & development projects in Africa (I'd be interested to learn otherwise).

3. Interesting about the Ajami. I imagine eventually someone will do a web site in that script -- perhaps permitting to "toggle" between the Romanized and Ajami orthographies (somewhat more challenging than for what one can do with simplified & traditional Chinese). For the moment, though, the concerns on the ground here are much more basic and focused on use of the Romanized orthography.

4. Thanks for the info on the Hausa mailing list, which indeed would be a logical place to ask a lot of these questions. I did try subscribing a while back probably to the one you refer to (at a .jp address) but had a response from the listserve that I interpreted, incorrectly as it turns out, to mean the list was not functioning. Please feel free to mirror or summarize this discussion there. I'll probably do the same in the direction of Afrik-IT.

5. The UCLA site is indeed an excellent resource.

6. You're probably right about the markets, but from where we are now with Word 2000, I can already create a document in full "proper" Hausa, Fulfulde/Pular, Zarma, Bambara, etc. orthography using Lucida Sans Unicode (supplied with the software) or Arial Unicode MS (the latter I had to download specially). Even Times Roman gives me the "eng" letter (not a concern in Hausa, of course, but it makes that font sufficient for typing, say, Wolof). All this without buying anything new. However I'd have to set up my own set of key macros to avoid having to use "Insert" every time I needed a special character.

For the sake of users, esp. novices in the matter, it would be very helpful to have a standardized keyboard setup. If Unicode indeed manages to get us past incompatible coding issues, we're still left with a limited number of key combinations used for various different things across platforms and software, that now have to accommodate special characters. What can an IT for development project do now?

7. Thanks for the info on your new book, which sounds interesting. I dabbled a bit with Ajami transcription of Pular when in middle Guinea, which continues to be used there.

Don Osborn
http://www.bisharat.net

*(see for example http://www.issco.unige.ch/staff/andrei/formRIFAL2000/)

After further checking, the Unitype products (Global Office & Writer) supposedly are Unicode compliant and the Linguist's Software fonts (incl. AfroRoman) apparently aren't.

"Alan Wood's Unicode Resources" at http://www.hclrss.demon.co.uk/unicode/index.html has a lot more info on Unicode fonts and software that supports Unicode (not limited to Windows!). When looking at his font list, any font including both "Latin extended-B" and "IPA extensions" will have the necessary special characters for Hausa.

Of course this doesn't address the keyboard set-up (or the broader issues of localization that will eventually be of concern), but some of the Unicode supporting software he lists might. However these unlike the fonts will likely run into those cost issues...

Don Osborn
Bisharat.net

Thanks for the information, especially about Alan Wood's Unicode page. It will be very useful for me, especially since I am running Windows, Mac and Linux computers in several languages.

For ordinary e-mail I have no trouble with using apostrophes in place of hooked letters, and I haven't had to publish anything in Hausa since I produced the camera ready copy for my book. Since then I have had trouble setting up AfroRoman on my W2K computer, which still does not support Arabic. (Being based in Japan I find it easier to use the Japanese operating system.) I may just use Mac next time, since I have Word2000 on that machine, but only Word97 on Windows.

Finding a Unicode font for Mac should not be hard, even if you don't install Word2000 for Mac. The problem is that the Macintosh language kit does not install many African languages, although there is a Unicode setting. Of course Macs are more expensive than Windows, and many university computers in Nigeria use Linux, since that is even cheaper.

I have a free Arabic word processor on my Linux machines, (you can find it by going to the "Linux Arabic" link in the computing section of my links page.) but I have not tried to do Hausa on them yet. You may inspire me to download the necessary fonts and try it.

You may have to get a grant to do some software development and develop fonts and keyboard maps for various operating systems. For better or worse the computing environment not only in Africa but even in the developed world will be confusing for some time to come. There will continue to be three basic operating systems, and users will continue to need to share information back and forth across them. That is why standards like html, java, TCP/IP and Unicode are so important.

http://www.tucows.com has a lot of download information that may be useful trying set up fonts.

Edited by author 09-04-2001 02:56 PM
What do other readers have to say on all this so far?

Should a rural IT & development project simply use the "substitute solution" for Hausa language -- that is use apostrophes with b, d, k, and y for the implosives ("hooked" letters) until something cheap and standardized comes along? Could people new to computers and of moderate literacy skills adapt to using such a temporary arrangement when printed material and an increasing number of web pages use the hooked letters?

Does anyone have experience setting up a Hausa keyboard for use of the resident Unicode fonts in Word 2000 or 2002 or another widely available software?

What are/should be the roles of commercial developers of international software and national applied linguistics agencies in this matter?

And what about interfaces with "pop-up" virtual keyboards, liek the multilingual e-mail service at http://www.africast.com ? (Which, BTW, supposedly can handle the special Hausa characters.)

Don Osborn
Bisharat.net

What about Hausa localization (or pan-West African localization) with Linux? There's apparently a significant effort on such for Indian languages - see http://www.indlinux.org/about.php

Don Osborn
Bisharat.Net

In the specific area of web design in Hausa there is a way other than Unicode to get the special characters to display. That involves creating a modified standard font to be driven by a special script. For a very small example see the "hooked b" in the word "ha'be" and "hooked d" in "Ha'dejia" at http://www.kanoonline.com (I add the apostrophes as QuickTopic won't support the hooked letters). Apparently the glyphs for "~" and "]" have been modified in the font. See reveal codes for the script.*

As I understand it, this method works with recent MS Internet Explorer browsers -- perhaps the same versions that support Unicode? -- but has limitations. (Thanks to Bartek Plichta of Michigan State Univ. for clarifying some of this for me.)

Of course this solution, interesting as it is, is not intended to address anything beyond web browsing in Hausa.

Are there any similarly creative approaches that could be used for computer access in Hausa (font & keyboard solutions) by people just getting to know the technology. Any chance of any kind of standardization?

Don Osborn
Bisharat.Net

* (See http://www.fakoli.net/Nkobooks.htm for an example of this approach with a completely different alphabet, the N'Ko script.)

I have just joined the bandwagon, thanks to Don's invitation. In Kano (Nigeria, the "heart of Hausaland"?!?) we have been using a font I narcisstically call Dr. Abdalla Uba Adamu, which contains the hooked Hausa characters of b, d, k. This uses Times New Roman as the base font, and using Fontographer, I just bent bits and pieces on the necessary charactes to create my Hausa fonts. The main motivation was to get a better form of expressing my Hausa-language musings. But then Business Centers (the main publishing phenomenon in Northern Nigeria) got interested as the Hausa people embraced PCs and started dusting off their old manuscripts and converting them to MS Word. I decided to make my font public domain (obtainable at http://www.kanoonline.com - just look around, you will see it!!).

This of course do not display well in Internet Explorer, but the font is more than sufficient for accented writing of Hausa language using the PC. As it is, the only way Hausa accentuation can be standardized would be if it rides piggback on a standardizded IE language such as Arabic, when it transforms itself into AJAMI - the Hausa script written in Arabic alphabets. John Phillips has done a lot of work on the development of Ajami as Hausa literary script, but not much was done in Hausanizing the PC. Again some years back we developed a Hausa word processor called Marubuciya (The Writer), complete with Hausa menu entries; but lack of encouragement, or support, and time contraints made it difficult to develop beyond the openining screen!

Edited by author 09-09-2001 01:05 PM
Hi! It's me again. I saved the QuickTopic page and chewed it at my leisure! Also followed the some of the links to Bishira and it's all kinda groovy.

I am not sure about the nature of the debates (if any!) on this Board and its relation to the Hausa language. From the little that I see, it would appear to be on ICT and Hausa language development. This is hunky dory!

To contribute to the lot, I've founded a Yahoo! Groups on Contemporary Hausa Novel. I have sent an invitation to the bishira webmaster and I am extending it to all who visit this Board and are interested in the debates about the development of the Hausa novel form. The group's url is

littattafan_Hausa_na_zamani-subscribe@yahoogroups.com

On the Files section are a series of papers I have written and published in Nigeria concerning the role of the Hausa Novel in social discourse. In particular, I have written about the relationship betweeen ICT and Hausa writing, giving detailed mechanism of how I developed the Hausa font character sets. The URL is as below, although you need to be a member to access it:

http://groups.yahoo.com/group/littattafan_...ovel%20and%20IT.doc

If you don't want to be bothered to join the group, you can email me at auadamu@yahoo.com for a copy of the paper.

Let's rock!!!

Abdalla

Abdalla, Thank you for your contribution which is definitely helpful and most welcome.

Since you ask, this discussion board was set up to get experience and ideas such as you and John Philips offer on ICT & Hausa. But where the group you have started focuses on the interesting and important area of Hausa literature and language development, the focus on this board is more the technical aspects of fonts and keyboards for Hausa (with a particular concern for applications in ICT for development). These two share some concerns, of course, and should be seen as complementary.*

One of the questions motivating creation of this discussion space is: As computers are made more available and internet connections multiply in areas like those in Nigeria and Niger where Hausa is the predominate language, how to provide for intercompatible fonts and standardized user interfaces (i.e., predictable & user-friendly keyboard setups)?

Up until now, individuals and agencies have generally figured out their own ways of adding special characters (in Hausa the "hooked" letters) to fonts, and then how to add key-macros or whatever to the keyboards to facilitate use of the fonts. As a result there are a lot of mutually incompatible fonts for the same languages. As long as people don't share digitized documents much, and the focus is on individual work or production of printed documents this is fine. But what happens when rural telecenters, cybercafés, and IT for development projects (like Fantsuam) try to provide font + keyboard setups for a broader number of users? And how useful is it to train anyone in use of Hausa on computers when most other computers will not be set up the same?

Some people who have a lot of experience with computers and African languages have suggested that there will never be a full intercompatibility or standardization. Not even with Unicode, which is proposed as a standard for coding of all characters. In fact, one representative of a commercial software company I recently corresponded with suggested that even Unicode could not address the character issues of all African languages.

Still the problem is there and it will only become more pressing as time goes on. Approaching the issue from the point of view of a single language -- Hausa -- which is one of the most spoken on the continent, may be more manageable. What kind of workable approaches are there to standardization of fonts and keyboards for Hausa?

Don Osborn
Bisharat.Net

* (One would be remiss not to mention again the Hausa language list mentioned by John in in message #3: http://www.humnet.ucla.edu/humnet/aflang/H...es/mailingList.html )

Thanks for all these links. I will check them out when I have time.

BTW, N'Ko was not the first African script to be encoded. The ancient Meroitic script of Kush (originally a modified hieroglyphics used as an alphabet which then evolved into a different script) is in Unicode. I don't know of any fonts for it, though.

In the aftermath of the abominable hijackings & destruction in New York, Washington, and western Pennsylvania, it is hard to proceed as if nothing happened. Permit me, then, to offer condolences to any of the kind readers of this forum who had friends or loved ones among the victims of yesterday's tragic events.

At the same time it is essential not to be deterred by such crimes and loss of life from the constructive work of building a more just, unified, peaceful, and indeed righteous world (to the extent our humble efforts can aspire to contribute to same). With that intention I'll try to pick up the thread as left off by Abdalla and John.

Thank you, Abdalla, for the links you provide. I will look into adding at least some of them to the Bisharat.net links page (http://www.bisharat.net/links.htm). There are a lot of diverse resources for doing multilingual pages, but I haven't looked into the font design programs that much yet.

The approach you took, Abdalla, to creating a useable font is a tried and true one. The only drawback is that since many people have done it, there are a number of modified font/keyboard solutions around. For web presentation, the code script that your colleague, Salisu Danyaro, used on Kanoonlone.com is an effective way of "driving" the font from the server, but only if the person viewing the page is using a recent MSIE version.

Another approach is the one we took on the Hausa pages at Bisharat.net, and that is to use Unicode. A Unicode font that already has the extensions (groups of characters) that include the special characters used in Hausa saves one having to design those special characters and is viewable on recent MSIE and Netscape browsers (for Windows) if you have the same font.

I believe all MS Windows 98 (& after) packages come with the "Lucida Sans Unicode" font, which has been around since 1993 (copyright Bigelow and Holmes). This is what we used at Bisharat.net's Hausa Unicode pages. If you don't have it, it can be downloaded from numerous sites including the following (and also the RIFAL2000 site I mention in an earlier posting on this board):
http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm .

You are right about SIL's fonts. They've done a lot of good work in this area. One reason that I haven't pursued their fonts or others that people or companies have worked on is that they are not mutually compatible. I had the experience in Bamako over a year ago of hassling with a non-standard font for Bambara on different computers and printers and that really got me interested in problems of standardization. So far as I am aware, Unicode offers the only hope of anything standardized.

Regarding African scripts, in addition to what John mentions there is also, of course, Ge'ez (aka Ethiopic, used in Ethiopia and Eritrea) which has south Arabian origins, but is uniquely rooted in the Horn of Africa. And Vai around what is now Sierra Leone. Not to mention Tifinagh used by Tuaregs & Berbers. There's a link to more info on N'Ko at Bisharat.net.

Don Osborn
Bisharat.net

For many people, e-mail is much more accessible and important than the web. A multilingual web-based email such as that at http://www.africast.com, therefore may not help them much. Recently I've been trying to see how much I can do with MS Outlook Express to send Unicode pages (with special characterrs like the hooked letters). This depends on the recipient having the same font or another Unicode font with the same character extensions. So the results are mixed.

Another item of possible interest is the African Multlingual Internet Names Consortium page and e-mail list (see http://www.afriminc.gh). The site is apparently new. The list will presumably deal with language and internet issues.

Don Osborn
Bisharat.Net

Edited by author 10-06-2001 02:58 AM
The title of this board is plural for a reason. Aside from the extended Latin character set used for Hausa (including several "hooked" letters to represent "implosive" consonants not found in major European languages), there is an Arabic character-based system of writing known as Ajami. Does anyone have experience working with Ajami (in Hausa, or in other languages of the Sahel) on computers or the internet?

The two writing systems do not necessarily present an "either-or" choice in the new technologies. In the future, one imagines it should be possible to develop a way to "toggle" between an extended-Latin and an Ajami version of a text in a wordprocessor or a web page, much the same as is possible between simplified and traditional Chinese characters.

Don Osborn
Bisharat.Net

Edited by author 10-19-2001 11:06 AM
do you know of any 'instant translator' internet sites that can translate all or some of the African languages in South Africa?

Edited by author 10-19-2001 12:30 PM
Pinksoftware http://www.pinksoftware.com/trans has an online demo of its translation software for Afrikaans, English, Swahili, Tswana, Zulu; Xhosa on the Web! http://mokennon.albion.edu/ is a simple on-line translator to and from English. Check out section 5 of the Bisharat.net annotated links - http://www.bisharat.net/links.htm - for other languages and links to sites that keep track of on-line translators.

There aren't yet many translation programs - either software or on-line - for African languages, and they are not as sophisticated as those available for the main European languages. There is nothing for Hausa.

There is an effort to create a translator for the Fula language that may be of interest on this list. At this early stage it involves both the Pulaar of northern Senegal and the Fulfulde of central Mali. What is interesting for Hausa is that the display test pages can display the two or three special characters (in Unicode) which that langage has in common with Hausa (hooked b & d, and in Niger also hooked y).

On a side note, it is interesting that Xhosa apparently has the same "implosive b" sound as one finds in both Hausa and Fula. If I understand correctly, Xhosa uses "bh" to write that, which is the same system long used for Fula (Pular) in Guinea. There is a move to standardize the transcription for this sound in West Africa using the "hooked b."

Don Osborn
Bisharat.Net

Edited by author 10-30-2001 12:46 PM
Looking again for a moment at the situation of language and the new ICTs in southern Africa, it is worth noting a significant "open source" project to localize software in South African languages at http://www.translate.org.za/ (as reported in the Kabissa-Fahamu-SangoNet n-l #35). Is anyone aware of similar efforts to translate or create software in Hausa or other languages of west Africa? This is really the step beyond getting standardized fonts and keyboard conventions, but it's important to look ahead (esp. if someone is already there!).

Also regarding this South African localization effort, it would be interesting to know what kind of collaboration there is with agencies and specialists in neighboring countries, for those languages that cross boundaries, and whether such might serve as useful models for other regions.

Don Osborn
Bisharat.Net

Edited by author 11-19-2001 02:24 PM
A useful new web resource for Hausa is the Hausa Database On-Line Dictionary (Hausa to and from English and German) at http://www.univie.ac.at/afrikanistik/oracle/KofarHausaE.html . However they substitute b' d' k' for the hooked letters in the standard orthography. (For those unfamiliar with the language, these represent "implosive consonants" which Hausa has in addition to the ordinary b, d, k, and the difference can be very important for meaning, aside from sounding right to the listener).

As previously mentioned (below /m19 ), a very young and small project for an automatic translation program for Fulfulde/Pulaar (which shares some of the same sounds and letters as Hausa) has used the full standard orthography on test pages. This is done with a Unicode font.

If you want to see a really sophisticated interface for an on-line dictionary of a language that uses a non-Latin character set, see the Inuktitut Living Dictionary at http://www.livingdictionary.com (Inuktitut is a language of northern Canada). They allow users to choose to use the Unicode font or one of several others.

It should be possible to do something similar for a language like Hausa that uses an amended Latin alphabet. The problems for display of special characters in an online dictionary or translator are not that insurmountable (indeed, as far as simple text weblishing goes, one will have less and less excuse for not using the full standard orthography as time goes on).

The greater problem is input of the characters. IOW, how do you type in a hooked-b? In order enter specialized characters in MS Word, I have to go to the top menu to insert or set up a key macro. To display the special characters on the Hausa pages at http://www.bisharat.net above I had to insert the decimal codes in the HTML. But the first won't work in the browser and the second is not practical for the user of an online dictionary or translator.

One possibile solution to the input problem, and this requires the ear of people working on software: design a keyboard setting somewhat similar to the "international English" keyboard in Windows, for African languages. In the "international English" setting, you type apostrophe and e and end up with é. Could a standardized African setting ("international English for Africa"?) allow you to use such 2 key sequences to get special characters such as the hooked letters? This gets back to the original question from Fantsuam (moving to http://www.fantsuam.com ?)about user-friendly Hausa interfaces for computers in Hausaphone areas that was the impetus for starting this message board.

Don Osborn
Bisharat.Net

Edited by author 11-19-2001 02:52 PM
A project in southwestern Nigeria that shares a common interest with Bisharat.net in agriculture and ICT in grassroots rural development, the Committee for African Welfare and Development (CAWD) -- http://www.cawd.info -- is apparently looking at use of computers with touch screens (remember them? I had thought they were out or on the way out). Pamela McLean of CAWD has suggested keyboards on the touchscreens that could be modified for any set or combination of characters. Could such a "soft keyboard" (at a more user-friendly angle than on the usual video monitor) be an innovation especially friendly to use of languages like Hausa and Yoruba in ICT and to multilingual computer/internet use in general? If such applications were to use Unicode (for compatibility etc. with other systems) then this could have really interesting implications for computer use in poorer countries.

Don Osborn
Bisharat.Net

The discussion so far seems to be fascinating.

a number of issues have arisen, and i thought i'd throw my two cents worth in.

these days my prefered font solution is to use unicode fonts. Diusplay of unicode web pages requires an appropraite font. There are a growing number of fonts available that contain the extended latin characters. Quite a few come to mind.

It becomes more complex if you wish to display diacritics to indicate vowel length or tone which some academic texts and dictionaries do. Depending on the orthography used, it may require complex script support for the latin script. Microosft, for instance hasn't implemented this support yet. So a couple of projects i'm working on are on hold.

With respect to input, there are utilities such as Tavultesoft keyman (http://www.tavultesoft.com/keyman/) which can be used to design keyboard layouts which allow oyu to type in unicode characters. There are a range of 8-bit font solutions available for Hausa, is it the same problem for keyboard layouts?

Given a standardised keyboard layout it would be possible to generate a keyboard file for keyman. I should be able to throw one together fairly quickly. I'm currently using keyman for a number of languages i am working with.

The main limitations with keyman is that it requires that the applications have appopriate support for unicode. supported programs and OS are listed at http://www.tavultesoft.com/keyman/support/FAQ002.php

A self contained unicode solution maybe a text editor called SC UniPad (http://www.unipad.org/main/). it operates on windows platforms and has its own internal font. The text editor also allows you to create custon keyboard files, and includes an onscreen virtual keyboard.

Andj
andrewc@ozemail.com.au

Edited by author 01-15-2002 02:21 PM
Thanks, Andj, for the info and pointers, which are most helpful. I will incorporate the references on the Bisharat.net links page.

A few quick replies:
1) It turns out that recent Netscape and MSIE in Windows have (from all I've seen and heard) at least one font that permits display of extended chararacter sets used in Hausa (and by extension - no pun intended - many other West African languages).

2) What are the several fonts you have in mind? Esp. if they are freely available and have all the (Latin-based) extensions, it would be helpful to publicize how to obtain them. On recent Microsoft products (which I'm not trying to sell, but this is what I have personally and it is of course widespread) several fonts have some of the extensions (notably Latin Extended-A and sometimes -B), but only one, Lucida Sans Unicode, has the Latin Extended-B and IPA Extensions that have the special characters necessary for Hausa (add Latin Extended-A, also in this font, and you can have most West African languages in their current orthographies). Arial Unicode MS is another option, but it has much more than one needs. Any other font info would be greatly appreciated.

3) The matter of diacritics as you point out is a potential problem, but for most purposes in languages that use extended Latin character sets, diacritics aren't necessary. It won't help your projects, unfortunately (are you publishing?), but for creation of web content or word processing basic texts (at least on PCs with Windows) there's no reason not to go ahead and use the proper orthographies for text in languages like Hausa (the problem for web content is the necessity of providing an alternate version for Macs and older browsers - as we did on http://www.bisharat.net).

4) The option to create keyboards as in the Tavultesoft software can be very useful. In the short run it may be sufficient (for organizations like Fantsuam Foundation* or cyber-cafés), as a cheaper and easier measure for West African languages, to add some key macros to a system that has a Unicode font. MS Windows software will let you do that (for instance I've set up left Alt+b to give me the "hooked-b" and so on). The longer term issue of some sort of standardization of keyboards - so that someone who learns to type Hausa on one computer won't have to learn all over again on another - is one that I don't think anyone is addressing (please correct...).

5) I will try the SC Unipad, but have noted with concern the probable limitations on clipboard use mentioned on the website, which could defeat its utility for some applications. What has been your experience?

May I ask what languages you're working in?

Don Osborn
Bisharat.net

* Fantsuam has a webpage at http://www.fantsuam.com . It was a question from them that was the impetus for starting this discussion board.

Hi Don,

wrt 1) and 2)

iu have a number of fonts I use on a windows 98 machine that would support Hausa. There are a few ofnts that i'm still investigating.

Arial Unicode MS : [Commercial] available in a range of Microsoft products.

Caslon : http://bibliofile.mc.duke.edu/gww/fonts/Caslon/Caslon.html

Code2000 : http://home.att.net/~jameskass/CODE2000.ZIP

Lucinda Sans Unicode : [Commercial] available in a range of Microsoft products.

Monospace : http://bibliofile.mc.duke.edu/gww/fonts/Monospace/index.html

SImPL : http://www.vector.org.uk/resource/simplttf.zip

Thryomanes : http://www.io.com/~hmiller/lang/

TITUS Cyberbit Basic : http://titus.fkidg1.uni-frankfurt.de/unicode/tituut.asp [available for non-commercial purposes only]

I'm currently searching for otehr suitable fonts. Arial Unicpode MS and Lucinda Sans UNicode are hte best fonts for screen display, the are the only fonts in the list that possess high quality hinting.

I'm in the process of creating a set of unicode fonts for a couple of Sudanese languages. It would be simple to add the additional characters for Hausa. The only problem will be that these fonts will have inadequate hinting, since i'm not a professional typographer.

3) diacritic uses differs form langauge to langauge. In most cases they are not used in normal writing or printing. But they are useful in dictionaries, glossaries and sme educational materials. But teh reality is that until micrsoft support the Latin script in tehir complex script rendering engine (uniscribe) it is not feasible to add diacrtics to the unicode fonts.

4) i've put together two simple keyboard files for Keymam (Hausa 1 and Hausa2). The file should install keyman and install the two keyboard files.

they're at http://home.vicnet.net.au/~agamlong/files/hausa.exe

the files are just quick keyboard layouts i threw together.

Hausa 1 uses the control-alt and shift-control-alt key sequences to insert the extended latin letters required by Hausa.

eg to get

U+0253 LATIN SMALL LETTER B WITH HOOK type Control-Alt-B
0181 LATIN CAPITAL LETTER B WITH HOOK type Shift-Control-Alt-B

and so on.

Hausa 2 uses deadkey key sequences to insert the extended latin letters required by Hausa. The deadkey is the "/" key.

e.g. to get

U+0253 LATIN SMALL LETTER B WITH HOOK type / + B

0181 LATIN CAPITAL LETTER B WITH HOOK type / + Shift-B

and so on.

I've laso included a way of typing U+02BC MODIFIER LETTER APOSTROPHE. In ASCII dociuments i've noticed that an apostrophe is some times used in transliterated Hausa to indicate glottalization and the hooked letters. In unicode this would most likely be U+0181 rather than teh other instances of apostrophes, ie a different unicode character.

it can be type by / + ' in Hausa 2
and Contol-Alt-' in Hausa 1

One other thing, keyman 5 will work with appropraite unicode fonts and MSN messenger to applow people to communicate via instant messaging in Hausa. Not sure if this would be of interest or not.

5) yes , tehre are limits with SC Unipad, but i find it a useful tool at times, since its easy to cut and paste between it and other unicode aware programs. The main feature i like in it is the ability to convert a unicode text document into HTML numerical character reference formats and back again.

two other text ediotrs that may be more suitable are emeditor and ultraedit, both support unicode and accept input from keyman 5.

What languages am i working in? The focus of my position assisting public libraries, community organisation and NGOs in Victora, Australia to provide multilingual computing resources and internet resources and web page development to/for our immigrant communities and by extension across their Diaspora. Potentially we're working with any language. Most of the African languages we're working with at the moment are form the Horn of Africa and Sudan. But with the growing number of immigrants from other parts of Africa, we're slowly researching other African languages.

If that makes sense.

Andj.

Thank you Andj, for these additional references (and pardon the slow reply). At this point it may be worth setting up a page of links and instructions on Unicode fonts and keyboards for African languages (unless you already have one or are aware of one).

The question of what characters to add is tricky if one isn't going to put in all of Latin extended A&B and IPA. Four extra letters would work for Hausa of Nigeria and Niger. A total of four to five other modified letters will give one perhaps all the languages across the Sahel with the exceptions of Kanuri and Tamashek (Latin transcription), and possibly also More. A few more and one will have those plus most coastal West African languages with the notable exception of Yoruba (unless diacritics are used).

That said, it may suffice for just the Hausa characters. The "hinting" issue you mention could be a major limitation to wider use, so it might not be worth putting too much of your time into it at this point.

Re the keyboards, this is helpful. I'm curious, though, if your keyboard will let you simply add keymacros so that left Alt + b = hooked b, etc. This works for me in Word products in Windows 98. I'm always looking for the simplest solution with user convenience in mind (thinking of novice and occasional users on shared computers in cyber-cafés and community telecenters).

The possibility of putting the Unicode characters in MS Messenger is indeed interesting. I've managed to do that with Outlook. The facility to do such things with popularly available software (plus perhaps a special keyboard) is something few seem to realize.

Are the characters you use for the Sudanese and other languages you work on listed by language in the Letter Database (http://www.eki.ee/letter)? Worth adding if not. Also, if you have glyphs available for some of those missing from the collection (they have the codes for all characters but not all the glyphs), I think Indrek Hein of EKI would be interested.

The address for the Arial Unicode MS font download that I used is: http://office.microsoft.com/downloads/2000/aruniupd.aspx

Thanks again! (And yes, all you wrote makes sense.)

Don Osborn
Bisharat.Net

Hi again,

I was intending to add all the latin characters, IPA charcaters and some spacing modifiers to the fonts. When i get time to work on it. For the Sudanese langauges (Dinka and Neur) that i'm workking on, diacritics will be important. Both languages use diacritics is day to day usage. Some of the character and diacritic combinations do not exist as precomposed characters and will ahve to be represented by two unicode characters, a base letter (open o or open e for instance in dinka and a combining diactritic. To get this working properly for both upper and lower case characters will require Latin script support to be added to uniscribe (Microsoft's unicode script processor). Microsoft have indicated that they will be doing this, but no indication of when such fuinctionality will be available. Uniscribe support for Latin is important in a number of African languages.

WRT to the two keyman keyboards, i just quickly threw them together as examples, not as definitive kyboard layouts. My use of deadkeys (liek the / key in Hasua2 layout) is a habit from working with Vietnamese. The Control-Alt key sequence (on an English keyboard) carries out the same function as the AltGr key (Left Alt key) on a variety of european keyboards.

I avoided using the alt key since so many programs still utilise alt key combinations. I've quickly thrown together an alternate keyman lyout which uses alt and shift-alt instead. Its available at http://home.vicnet.net.au/~agamlong/files/hausa3.kmp

But i stress again that these keybaord layouts are examples. Ideally there should be some discussion on what the best unicode keyboard layout for Hausa would be. Such a keybaord may be able to support other langauges in the region as well. And obviously decent installation and usage instructions would be required.

yes, a web page discussing Unicode issues in relation to African languages would be a good idea. I haven't seen any pages that specifically address unicode issues relating to African languages. If you wish i could start developing something along those lines. I was already intending to do something similar for Dinka. But since my knowledge of African langauges is limited what i'd write would need to be proofed. Although there maybe other people available more suited ot the task.

Unicode seems to become more common, many of the newer dictionary and digital library projects i've seen are unicode based. One of the projects i have going at work is a small web directory for public libraries indexing a range of web sites in 14 languages. We hope to keep adding langauges to teh directory. All the data is stored in a database as unicode.

I'm familiar with the letter database and most of the langauges i work with exist already in the database. I'll need to double check it.

Andj

"Unicode-Afrique", un nouveau e-groupe/forum de discussion sur l'utilisation de l'Unicode pour langues africaines vient de débuter à : http://fr.groups.yahoo.com/group/Unicode-Afrique

Edited by author 01-24-2002 03:26 PM
Andj, A few quick comments - first technical then policy/collaboration related.

Regarding keyboards, my "alt keys" left and right are different (US keyboard, international English language setting in WIndows). I think the left alt thus functions as the AltGr key on French (& some other) keyboards. Most of the right alt + letter combos are already assigned to various functions.

I haven't installed your keyman keyboards yet, but if I understand correctly, the key combinations will give the same character result in non-Windows but Unicode-aware software?

As for discussion on standardized keyboard layouts for Hausa, that's one of the things I hoped this discussion board would promote. Ultimately one may want to talk of regional or even continental keyboard setups. If we keep using hard keyboards like we have now, what are the possibilities for some modified Eng and modified Fr keyboards to accommodate a whole range of spacial characters used in African languages.

You mention open e and o in Dinka -- these are not used in Hausa but are in the Manding languages and others further west & south. Tones on these same letters in Bambara, for instance, can be indicated with accents (same issue as you have, it seems). (Hausa vowel tones don't pose a problem in this regard.)

From here it looks like there would need to be a lot of communication & cooperation on the issue of keyboards. Various conferences on African languages (most recently in Harare in 1997, see http://www.bisharat.net/bdocs.htm) have mentioned the need for collaboration on languages that cross borders (as so many do), but this little keyboard issue must accomplish the same task and more. It would make little sense to have one keyboard setup for Hausa and another for Zarma etc., and equally little to have completely different key combinations for Hausa of Nigeria and Hausa of Niger. So at least on a subregional level, it would be helpful to have some standardization of keyboard maps to facilitate work in African languages. One might be able to comfortably live with something different for East & West Africa, but if keyboard setups for specialized characters could be harmonized (e.g., left alt/AltGr + right square bracket = open o in Mali or Sudan ... or Australia) it would make a lot of sense to do that.

Folks involved in the work have to make it happen, though. And that takes extra effort that may not seem a priority to many. In this context, I should mention that I was disappointed to learn that the South African translation project mentioned below /m20 does not have collaboration on the languages that cross its country's borders.

BTW, is the web directory you mentioned on the web?

Don Osborn
Bisharat.Net

Hi

French keyboards (along with a range of other european keyboards) have an AltGr key which is used to access additional characters. This is physically located as the right Alt key.

On English keyboards, there is no AltGr key, the left and right Alt keys do exactly teh same thing. The AltGr key is a short cut for a Control-Alt key sequence.

On the Hausa 1 keyboard I threw together, on an english keyboard you need to use the Control-Alt keyboard sequence.

If you installed a french keyboard layout and then selected hausa 1 keyman layout you could use teh AltGr key instead of the Control-Alt key sequence. Although i'd have to redesign the keyman layout so that all key sequences work according to a french and english layout. As the file currently exists when used with a french keyboard i cann't type in U+02BC.

This exercise probably highlights teh need of either

1) separate keyboard layouts optimised for US English and French keyboards.

or

2) a keyboard layout independant of the system keyboard.

although I think i prefer option one, so that a key with "a" printed on it will give you teh letter "a" irregarless of wether its a qwerty or azerty keyboard.

Wrt keyman 5, it only operates on the windows platforms. On Winodws 2000 and XP it will input into any unicode aware program.

On Windows 98 (wince windows 98 doesn't support unicode internally) keyman will only work with programs that are unicode aware and support the WM_UNICHAR function. So its application on Win98 is limited.


Keyman System Requirements
http://www.tavultesoft.com/keyman/downloads/sysreq.php

Unicode Support Information
http://www.tavultesoft.com/keyman/support/FAQ002.php

to that list oyu can addd MSN Messenger under win98/ME/NT4/2000/XP

As consensus develops on what the keyboards should be like, I'm willing to put together a keyman layout to match it.

Thanks for the "Unicode-Afrique" group. I'll have to polish up my virtually nonexistant french. Are messages in English welcome? I'll start putting together a list of links that are relevant.

It is also possible to develop keyman keyboard layouts for 8-bit fonts rather than unicode fonts. I have one Maori keyboard that supports both 8-bit and unicode on the same keyboard. If the application can handle unicode input then data is enetered as unicode, if the application isn't unicode aware then data is entered based on an 8-bit encoding.

Our project home is at http://www.openroad.net.au/ and the directory is at http://www.openroad.vic.gov.au/

there are 14 languages in teh directory. No African languages yet. I hope to beable to commence Oromo, Amharic and Tigrigna sections within the next couple of months.

Andj

A couple of additional thoughts:

to successfuly use Unicode for Hausa, requires not only appropriate unicode fonts, and a unicode input method, but also

* mapping tables between 8-bit charcater sets currently in common usage and unicode.

* conversion utilities to convert to and from unicode. Given the development of mapping tables, there are utilities that can be used for character set conversions.

* locale data for Hausa, have a look at ICU LocaleExplorer [1] for examples. (including regional/national variations)

Andj

Edited by author 02-02-2002 02:33 PM
Thanks again, Andj, for the info and thoughts. I'll quickly try to respond as best I can in this posting to the keyboard issues in your message before last - /m30 - and to your other points in a separate posting.

Re Alt keys on English keyboards, the left and right Alts on my PC with Windows 98 have different functons, I believe because I've set the keyboard up as "International English." I haven't tried resetting it to US English to see what happens. At the risk of belaboring the point let me give details in case this could be of use to other readers.

If I type the right Alt + n I get ñ (this is part of the Int. Eng. setup) in all applications; the left Alt + n was blank so I set it up in my Word 2000 for the "eng" (velar n for several W. African languages but not Hausa [left alt b, d, k, and y give me the Hausa hooked letters]). The left Alt combos do not work outside of Word, which is not really a problem for the things I do. Using shift gives the capital letters. Ctrl + Alt (left or right) doesn't seem to work, so my impression is that the left Alt functions as the AltGr. Or something different.

If this is a feature that could be used in some other English keyboards, then it might be simpler to develop some sort of simple system of two keys (left Alt + letter on qwerty keyboards and AltGr + letter on azerty keyboards) for the special characters used in many languages.

IOW, an "African English" or "African French" keyboard setting in Windows (and other operating systems for PCs) could have preassigned settings for the left Alt or AltGr + letter combinations that are otherwise left to be user defined. Somebody must have thought of this already.

There's a lot I don't know about the keyboard hardware and software settings. Is is possible in a keyboard such as you have put together to treat the left & right alt keys differently? If not, why does it work in the software I have?

In any event, I think that you are right that we would be talking about separate keyboard layouts for English and French keyboards.

Don Osborn
Bisharat.Net

Andj, Here's a quick reply to some of your other points.

Re the mapping and conversions, two thoughts:
1. This seems to have been part of what the RIFAL project was trying to do with several African languages in several countries (including Hausa, Zarma, Fulfulde, Tamashek, and Kanuri in Niger; see http://www.issco.unige.ch/staff/andrei/formRIFAL2000/ ).
2. If I understand correctly, the mapping and conversion you are talking about is important to the extent that there are texts in 8-bit systems. But this is not a problem for using Unicode for new work. In most countries in this region, there is nothing in what one could really call common usage for African languages, but usually a few non-intercompatible fonts used by some specialized agencies and NGOs. There are also some non-Unicode fonts in fairly wide use among scholars outside of Africa (notably AfroRoman from http://www.linguistsoftware.com).

Unicode (whatever some might see as its limitations) not only facilitates multilingual work in Africa but it also in my mind makes the many non-Unicode fonts obsolete, because the latter are not intercompatible. The dual purpose Maori system you mention in /m30 might have limited use for African languages despite the plurality of non-Unicode fonts being used.

Re locale data, how would that be added?

Don Osborn
Bisharat.Net

Having mentioned one commercial font company below that has a non-Unicode font for African languages, I should mention another whose products apparently use Unicode and support Hausa (and many other languages): http://www.unitype.com . Has anyone had any experience with this? (Needless to say, I'm not endorsing this or any other commercial product.)

Don Osborn
Bisharat.Net

Unitype make two software products: Global Writer 98, and Global Office. Global Office is an input system that works with Microosft office 97. Global Writer is a simple standalone word processor.

They both support Unicode. Although most of the fonts they use are not unicode fonts. The programs use a bit of internal trickery to display unicode text with non-unicode fonts.

Global Writer can cut and paste unicode into other unicode aware programs, if the writing script are supported.

Some of the public libraries here use Global Writer for their free public access word processing facilities. Great for those langauges and scripts that Microosft doesn't support yet from Central, South and South-east asia.

Maybe a bit of overkill to use them to provide Huasa unicode support.

Hi Don,

Yes, you are correct, The US-International keyboard does use the AltGr key.

On the standard US English keyboard, the left and right Alt keys preform tah same function (Alt).

On the French standard keyboard the left Alt key is an Alt key, While the right Alt key is an AltGr key (Control-Alt).

There are two English keyboards that also have an AltGr key instead of a right Alt key. These are the US-International keyboard and the Canadian Multilingual keyboard.

Information about the Windows 2000/XP keyboard layouts is available at

Windows Keyboard Layouts
http://www.microsoft.com/globaldev/keyboards/keyboards.asp

So, the Hausa 1 keyboard for a standard english keyboard would need to use the Control-Alt and Shilft-Control-Alt key sequences.

If you are using teh US-International or Canadian Multilingual keyboards, the you would use the AltGr and Shift-AltGr key, with the AltGr key being a simplier way to enter a Control-Alt sequence with a single key.

How the keyboards i have designed work depend on what the current active system keyboard is. If you change between teh US standard and the US-international keyboards while using Hausa 1 the behaviour will change. The standard US English keyboard treats the two Alt keys as exactly teh same key, just as the two shift keys behave identically.

i'm hopeing that in a future realease of keyman it will be possible to tie a keyman layout to a specific system keyboard.

The mapping is important for two reasons:

1) to allow conversion of old documents to unicode

and

2) for software and font developers to agree on what codepoint to assign to characters. If one person uses a "lowercase schwa" and anotherone uses a "lowercase turned e" (both characters look the same) data could not be searched effectively.

WRT locale data, normaly locale data is integrated into the operating system, it handles collation/sort by langauge or region. Number, time, date formats. Currency and number formats, etc. Although this level of integration is unlikely. OS developers don't appear to be very interested in integrating african langauges into teh OS.

LOcale data would also be useful for web services/applications and web sites developed to support african languages. It would provide the application developers and web designs with an agreed standard method of expressing data.

Andj

Edited by author 03-19-2002 03:04 PM
Perhaps it is a good time to take a quick look at what has been covered in this message board thusfar. Quite a few topics have been touched on and a number of resources discussed, including 2 keyboard layouts specifically (though quickly) put together for Hausa.

A quick list of some of the topics discussed follows, with click-on references to the messages on the board. Among the issues not explored much is the potential for open-source software for Hausa (and other African languages). The list follows (Hausa is in one way or another a topic of most messages, so only a few specialized Hausa topics are noted in the list) ...

Don Osborn
Bisharat.Net

Some of the topics covered so far on
Hausa charsets & keyboards

Reason for and goals of establishing this message board /m2 /m12

Diacritics (accents etc.) /m23 /m24 /m26 /m27
Dictionaries on-line /m21
Display of special characters /m9 /m10 /m13 /m19 /m21 /m23 /m24
E-mail /m16
Fonts /m10 /m13 /m15 /m25 /m26 /m33 /m34 /m35
Fonts – info on setting up /m6
Fonts driven from website /m9 /m10
Hausa Ajami script /m3 /m4 /m17
Hausa language resources /m3 /m11 /m21
Hausa Latin script origins /m3
Keyboards – “Hausa 1” & “Hausa 2” /m25 /m30 /m36
Keyboards - key functions /m29 /m30 /m32 /m36
Keyboards - layouts, Keyman /m25 /m27 /m30 /m36
Keyboards - popup (on-screen) /m7 /m23
Keyboards - soft (touch screen) /m22
Keyboards - standardization /m29 /m30 /m32
Language policy /m29
Letter database, mapping tables /m26 /m31 /m33 /m36
Linux /m3 /m6 /m8
Macintosh computers /m3 /m4
Other languages - Arabic (wordprocessor) /m6
Other languages - Bambara /m2 /m4 /m15 /m29
Other languages - Dinka /m27 /m29
Other languages - Fulfulde/Pular /m2 /m4 /m19 /m21 /m33
Other languages - Kanuri /m33
Other languages - South African /m18 /m19 /m20 /m29
Other languages - Tamashek /m33
Other languages – Yoruba /m22 /m26
Other languages - Zarma /m4 /m29 /m33
Translators /m19 /m21
Unicode resources /m3 /m5 /m23 /m28 /m30

Thanks for this synopsis .. it's most useful!
Maggie Canvin

--- Quick Topic daily digest <qtopic+8-JxKHyg9ccPUVB@quicktopic.com> wrote: >

=====
Please visit my fun Ezboard:
http://pub80.ezboard.com/bkingdomofchatalot

http://web.icq.com/wwp?Uin=75652576

Live in the UK? Try http://www.feesch.co.uk/75652576 Join Feesch and make lots of ICQ friends.
__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page
from News and Sport to Email and Music Charts
http://uk.my.yahoo.com
< replied-to message removed by QT >

Edited by author 03-24-2002 01:46 PM
A new electronic working group, A12N-Collaboration, has been established to deal with some of the issues brought up on this message board. For more info, see: http://lists.kabissa.org/mailman/listinfo/a12n-collaboration

Current plans are to phase out Hausa charsets & keyboards in the near future as an active message board. The contents, though, will remain available. More info upcoming.

Don Osborn
Bisharat.Net

[reposted from A12N-Collaboration]

There are three or four modified letters in the Hausa alphabet, all consonants:
"hooked-b" (implosive b)
"hooked-d" (implosive d)
"hooked-k" (implosive or ejective [?] k)
"hooked-y" (implosive y) (in Niger, not in Nigeria)

Siva's site and Indrek's database mention these. Siva should mention that the hooked-y is used only in Niger; in Nigeria 'y is used as two keystrokes as far as I know, but theoretically it could be composed (I will have to check this and also if the y' appears anywhere). The ALRC list does not yet account for Hausa.

The old correct forms for all these letters used apostrophes. For the implosive b, d, & ejective k, Bargery in his classic dictionary placed them before, but most now put them after if they need to resort to the old system (the Hausa pages at http://www.bisharat.net are presented in both the Nigerien orthography in Unicode, and in the non-unicode system with apostrophes).

Tones and vowel lengths are noted in specialized texts with various marks, as Indrek indicates, but this gets rather complicated. Russell Schuh's Hausa language site at UCLA has a very nice page on Hausa orthography, including details on these marks in the Latin transciption, and also a description of Ajami transcriptions:
http://www.humnet.ucla.edu/humnet/aflang/H...iation/writing.html
Note the variant forms of notation taken by diverse sources. This would indicate need for caution on accounting for such notations in code pages. Indeed, might systems of notation unique to one work or author be safely ignored (with all due respect) for our purposes?

BTW, I understand from John Phillips, who has worked a lot on both systems, that Unicode does support Hausa in Ajami, but that is another subject.

Don Osborn
Bisharat.net

Hi Don

Siva's site does mention that Hausa uses an apostrophe, although the
code point given is to a modifier letter prime, rather than a modifier
letter apostrophe.

thus the previous question i asked on a12n-collaboration about Siva's
codepoint allocation.

Btw there was a brief discussion on the Unicode mailing list last week
about the reasons for not using the ASCII apostrophe in languages, but
rather using a modifier letter apostrophe.

A similar discussion took place on the aust-lang mailing list about the
use of the modifier latter apostrophe in the Yolngu-matha languages.


Andre

Andrew, Here's a really belated response...
Probably the apostrophe should then be listed as a "special character" ("extended Latin character" ?) in a chart of codes for Hausa characters.

In practice however I think most people would use the plain apostrophe, and not consider the difference from modifier letter apostrophe (life is complicated enough already!). Eventually language settings in a wordprocessor would have to make the change? And the search capacities of search engines (such as in /m43 above) would have to treat the two as the same terms searched?

Don Osborn
Bisharat.net

I take the liberty of reposting a message Andrew sent to the A12n-collaboration WG which would be of interest to readers of this board. NB- the Bisharat! pages do not have language tags.

Don

[letter follows]
While working on a project at work, I started playing with creating sidebars tabs for netscape 6 that allowed unicode input into search engine query fields.

I created one that included a small virtual keyboard to allow input of extended latin characters for Hausa into a search prompt. there is currently a search field for Google and one for Alltheweb.

the web page is at http://www.openroad.net.au/languages/tools/hausa.html

A icon on the bottom of the page can load a Netscape 6 sidebar tab for those who use Netscape 6.

Obviously there aren't many Hausa unicode web pages out there. it was able to search and locate the unicode pages on the Bisharat! web site. The page was created as an experiment, to test input and searching on win95//98/ME platforms where it wasn't possible to directly type in Hausa.

As far as I know the Bisharat! web site does not use langauge tagging in its HTML, so I wasn't able to test if it was possible to limit a search results to the Hausa language. That experiment will have to wait for another day.

Have a play. Comments and suggestions welcome.


Andrew

Hi All,

QT - BisharatNet wrote:

 >
 >.
 >
< replied-to message removed by QT >

Hi All

Yes, this is the problem.

As far as the unicode standard goes, characters are differentiated by
function/purpose not by appearance.

Essentially there is a different function between U+0027 APOSTROPHE and
U+02BC MODIFIER LETTER APOSTROPHE


Most likely teh end user will treat the two the same because they're
visually similar. And I suspect that most people will just type in U+0027.

I also suspect that softwrare will treat them very different. For
instance in the Google search engine a single apostrophe will be treated
similar to a wildcart character, so something like "ha'ka" would yield
"ha ka", "ha'ka", "ha-ka", etc but probably wouldn't matcg U+02BC

apostrophes in pairs are used as string delimiters in a range of
scripting and programming languages. In such scripts the presence of an
apostrophe in the wrong place could break the script. Although this may
mean we just have to be extra careful to escape the character.

some programs when sorting words will treat an apostrophe as a space or
may ignore its presence altogether ...

To lots of programs an apostrophe is not a letter, at most its
punctuation or a character to be ignored.

Word processors (like MS Word) already do a conversion on U+0027, its
converted to U+2018 or U+2019 Left/Right single quotation marks. Or the
languages equivalent Unlikely taht in US-centric or Euro-centric
softwrae it will ever do a conversion between U+0027 and U+02BC.

in the average case like emails, it probably wouldn't make any
difference. If you wnated to use U+0027 as a character in MS Word you'd
need to turn off the autocorrecting feature so that it isn't mistakenly
converted to either U+2018 or U+2019. But that means if you actually
wnated to type in U+2018 or U+2019 yoiu'd have to go through a more
complicated process ... of using Insert/Symbol menu command.

Andrew

An article on the BBC site, "Urdu website breaks new ground," may be of interest for any interested in producing Hausa web content in Ajami:
http://news.bbc.co.uk/hi/english/sci/tech/...2139000/2139300.stm

The Urdu script, which is also Arabic-based, is not standard on any OS, and apparently fonts must be hard to get. BBC developed a Unicode font and offers it for download, which enables people using a range of systems to read their Urdu site.

Are there any Ajami Unicode fonts suitable for Hausa or other sub-Saharan African languages? I guess there's not much chance that BBC would invest in this too...

Don Osborn
Bisharat.net

A few months ago I wrote in /m39 that plans were to phase this board out in the near future. Plans have changed for the moment and it will continue to function for occasional information and discussion on computer/internet issues specific to the Hausa language. The focus will remain on how to enter and convey Hausa text on computers and the internet.

The main reason is that this narrow set of issues remains important, with still unresolved questions, for facilitating use of Hausa in ICTs. Also, this board has received quite a number of visits (even though few posts) since its creation almost a year ago, which may indicate a certain level of onging interest. As more people attempt to use Hausa and other languages in ICTs, the interest in and need to know about the kinds of technical topics raised on this board may increase. So, it seems at the moment premature to close it.

Thank you for your continued interest.

Don Osborn
Bisharat.net

Edited by author 08-25-2002 03:46 AM
Here are a couple of reference pages on the Latin alphabet for Hausa in Nigeria and Niger from Alphabets of Africa (R. Hartell, ed., UNESCO & SIL, Dakar, 1993) as recently posted on the Rosetta Project site:

Hausa (Nigeria)
http://www.rosettaproject.org:8080/live/se...scale=six&version=3

Hausa (Niger)
http://www.rosettaproject.org:8080/live/se...scale=six&version=2

Also an excellent reference on written Hausa (both Latin and Ajami transcriptions) is available at the UCLA Hausa page:
http://www.humnet.ucla.edu/humnet/aflang/H...iation/writing.html

The "letter apostrophe" that Andrew referred to in /m45 functions as a letter in the alphabet (see also /m41 & /m42). Although it may not make much difference how one arrives at the apostrophe shape for printing a document, it could make a difference for manipulating digital text (sorting etc.) or presenting & accessing it on the web. Hence the attention to the issue of the different apostrophes (there being several, for better or worse).

Don Osborn
Bisharat.net

The company that runs the search engine "Google" has a "Translate Google Into Your Language" program and Hausa is among the languages for which they are developing support (although 0% of the site has been translated so far). See: http://services.google.com/tc/Welcome.html

Since many people operate systems that can't handle Unicode, I'm not sure how they'll deal with the hooked letters and all.

It would be very interesting if an eventual Hausa language version of Google could incorporate the feature that Andrew developed for entering the special characters (see /m43).

Don Osborn
Bisharat.net

re /m49

It would be good to see a google interface in various languages.

The interface I designed could be easily adapted. I just have to replace the English labels with Hausa labels and change a two letter flag to use the google hausa interface rather than the english or french interface when its available. the conversion should only take a minute or two once a Hausa google interface is available.

The problem is more with access to fonts and typing, most web browsers support unicode (utf-8) ... maybe eot fonts could be incorporated ... have to think about that ...

One word of warning ... I've been testing Google with Vietnamese and a few other languages. Just because Google has a translated interface for a language, doesn't mean that Google is optimised for that language.

What this means in practice is that if you search in Unicode, you'll only find Hausa websites in Unicode. You essentially have to search in each relevant character set to find a cross section of sites. For vietnamese for instance I have to do searches using precomposed Unicode Vietnamese, Microsoft's keyboard layout for vietnamese which uses a mix of precomposed and uncomposed characters, I also ahve to search using the legacy character sets: VISCII, TCVN-3, VPS, VNI, Windows-1258, etc.

For Hausa, this means that it would be necessary to search in the range of character sets that are use in the wild,

or

it will require a concerted effort to promote the use of Unicode for Hausa on the web. Google would really only be useful if most Hausa websites that are developed use unicode.

If we could build mapping tables between the legacy character sets and unicode it would be possible to develop conversion tools to assist in the process of creating unicode web sites.

Andrew

re /m50

Just an additional thought:

as well as a unicode based search engine, what may be useful is a unicode based Hausa web directory.

There are two possibilities:

* Using the Open Directory project to create a Hausa web directory, eg http://www.dmoz.org/World/, or

* we have an old tool (circa 1999) that we were using in-house, it was designed to be used as a bilingual unicode web directory. We can also use embedded open type fonts with it to get around some of the font access issues (for those using IE5/5.5/6). [1]

Just a thought. A selective unicode web directory may be of more use initially that a serach engine, until the number of Hausa web sites expands.

re /m50

I would strongly urge you to use UTF-8 for a Google Hausa translation. Most newer Windows and Explorer and Netscape work just fine displaying the hooked characters in several fonts that come with the OS. Arial Unicode MS is the obvious one.
This will spur on further support for properly spelled Hausa in the future as it will be more widely seen. There is a major chicken and egg problem with minority languages on the internet where people feel the need to "dumb down" their languages for old software because they feel this is easier than hoping software will change to support them.

The problem with Vietnamese on Google is also due it being a minority language on the internet. Localizing the Google interface is dont by the community, internationalizing the search engine itself is done by Google internally. French and Spanish Google will search properly for words even if you've typed them without accents, German google treats ae and a-umlaut identically. It also appears that Google indexes are kept in Unicode, converted from the native encoding of the web pages. I think either Google has ignored the Vietnamese encodings and doesn't convert them to Unicode or possibly many Vietnamese pages don't correctly declare their encoding. I don't think Google tries to make guesses about undeclared encodings. The fact that some Vietnamese encodings require combining characters, some don't, and some don't even allow all possible diacritic combinations make this a tougher problem than European languages. Google could certainly do it right if they felt the market was there.

Likewise it would be possible for Google to create a Hausa search mode where it would treat hooked Unicode letters and ASCII letters with apostrophes alike. It would probably take some pressure for them to do this.

Also note that having hooked characters in the translation is quite a separate issue from being able to type them into search queries or being able to match them in searches.

Having said that, you should approach Google because they are one of the coolest internet companies out there.

re /m52

Yep, I agree that unicode is necessary. All of my non-english work has been unicode based since 1997.

The inherent problem with Google is that they do not attempt to make a language aware search engine. In some sense it would be not be possible to. Google is a great tool for English and soem other languages, but its a partial tool for others.

Vietnamese is probably a bad example. although the easiest search engines to use for vietnamese are those with inbuilt conversion for vietnamese character sets and encodings. At least with vietnamese, I need to enter the same search term 7 times using different character sets. Not efficient. Its irksome that Google doesn't even use normalisation.

orthographic variation could be another source of problem, dpeneding on wether an ascii apostrophe was used or a modidier letter apostrophe. There are three cases: unicode hooked letter, ascii apostrophe and letter, and modifier letter apostrophe and hooked letter. As to whether you want to equate hooked-letters with either apostrophe and letter is questionable. It would depend on how it would impact on other languages that may use the apostrophe in other ways.

I'll have to try to locate my notes of some test searches i tried using the ascii apostrophe and modifier letter apostrophe on google.

Search engines including Google do not handle apostrophes well (at least last time i ran some tests).

Multilingual search engines are a cute and sometimes useful tools (like Google or Alltheweb), but unicode based search engines tailored to the needs of specific languages tend to be more useful [at least form the point of view of multilingual reference and information services].

personally, I look forward to more google language interfaces. Although I am aware of the limitations. The reality is that google as is would be better for some African languages than others.

Andrew

Some quick responses.

Re Unicode pages in Hausa, there are still very few. The main Hausa-language sites use the ASCII characters and contributors to Hausa-language discussion groups are obliged to do the same. There will be more webpages in Unicode (I know of one in process including the Hausa for Niger), and one wonders when setups like Yahoogroups will facilitate Unicode use.

Re the pluses and minuses of Google in this regard, I'll send a note - it would be interesting (though I don't expect it) to have some feedback on some of the questions raised in the last few messages.

Re building a Hausa web directory /m51, this could be an interesting project in any case.

Don Osborn
Bisharat.net

re /m54

a unicode based web directory would provide access to both unicode and non-unicode web sites, and would be advantageous at this stage of Hausa web development.

Long term unicode search engines would be preferable, but the current level of Hausa web content may make a directory a more efficient resource location tool. Search engines become practical once a certain saturation l;evel has been achieved. At least thats my twop cents worth.

re. Yahoogroups .... my employers have a similar sort of tool which was developed for a state government project here (My connected communities). Currently they're working on some bugs related to the integration of email and forum posts. The site is unicode based. It has irc, forum, calendar, file storage. Has teh ability to create simple templated websites for people who don't know how to create web sites. And a few other features. Hausa unicode chat would work well using chatzilla (Mozilla IRC client) assuming you have Hausa unicode input software that works with Mozilla.

There are plans to enhance its multilingual capabilities in the future. But seems to work relatively well (once they fix that email-forum integration problem).

Although ... it has the same problem as any unicode based web service (wrt to Hausa). Direct input of unicode into a web browser is straight forward on Win2000/XP but is more limited on Win95/98/ME.

By the way, has anyone tried to use MSN messenger for chatting in Hausa using Unicode?

Andrew

Edited by author 10-06-2002 07:23 AM
This message board has been in existence now for a little more than 13 months. Many topics have been touched on and a number of resources discussed.

The advantage of this board format is the ease in accessing and posting, with the option but not necessity to subscribe. The disadvantage is that there is no search mechanism to find specific topics of interest. To assist, I put together an index in /m37 which I have updated below. Another search option is to select to display all messages and then search for keywords in the text with your browser.

Don Osborn
Bisharat.Net

Some of the topics covered so far on
Hausa charsets & keyboards

Reason for and goals of establishing & maintaining this message board /m1 /m2 /m12 /m47

Chat /m55
Diacritics (accents etc.) /m23 /m24 /m26 /m27
Dictionaries on-line /m21
E-mail & e-mail lists /m16 /m54 /m55
Fonts /m10 /m13 /m15 /m25 /m26 /m33 /m34 /m35
Fonts – info on setting up /m6
Fonts driven from website /m9 /m10
Hausa Ajami script /m3 /m4 /m17 /m46
Hausa language resources /m3 /m11 /m21 /m48
Hausa Latin script extended/special characters /m40 /m41 /m42 /m45 /m48
Hausa Latin script extended/special characters (apostrophe as letter) /m40 /m41 /m42 /m45 /m48 /m53
Hausa Latin script extended/special characters (display of) /m9 /m10 /m13 /m19 /m21 /m23 /m24 /m43
Hausa Latin script origins /m3
Keyboards – "Hausa 1" & "Hausa 2" /m25 /m30 /m36
Keyboards - key functions /m29 /m30 /m32 /m36
Keyboards - layouts, Keyman /m25 /m27 /m30 /m36
Keyboards - popup (on-screen) /m7 /m23 /m43
Keyboards - soft (touch screen) /m22
Keyboards - standardization /m29 /m30 /m32
Language policy /m29
Letter database, mapping tables /m26 /m31 /m33 /m36 /m40
Linux /m3 /m6 /m8
Macintosh computers /m3 /m4
Other languages - Arabic (wordprocessor) /m6
Other languages - Bambara /m2 /m4 /m15 /m29
Other languages - Dinka /m27 /m29
Other languages - Fulfulde/Pular /m2 /m4 /m19 /m21 /m33
Other languages - Kanuri /m33
Other languages - South African /m18 /m19 /m20 /m29
Other languages - Tamashek /m33
Other languages – Yoruba /m22 /m26
Other languages - Zarma /m4 /m29 /m33
Search engines & Hausa /m43 /m49 /m50 /m51 /m52 /m53 /m54 /m55
Translators /m19 /m21
Unicode (& Hausa pages) /m43 /m50 /m51 /m52 /m54
Unicode resources /m3 /m5 /m23 /m28 /m30
Web directory in Hausa /m51 /m54 /m55

Here are excerpts from yesterday's announcement of a new e-mail list on Hausa topics: H-Hausa. It is actually the successor to the longtime Hausa-L list and is moderated by John Philips (who has also previously contributed to this message board).

Don Osborn
Bisharat.net
__________________________
ANNOUNCING H-HAUSA: H-Net Network on Hausa language, literature and culture. Sponsored by H-Net, Humanities & Social Sciences On-line, Michigan State University

ABOUT H-HAUSA

H-Hausa is for the discussion of issues related to Hausa language, literature and culture. It began as a list concerned exclusively with language, but it has also hosted discussions ranging from traditional, Arabic-based orthography to computerization of Hausa, from grammar to culture and other subjects. It has attracted academics, missionaries, former expatriates and Hausa themselves to the only list which deals primarily with Hausa issues.

[ . . . ]

Logs and more information can also be found at the H-Net Web Site, located at: http://www2.h-net.msu.edu/~hausa/ [see this site for subscription info and the full text of the announcement]

[ . . . ]

For more information about H-Net, [ . . . ] point your web browser to: http://www2.h-net.msu.edu

Edited by author 10-11-2002 11:03 AM
In addition to the new H-Hausa list (see below or click /m57) , there are several other discussion lists on Hausa-related topics (most are mainly in Hausa). This is not to say they are all comparable - in fact one of the interesting features (and a reason not to close this board yet) is that each may have a separate focus and together reflect & serve an increasing number of internet users among Hausa speakers.

It is interesting to note that without the option of Unicode fonts and appropriate input methods, folks type in Hausa using just the ASCII characters - and from what I saw in a quick look, without using apostrophes to mark the implosive and ejective consonants (which the hooked letters indicate in printed text). I don't know if either H-Net or Yahoogroups are planning anything to facilitate use of Unicode extended characters.

The list of Hausa e-groups (there may be more groups than these - I just checked Yahoogroups):
Finafinan_Hausa (in Hausa; quite active; 100+ members)
http://groups.yahoo.com/group/Finafinan_Hausa/
Hausa (mainly in Hausa; rather inactive; 50 members)
http://groups.yahoo.com/group/hausa/
Hausa_Muslim (languages?; rather inactive; 28 members)
http://groups.yahoo.com/group/Hausa_Muslim/
Littattafan_Hausa_na_zamani · Contemporary Hausa Novel (founded by Abdalla Uba Adamu, who has also previously contributed to this message board; in Hausa; rather inactive; 20 members)
http://groups.yahoo.com/group/littattafan_Hausa_na_zamani/
Hausadahausawa · Hausa Da Hausawa - Promote Contemporary Hausa Literature and Culture (mainly in Hausa; rather inactive; 12 members)
http://groups.yahoo.com/group/hausadahausawa/
Finafinanhausa · Finafinan Hausa (languages?; rather inactive; 5 members)
http://groups.yahoo.com/group/finafinanhausa/

Don Osborn
Bisharat.net

Edited by author 10-14-2002 02:26 PM
Andrew asked in /m55 about whether anyone used Unicode (which has the hooked letters) in Hausa chat - referring to the active messaging and not the "bulletin board" type "chat" services. It is an interesting question now whether anyone actively uses Unicode for much beyond a small percentage of web content. This will change, I'm sure, as Unicode becomes more prevelant, but in the meantime I suspect that in chats, like in the e-mail lists mentioned below, people currently just use the ASCII characters.

It would be worth it to experiment with a Unicode font in a MSN Messenger chat.

Don Osborn
Bisharat.net

How many (what percentage) of people who access websites with Hausa themes actually can read Hausa? This may be a question webmasters and designers of web sites ask when considering how much effort to put into the Hausa language web content. The Hausa Voices site at http://www.hausavoices.com/ has a poll that, even though unscientific, has some interesting results on this question.

The more Hausa-literate internet users there are and Hausa-language web content is produced, the more important it would seem to have appropriate fonts & user interfaces to facilitate use of the language on computers and the internet.

Don Osborn
Bisharat.net

Edited by author 10-20-2002 05:16 AM
I've just learned about a font and keyboard package for Nigerian languages called Konyin (with thanks to Walter Oluwole). See http://www.konyin.com for more information. Does anyone have any experience with this?

Don Osborn
Bisharat.net

Edited by author 11-04-2002 04:16 PM
Here is a Hausa (Latin transcription) keyboard for Macintosh (OS X 10.2): http://homepage.mac.com/thgewecke/fs/ (with thanks to Tom Gewecke). NB- This URL works only from Mac computers and the keyboard only on Mac computers.

Don Osborn
Bisharat.net

PS- I'm in the process of updating the A12n gateway at http://www.bisharat.net/A12N. Also, there is a new web address with access to this and three other language/country specific discussion boards: http://www.quicktopic.com/share?s=QSpo

A retransciption of the Universal Declaration of Human Rights in Hausa is now available at:
http://www.bisharat.net/Demos/UDoHR_HAU.htm

This is the same as that on the website of the UN High Commisioner for Human Rights but with the hooked letters inserted.

Don Osborn
Bisharat.net

This is reposted from A12n-collaboration. It concerns keyboard layouts for Unicode fonts on Macintosh designed by Eric Rasmussen to facilitate typing of extended Latin characters used in African languages.

> I have finally posted the first of my Latin Extended keyboards at:
> http://homepage.mac.com/chinesemac/LatinExtended/index.html
>
> These are Mac OS keyboards, so the compression format used for the
> downloads is SIT.
>
> I went with a minimum set of African characters, leaving out a number
> of IPA characters in Hartell: small alpha, small u with stroke, small
> turned v, and small r fishhook. [Note that u with stroke and r with
> stroke can be composed using dynamic input.] I also left out a number
> of IPA characters already adopted into Unicode for African (with
> capitals in Latin Extended-B): upsilon and t with retroflex hook,
> neither of which are in Hartell. I added the modifier letter small
> raised w (U+02B7), which is used for Twi.
>
> There will be a version 1.1.1, so additions are still possible...
>
> Thanks for your help,
>
> Eric

Warm welcome to Alnemat Arabic Christian Internet Magazine, We love you! Please visit us at:
http://www.TheGrace.com
http://www.TheGrace.net
äÍííßã ÈãÍÈÉ Çááå æäÊÃãá ÒíÇÑÇÊßã ÇáßÑíãÉ áãæÞÚ ÇáäÚãÉ /
ãæÞÚ ãÌáÉ ÇáäÚãÉ íÞÏã ßáãÉ Çááå ÇáßÊÇÈ ÇáãÞÏÓ ÇáÅäÌíá ÑÓÇáÉ ÇáÓíÏ íÓæÚ ÇáãÓíÍ ãÌÇäÇø æÞÑÇÁÇÊ ãÎÊÇÑÉ ÊÃãáÇÊ ãÕíÑíÉ ÞÕÕ æÇÞÚíÉ ÔåÇÏÇÊ ÔÎÕíÉ ÊÑÇäíã æÃÔÚÇÑ ÑÏæÏ ãÄßÏÉ ßÊÈ ÈäøÇÁÉ ãæÇÞÚ ãåãÉ Bible Readings in Arabic Studies Stories Testimonies Hymns and Poems Answers Books Links/Alnemat Journal Arabe Chrétien ( La Gràce ) La Revue Arabe sur Internet offrande La Sainte Bible ( Al-Injil ) L'Evangile de Jésus Christ gratuit /

Thank you for your message. It illustrates the problem that anyone would have with non-Latin scripts in certain internet applications. The text you had presumably in Arabic script comes through as a garbled string of Latin characters (the same would have happened of course with a Hausa text in Ajami).

Below is a test of three Arabic characters - ba shin ra - using the Unicode decimal codes (which is presumably all that subscribers will see in their e-mail):
ب ش ر

I think that any Ajami would have to be entered similarly, which would be a bit tedious - or require conversion of a Word document (for instance) to HTML and then pasting the HTML code.

That said I would ask contributors to please stay to the topic of the board which is technical aspects and applications of Hausa on computers and the internet.

Don Osborn
Bisharat.net

Another kind of keyboard is one based on a graphics tablet where input is made with a stylus. One intended for African languages that would work for Hausa is the Aim8 Pan African Language Tablet - see http://www.largeformatcomputing.com/LTAB/PanAfricanLTAB.htm

My understanding is that this will be revised in the not distant future.

Don Osborn
Bisharat.net

A draft tablet keyboard layout designed by Lee Pearce and intended specifically for Hausa and other Nigerian languages can be viewed at: http://www.bisharat.net/A12N/Projects/NigeriaTabletKeyboard.htm. This is part of the redesign process mentioned in /m67. As I understand it, the resulting product will eventually be offered at a new site: http://www.language-keyboard.com/.

Other draft layouts for Africa as a whole and some languages and regions are at http://www.bisharat.net/A12N/Projects/ (top of the page for traditional keyboards; bottom for tablet keyboards).

Don Osborn
Bisharat.net

An issue has come up re the possible use of the ts ligated digraph U+02A6 in Hausa. The first question I guess is whether anyone uses this (e.g., in publications). On the other hand, I understand from previous discussions that use of two characters in sequence is preferred over use of combined digraphs.

However, in a language like French, ligated digraphes such as the œ (oe) U+0153 is used, perhaps only as something the software automatically resolves to (that's what Word does when you set the language to French). No one types a "special character" for the oe when typing oeuvre for example, the software can choose it, giving you œuvre. As far as I've seen, this does limit the search capacities - i.e., searching oeuvre doesn't find œuvre and vice versa; the two are not considered equivalent in the software.

So there are a couple of questions:
1) First are search capacities (in software, over the internet) overcoming this limitation in French?
2) If so what would be required to get Hausa to the same place? - if whatever standards groups/agencies deal with Hausa decide that the ts graphic should be recognized in computer usage.

Don Osborn
Bisharat.net

Edited by author 05-26-2003 04:45 PM
On a recent trip in Niger I noticed signs with village names spelled with the hooked-K (uppercase) had the hook on the left stem hooking right, which is like the lower case hooked-k and not like the upper case in Unicode (and some earlier references) which shows the hook on the right: Ƙ

BTW, the ts ligature mentioned in /m69 is: ʦ. I am not aware of its active use in any Hausa typography, though.

Don Osborn
Bisharat.net

Dear Members:

We have keyboard that has all Nigerian alphabets. If you are interested
please let me know. Once you have this keyboard you don't have to change
your keyboard before you type in English...it is all in all keyboard.
It is called KONYIN keyboard and the driver software is Windows ® OS based,
which allows the driver to be application independent. The keyboard works
like any other standard keyboard driver, with enhanced functionalities for
Nigerian users. In short, it is a standard keyboard driver for everyday use
and it allows users to type Nigeria specific unique alphabets without
changing how you type today. Users do not have to remember any complicated
codes or shortcuts. The software is designed for use both at homes and
offices.
The Keyboard driver software uses proprietary source codes to generate all
the identified unique alphabets. The source codes further create a Nigeria
specific code-page in your computer registry, which means that you will be
able to see the Nigerian currency sign any time you use a spreadsheet
application like Microsoft Excel or Lotus.

System requirements: Windows ® (9x, ME, NT, W2k, XP professional). KONYIN
basic package includes keyboard and wrist pad, keyboard driver software only
$99.98 please add shipping and handling cost. If you are buying from
California please add 8.25% tax. Thank you for your inquiry. Contact oladokun@ix.netcom.com

Oládòkun

I recently received this request for info on input for Hausa...

----- Original Message -----
From: tanko abubakar
To: bisharat@bisharat.net
Sent: Sunday, January 04, 2004 5:15 PM


Best compliments of the season.

I am glad to have visited your website through google search. I think you are providing a great service to writers and researchers.

Kindly help me, if you can: I need a keyboard or software or wares that will have both the characters of the IPA and also the characters of HAUSA alphabets. I am working on a project that has to do with broad trancription as well as continuous writing in Hausa. I would be most pleased to receive the most helpful information including price list and how to order.

Waiting to hear from you soonest.
Abubakar.

Dear Tanko, Sannu!

Thank you for your letter and your kind words.

For commercially available packages there is the Konyin system specifically intended for Nigerian languages (see http://www.konyin.com/ ). Also a couple of others (such as Unitype http://www.unitype.com) that cover a lot of world languages and are probably not worth the expense if your main work will be with Hausa and English.

You may be able to use a unicode font such as Lucida Sans Unicode, Arial Unicode, or Gentium (the first two are with many Windows systems and the third is available free for dwnload at http://scripts.sil.org/cms/scripts/page.ph...rsi&item_id=Gentium ). These three fonts also have the IPA range. With these, you could chose one of two keyboard solutions:
1) If you are working with Word, you could set up key macros to type the special letters you need, or
2) If you are working across appications in Hausa, you could use a keyboard layout such as the Keyman keyboards accessible via http://www.bisharat.net/A12N/Projects . These would require the Keyman 6.0 software from http://www.tavultesoft.com which is free for individual use.
This font + key(board) approach would be cheaper and maybe just as effective.

I'll copy your letter and this reply to the A12n groups and the Hausa Charsets & Keyboards forum for any other feedback.

Hope this helps.

Don Osborn
Bisharat.net

The Nationmaster.com site has a page on writing Hausa webpages. See http://www.nationmaster.com/encyclopedia/Hausa . It's not entirely clear to me from a cursory reading how the routine they describe works, though inserting the decimal codes in HTML for the hooked letters is standard.

Don Osborn
Bisharat.net

An online Hausa verb conjugator is available at http://www.verbix.com/languages/hausa.shtml (the Verbix.com site has a number of features and shareware for diverse languages).

Don Osborn
Bisharat.net

A draft of a list of Hausa plant names compiled by Roger Blench (who is or was working with with ODI) is available in PDF format at:
http://homepage.ntlworld.com/roger_blench/...20plant%20names.pdf

I tried to block and copy names with the hooked letters, however, but those letters were transformed to other ANSI characters. (Adobe's coding? Or my Windows 98 system?)

Don Osborn
Bisharat.net

Just learned on the Internet-Niger list that an organization in Saudi Arabia called the King Fahd Complex for the Printing of the Holy Qur'an (Matattarar Buga Al-Kur'ani Mai Girma ta Sarki Fahad) has a multilingual site with Hausa content. See: http://www.qurancomplex.org/

Don Osborn
Bisharat.net

Edited by author 09-19-2007 12:00 AM
The A12n-forum has a discussion on Hausa and ICT beginning with an inquiry by Peter Rogers concerning Hausa orthography and how widely it is used (8/23/04). See:
http://lists.kabissa.org/lists/archives/pu...forum/msg00136.html

Don Osborn
Bisharat.net

This item may be of interest. I understand that there is also an open source localization effort for Hausa (no details yet).

Don Osborn
Bisharat.net

"Microsoft Endorses Due Process in IT" This Day (Lagos)
December 15, 2004, Posted to the web December 16, 2004
http://allafrica.com/stories/200412160060.html

" ... Microsoft Nigeria has also begun a process of software localisation in the three major Nigerian languages as part of its commitment towards reinvesting into the Nigerian society. By this, Microsoft intends to come up with a software programme that would enable functions to be performed in the three local languages of Igbo, Yoruba and Hausa. Ilukwe disclosed this in Lagos last week, saying that such a programme has already been produced in South African for Swahili language."

Greetings and Happy New Year.

Kindly see our modest contributions towards the promotion of Nigerian languages since 2001. Our award winning word processor and translator - Paradigm Lingua® makes it easy to produce documents in several African languages on the PC.
http://www.paradigmint.net/lingua.htm
For details, please see our News Page.

Please have a look at the following wiki page on the Hausa language:
http://www.bisharat.net/wikidoc/pmwiki.php/PanAfrLoc/Hausa

... it is part of a survey of localization in Africa for the PanAfrican Localisation project, and as such is intended to provide a minimum of linguistic background plus useful information on use of the language on computers and the internet. Comments, corrections, and new information is welcome.

Don Osborn
Bisharat.net

There are a couple of notable developments in ICT localization in Hausa, even if these are in an early stage:
1) There is a Hausa wordprocessor in an early stage. More on that as soon as I can get installed. (Delays have been on my end.)
2) I received a message from Ramon Mathias Soares Pontes concerning a keyboard layout done with MSKLC (on Windows XP) for Hausa Ajami.

More on both of these soon.

Don Osborn
Bisharat.net

If your kids speak only english and you want to teach them Hausa using your PC, there is only one keyboard in the world that gives you all Hausa alphabets and all English alphabets on a single layout for easy direct access typing.

The keyboard is KỌNYIN Nigeria Multilingual Physical Keyboard made by LANCOR Technologies of Boston, MA (http://www.konyin.com)

KỌNYIN Multilingual Keyboard

FYI, from http://www.conferencealerts.com/seeconf.mv?q=ca1h8600 . I have not been able to access the site so don't have any details on topics to be covered. DZO

LANGUAGE,CULTURE AND GLOBALIZATION
21 to 24 November 2005
Owerri, Imo State, Nigeria

Website: http://www.apnilac.4t.com
Contact name: ANOPUE CALISTUS CUSSONS
E-mail: callycussons_AT_yahoo.com (to e-mail the conference organizers, please
replace _AT_ with @)

Organized by: ASSOCIATION FOR PROMOTING NIGERIAN LANGUAGES AND CULTURE.
Deadline for abstracts/proposals: 20 October 2005 (Check the event website for latest details.)

NITDA to promote Nigerian keyboard

Everest Amaefule, Abuja

The National Information Technology Development Agency has pledged to promote the study of Nigerian languages through the use of information technology.

Director-General, NITDA Prof. Cleopas Angaye, made the commitment on Friday in Abuja when a team from Lancor Management Limited, presented a special computer keyboard and software known as Konyin, that can write most Nigerian languages to the agency.

Angaye said the promotion of Nigerian languages through IT has become imperative at a time when several languages are becoming extinct because of generations that have been alienated by their mother tongue.

He congratulated the Lancor team for integrating both software and hardware to achieve the feat of the special Nigerian keyboard and pledged to promote the keyboard especially in the public sector.

Full Story... http://www.punchng.com/computer/article05

The PUNCH, Monday, January 16, 2006

Issue: Now I can type words from Nigerian native languages on my PC, but I cannot save these words into my custom dictionary. Every time I tried to save words with Hausa, Ìgbo or Yorùbá characters, I get a message that: “The custom dictionary is full. The word was not added.”

Solution: The custom dictionary is a notepad that is defaulted to ANSI encoding, to save non ANSI characters, like Hausa, Ìgbo or Yorùbá words that includes Ẹẹ, Ịị, Ọọ, Ụụ or tonal marks, you will need to change the custom dictionary encoding from ANSI to Unicode.

How to change Custom Dictionary encoding in Windows XP:

1. Click - .START button on the left bottom of you screen;
2. Click - My Computer from the menu;
3. Open – Local Disk (C:) from the menu;
4. Open – Documents and Settings folder;
5. Open – [Your User Profile] folder;
6. Open – Application Data folder;
7. Open – Microsoft folder;
8. Open – Proof folder;
9. Open - CUSTOM file;
10. Click – File at the top left corner;
11. Click – Save As from the menu;
12. On the line for Encoding – click the dropdown and select Unicode;
13. Click – Save button;
14. Click – Yes button;
15. Click –CLOSE ALL FOLDERS

KỌNYIN Nigeria Multilingual Keyboards
http://www.konyin.com

Kwamfyutar Cinya Ɗaya ga Kowani Yaro
http://laptop.org/index.ha.html

Happy New Year 2007! Hope the holiday season was good (whichever holidays you observed)!

I've opened up a new "wikigroup" on the PanAfriL10n.org website/wiki for Nigerian localisation: http://www.panafril10n.org/wikidoc/pmwiki.php/NG-L10n/HomePage

The object is to provide a more flexible space for Nigerian localisers to list their contact details, websites, and projects. This is in some ways an online "virtual plaza/market," interactive, with the added advantage that it links with the larger PanAfrican Localisation wiki and other country-specific "wikigroups."

Among other things, I am trying to set up an RSS feed from this forum to the new wikigroup.

Don Osborn
Bisharat.net
PanAfrican Localisation project

The PanAfriL10n.org page on Hausa has been updated. See http://www.panafril10n.org/wikidoc/pmwiki.php/PanAfrLoc/Hausa (corrections, more updates are invited).

Don Osborn
Bisharat.net
PanAfrican Localisation project

Edited by author 02-17-2007 02:01 PM
Just modified the board name from Hausa charsets & keyboards to Hausa charsets and keyboards. The ampersand was making a problem for the RSS feed, strangely enough. Now the feed works. See: http://www.panafril10n.org/wikidoc/pmwiki.php/NG-L10n/L10nForums

Don Osborn
Bisharat.net
PanAfrican Localisation project

Hi!
Nice design on www.panafril10n.org, good info, thank you.

Re /m100, thank you. If there are any changes / additional information you think is necessary for the Hausa page (see /m98) or the NG-L10n page (see /m99), please let us know.

Don Osborn
Bisharat.net
PanAfrican Localisation project

Hello, (Barka da warhaka)

I realy appreciate what I came across and surely I have contributions to make.

Sai anjima

Sannu, Thanks for the feedback.

I meant to mention earlier that Abdalla Uba Adamu posted information about some blogs in and about Hausa on the H-Hausa list.* These are:

• Bahaushe Mai Ban Haushi! http://www.ibrahim-sheme.blogspot.com/
• Sha'irai (Hausa Poets) http://www.shairai.blogspot.com/
• Nishad'in Hululu (Hausa Popular Culture) http://arewanci.blogspot.com/

There was a brief discussion on H-Hausa about the non-use of the hooked letters for implosive consonants (and ejective, which for specialists is I think the term for ƙ ).

Don Osborn
Bisharat.net
PanAfriL10n.org

*The H-Hausa list is at http://www.h-net.org/~hausa/

FYI, the Teaching and Learning with Technology site of the Pennsylvania State University has a page on "Computing With Hausa" at http://tlt.its.psu.edu/suggestions/interna...language/hausa.html

(there is an error: it calls the implosive d "implosive g"; also it does not mention the Nigerian usage of 'y for the ƴ. Also, technically I think the ƙ and ƴ or 'y are "ejective" and not "implosive")

Don Osborn
Bisharat.net
PanAfriL10n.org

The "Hausa Online" site at http://hausaonline.wordpress.com/ has some interesting material, including links to other sites at http://hausaonline.wordpress.com/links/

Don

FYI, there are 5 countries outside of Africa that have shortwave broadcasts in Hausa and webpages in Hausa concerning their broadcasts and general news:

• Deutschland (RDW) http://www2.dw-world.de/hausa
• France (RFI) http://www.rfi.fr/langues/statiques/rfi_hausa.asp
• United Kingdom (BBC) http://www.bbc.co.uk/hausa/
• United States (VOA) http://www.voanews.com/hausa/
• Zhongguo (CRI) http://hausa.cri.cn/

(Note that the text in all cases is in ASCII, not Hausa boko as used in Nigeria and Niger.)

Edited by author 09-20-2007 04:51 PM
Last year, a new language resources site began called "African Language Technology" or "AfLaT" at http://www.aflat.org/ . One of the utilities they had available was one for the Gikuyu language that would insert the tilde mark (~) over appropriate vowels (some i's and u's) in text that a user copied in or typed in - http://aflat.org/?q=node/8

I thought this could be adapted as a utility that would replace plain b, d, k (and y in Niger) with the hooked letters ɓ, ɗ, ƙ and ƴ in the appropriate places. Sort of like a spellchecker, but limited to correcting the orthography.

Such a utility for Hausa could be developed for use in two ways:

1. Similar to the Gikuyu utility, allowing a user to plunk in a good ASCII version of a Hausa text and get the correct boko version. This could be useful for creation of documents or creating or upgrading web content or blogs in Hausa
2. Also allow a user to enter the URL of a Hausa webpage (such as the blogs in /m104 or radio pages in /m107 ) and get a corrected version of the page in Hausa boko to look at. This would be very similar to what you can do at online translation sites for other languages like Systranet and Babelfish, but much simpler technically to set up.

In a way, such a utility would make expanded use of Hausa much easier, on both the user side and the content creation side.

Any interest, feedback?

Don Osborn
Bisharat.net
PanAfriL10n.org

Hi Don,

I can't comment on the feasibility, although I assume that it comes down to how many confusable words would be present in Hausa.

Such a tool if feasible would be useful for converting existing data into the standard orthography, and would allow the conversion of a large amount of material.

But I'd be reluctant to use it as a tool for creating new content.

The way I see it, such a tool would be a stop gap measure. A more important issue is the creation of standards compliant web sites and web services in Hausa (using the correct orthography).

Hausa doesn't need smart fonts, it doesn't use complex rendering like numerous African languages, and form the point of view of languages in developing countries in other parts of the world processing Hausa is simple. From the point of view of fonts and input systems, Hausa is an easy language to accommodate on computers.

The question is how do Hausa web developers move forward? And what are the currently blockages? Since the blockages to developing web pages and displaying websites in Hausa aren't technically based?

Andj.

Hi Andrew, Belated thanks for this feedback /m109 . You raise some interesting points. In principle, as you note, getting Hausa Boko right on computers and the web is not at all complicated technically.

A tool for correcting ASCII versions of Hausa would be useful. Using that for new content when users could just compose with the extended characters (the hooked letters) would be a fallback. For the latter, maybe the hurdle is the lack of a keyboard standard that is widely known and available.

Don

I haven't had the chance yet to look at the following, but pass it on now. (Thanks to Rui Correia for forwarding)

---------- Forwarded message ----------
From: google.public.translators group <noreply@googlegroups.com>
Date: 4 Nov 2007 09:02
 
* Google in Hausa help! - 1 new
  http://groups.google.com/group/google.publ.../t/08f486863f2ab46b
 
-----------------------------
I tried myself to translate Google in Hausa but still a lot work to do, so
please do something if you think you are concerned. Thank you very much. - Sat
3 Nov 2007 20:53
1 message, 1 author
http://groups.google.com/group/google.publ.../t/08f486863f2ab46b

Where I can find good quality films?
Can anyone help me?

Edited by author 11-17-2007 11:59 AM
Re /m112, if you are looking for quality films in Hausa, you might try asking on the Finafinan_Hausa list at http://movies.groups.yahoo.com/group/Finafinan_Hausa/

The H-Hausa list at http://www.h-net.org/~hausa/ is also an excellent place for general information and might provide answers.

This forum, Hausa charsets and keyboards, focuses on Hausa language on computers and the internet (and mobile devices). There are some other links of possible interest in Hausa mentioned below in /m58.

Hope this helps.

Several Nokia mobile phone models apparently have text menu support in Hausa. Does anyone have any experience with using this? How good is it?

See http://www.panafril10n.org/wikidoc/pmwiki.php/PanAfrLoc/Nokia for a list of models.

TIA for any info.

Don Osborn
Bisharat.net
PanAfriL10n.org

Edited by author 11-23-2007 01:00 PM
During the last four weeks there has been a discussion about keyboard layouts on the a12n-collaboration list. The main topic is plans by the One Laptop Per Child project for a multilingual keyboard for its "XO" laptop. The keyboard layout would be intended to support languages of Nigeria (including Hausa) or the whole of West Africa.

If you are interested, see http://lists.kabissa.org/lists/archives/pu...a12n-collaboration/ (Note- this list is not an official OLPC forum)

The OLPC layouts are shown at:
http://wiki.laptop.org/go/Image:WAfrica-Alt-1.png
&
http://wiki.laptop.org/go/OLPC_Nigeria_Keyboard

See also /m96

Don

very interesting, but I don't agree with you
Idetrorce

Re /m116, thanks for your comment but please be more specific about what you find fault with. Thanks in advance. Don

The One Laptop Per Child project (see /m96 & /m115) has a page for people who want to work on localization at http://wiki.laptop.org/go/Pootle#Sign-up . For languages like Hausa not yet in the table, it looks like you will have to add appropriate rows in order to enter your name.

Note link to their page on Hausa.

Don Osborn
Bisharat.net
PanAfriL10n.org

A reply to a blog posting I wrote on machine translation (MT) at http://donosborn.org/blog/2008/06/30/parad...achine-translation/ mentioned that "Language Weaver (MT company) has a Hausa-English engine..." I have been trying to track down information on that, and any help would be of interest.

I did find a notice from Language Weaver last September that I was unaware of (in an old ad on http://www.translationdirectory.com/transl...ranslation_jobs.php , where they did not delete the it as requested since the job is over):

Source language: Hausa
Target language: English

Details of the project: 100,000 words of articles in Hausa from international news sources (BBC World Svc, Deutsche Welle, VOA, etc) as well as national news sites (Gaskiya, Nigerian Newsday.) We are willing to divide the job over several translators if you want to do only a portion of the work.

We will pay 10 cents per source word. Deadline for completed translation: Oct 11.

We need a simple, direct translation without localization. The results will be used as a reference for testing the performance of an automated translation system.

Special requirements to the applicants: To determine the quality of your work, we will provide a 9-sentences sample article for you to translate and return to us.

Deadline for applying: 9/21/07

Delete this ad from the site soon after the deadline for applying

Barbara Blankenship
Language Weaver, Inc
USA

Language Weaver's site is http://www.languageweaver.com

Don Osborn
Bisharat.net

The "Hausa Online" blog has a posting on using the abcTajpu extension in Firefox to be able to type the hooked letters. See: http://hausaonline.wordpress.com/2008/06/2...d-letters-in-hausa/

Microsoft's Local Language Program is working to localize Windows Vista in Hausa.

See this story on 'Gbenga Sesan's Oro blog:
"Vista, MS Office in Hausa, Igbo, Yoruba"
Friday, December 12th, 2008
http://www.gbengasesan.com/blog/?p=307

The translation of terminology in Hausa is available for review at:
http://www.pinigeria.org/microsoft/hausaglossary.pdf

You can check out more information about Windows Vista Language Support in Hausa and other languages at:
http://www.microsoft.com/globaldev/vista/V...nguage_Support.mspx

Don Osborn
Bisharat.net

Edited by author 05-18-2009 10:28 PM
The Google web search engine now has a Hausa interface. See http://www.google.com/intl/ha/ . It appears to be in an ASCIIfied version of the Boko orthography.

(Not sure if this is related to the effort mentioned in /m111 )

Edited by author 06-12-2009 11:31 AM
Excerpt from a message posted by Manuela Noske on A12n-collaboration:

FYI that Microsoft Corporation has released the Windows Vista Language Interface Packs for Hausa, Igbo and Yoruba on June 2nd. The LIPs are available at the locations below. They are free downloads that install on any PC that runs Windows Vista with either SP1 or SP2.

Hausa Details Page and .mlc: http://www.microsoft.com/downloads/details...7f5b&displaylang=ha

http://download.microsoft.com/download/6/A.../LIP_ha-Latn-NG.mlc

...

(See also /m126 on this message board.)

I received a request for Unicode text of African languages written in Latin script with extended characters. The purpose is testing some new fonts. Does anyone have access to such digital text of Hausa?

Don Osborn
Bisharat.net

FYI from the Hausa Online blog at http://hausaonline.wordpress.com/2009/07/2...kipedia-is-growing/ :

Hausa Wikipedia is growing

Posted by hausaonline on Wednesday, July 22, 2009

About a year ago, I warned in a blog post that Hausa Wikipedia [ http://ha.wikipedia.org/ ] could soon be closed, due to lack of activity. I don’t know whether it was because of this post or due to some other factors, but since then, Hausa wikipedia is growing fast. Have a look, and maybe you feel like contributing something yourself or editing (please, don’t vandalize!) the articles that have been written.

If you would like to contribute something, but don’t know how this works, here are some guidelines [see the blog posting for links]:
* Contributing to Wikipedia
* Policies and guidelines
* Tutorial

You could start with some topic you are familiar with or translate an article from another language.

A good point to start from could be some online news articles. Some old Hausa school books like "Ikon Allah" would be good source material, too. But make sure you do not just copy what is written there, in order to avoid copyright issues.

For background to /m130 re last year's proposal to close Hausa Wikipedia, see http://tech.groups.yahoo.com/group/afrophonewikis/message/454

Siebrand Mazeland is asking for help translating Mediawiki* into Hausa. Mediawiki is the software used for Wikipedia, as well as being available for free to anyone wanting to set up a wiki on a website. Having it available in Hausa (as well as other African languages) would help make it more useful to Hausaphones (etc.).

See http://translatewiki.net/wiki/User_talk:Siebrand where you can add a message indicating your interest in helping. You can also post here.

Don Osborn
Bisharat.net

* http://www.mediawiki.org/wiki/

The ANLoc* locales project just announced that the Hausa locale for Nigeria is complete. Hausa for Niger is apparently already complete. See http://www.it46.se/afrigen/statistics.php (which has summaries of work on many languages). This project is being coordinated by Martin Benjamin.

Locales are files of data for computers and software to adapt presentations for different language preferences. They are generally defined by language and country.

Don Osborn
Bisharat.net

* http://www.africanlocalisation.net/