From: Don Osborn (firstname.lastname@example.org)
Date: Tue May 13 2003 - 15:10:56 EDT
Thank you, John, this is interesting and I appreciate your taking the time
to go into such detail.
Before responding, I think it's helpful to keep separate the "African
Reference Alphabet" (ARA) of which I posted the images from the Niamey
conference proceedings on the Bisharat site (ref. below), and the "Niamey
keyboard" that somehow evolved from that, which had, as I understand it,
only lower case letters including many modified characters. With that in
mind, I respond in text...
----- Original Message -----
From: "John Clews" <email@example.com>
To: <firstname.lastname@example.org>; <email@example.com>
Sent: Tuesday, May 13, 2003 10:41 AM
Subject: ISO 6438, the Niamey keyboard, and ISO/TC46/SC4/WG1
> Dear Don
> As a librarian involved in standards for 15 years, I can tell you the
> following (see embedded comments).
> In message <00e801c318cd$dd197020$90e4fbc1@gktg001> "Don Osborn" writes,
> Re: Sources for characters in ISO 6438:
> > Does anyone have any leads to sources of information on how the ISO 6438
> > "African coded character set for bibliographic information interchange"
> > developed? A 1979 DIN document (without explanatory text), "Coded
> > Set for African Languages"
> > which evidently is a forerunner to ISO 6438 (a 1998 version of which is
> > http://anubis.dkuug.dk/jtc1/sc2/open/02n3129.pdf)* differs in several
> > from usage illustrated in the "African Reference Alphabet"
> > (http://www.bisharat.net/Documents/Niamey78annex.htm) published after
> > 1978 Niamey "meeting of experts on the transcription & harmonization of
> > African languages," which raises the question.
> I am 100% sure that there was no contact between those involved in
> the Niamey project (mainly linguists and language planners) and those
> librarians (mainly from national libraries) who participated in the
> development of work towards ISO 6438. In my view, developments in
> ISO/TC46/SC4/WG1 were sometimes very ad hoc compared to
> what was ISO/TC97/SC2 (which became ISO/IEC JTC1/SC2).
Interesting - confirms my impression.
> I have a question: how much difference in repertoire was there
> between the Niamey keyboard layout, and that of ISO 6438?
> Also are both now satisfactorily included in Unicode and
> ISO/IEC 10646? I assume they both are by now. Certainly the
> ISO 6438 repertoire will be.
ISO 6438 is (so far as I've checked). A character in the Niamey ARA, and
forms of characters in the latter when they differed from ISO 6438 are not
in ISO 10646/Unicode. Peter Constable brought some of these up on this page:
... and another came up on this page:
An additional difference was in the capital esh - sigma form in ISO 6438 and
large version of the lower case in ARA.
This not to say that the ARA was necessarily more "correct" although in at
least a couple of instances one could make that case. On the other hand ISO
6438 has two or three characters omitted by ARA. As I mentioned, these
orthographies have been changing (the last revision of orthograhies in
Niger, for instance, was in 1999, though this did not involve new
> I am also 100% sure that there were no implementations in libraries
> of ISO 6438 or any "lookalikes" mentioned above, as coded character
> sets for bibliographic information interchange.
> Unesco was involved in both fields in a sense, in that it took a
> active role in promoting the development and availability of
> automated information systems in scientific special libraries in
> particular (then still somewhat new), and standards for these. At the
> same time, more directly, Unesco (probably via another division in
> Unesco) sponsored the Niamey conference, which led to the Niamey
> keyboard layout.
> The repertoire in ISO 6438 would have been reasonably well researched
> however, among the mainly national systems librarians involved, based
> on repertoires of African letters in the collections of national
> libraries, and with advice from the area specialists and language
> specialists in the national libraries concerned.
Interesting, though, if despite all that, the conclusions of expert meetings
also sponsored by UNESCO may not have been accounted for.
[ . . . ]
> Going back to ISO 6438, many of the national systems librarians
> involved are long retired, though I could give you some names
> off-list, if you needed to research this further.
Please do. It's not my intention to research this deeply, but a better
sense of the history I always find helpful (if I can find the time).
> In my view, ISO 6438 is perhaps best regarded as history, as indeed
> are most of the 7-bit coded character set standards for bibliographic
> information interchange which were introduced by escape sequences,
> which were developed by ISO/TC46/SC4/WG1.
[ . . . ]
> Finally Don (if you've made it this far to the bottom, past all the
> character set history, without falling asleep), the Niamey keyboard
> layout was developed by David Dalby, as a consultant to Unesco, at
> the time of the conference.
Of course know of David Dalby and also of his role in Niamey a quarter
century ago. I've tried contacting him previously and next time I'll ask
specifically about the keyboard. It would be interesting to see what it
> Do you have any evidence of the use of the Niamey keyboard layout in
> practice, on keyboards developed by manufacturers of typewriters or
> computer keyboards? I ust don't have much detail on that, either way,
> so I'd be glad of some further information.
I have never actually seen the Niamey keyboard (and I'm actually in Niamey),
and indeed it never seems to have been widely produced or used, if at all
(?). On the other hand, IBM by the 80s had at least one Selectric type ball
with some of the modified characters upper and lower case (I used one for
Fula in 1987). Not sure what else was done or what references they were
working from (i.e., if there might have been talk of a Niamey type ball).
Thanks again and all the best!
Don Osborn dzo_@_bisharat.net
*Bisharat! A language, technology & development initiative
*Bisharat! Initiative langues - technologie - développement
This archive was generated by hypermail 2.1.5 : Tue May 13 2003 - 17:11:22 EDT