[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

My next Arabeyes projects...

Now that Akka 1.0 is almost being released, a few thoughts about what I
might want to work on, you guys tell me what you think is the most
important, and bring your suggestions as well, as to what I might be
most useful...

- Akka 2.0 of course,
aside from unicode support, I don't see what addition I could put. I
don't think unicode support is worth the effort though, it is heavily
system and architecture dependent at the console level, and even the
Linux support is really awkward, so my effort would produce a lot more
if invested elsewhere... Something I would want to do for my fun is to
port Akka as a GNU/Hurd server. But this is really for fun, coz the
audience for now is very limited. Akka-izing xterm, gnome-term, kvt,

- Teach Your Kids Arabix (TYKA)
the reason I keep bringing in kids is because I know their situation
concerning Arabic support in my real life. In Tunisia like in most Arab
countries computerization is limited. It happens that many (7-8 on a
nearly daily basis during summer) of my little neighbors often come at
my home to play with my computers. Most of them are not able to read
Latin characters, or didn't reach a level necessary to have Latin words
make any sense yet. In fact, that, and my teaching experience in
database development to Palestinians, which revealed deep flaws in the
understanding and learning process due to the total absence of mnemonics
in the mother language, is what pushed me to invest myself in
Arabization. So you can guess how important this project could be for
So what this project would be about?
1) First, tuning unix in order to make it as Latin free as possible,
including some programming and formatting languages for kids.
2) Second, making it kids-friendly. For that, the SEUL project (didn't
check it for a couple of years at least, so I'll have to review it)
exists, and I suggest we coordinate/do the same through Arabization.
3) Supervise projects that would install computer labs in schools
throughout the Arab World. I am ready to invest my personal money if
needed to move around on this one. All we need are enough donations to
equip such a lab, and local volunteers to help. This is a work I've done
a few times, and there are more schools than you could imagine which are
ready to cooperate. To give you an idea, we could buy 20 used units for
an average school at 150$ each (3000$ total) in an excellent state in
Madrid, ship them for nearly free (it's easy to negotiate that when
you're not doing it for profit), and install a unix based network where
kids can learn all the basics of computerize without the hinderance of
languages. I have the logistics necessary for most parts, what I would
need are the hardware funding and coordination with the volunteers on
the ground. Note that I can receive US tax exempting donations for such
projects in my current NGO, though I would rather create a different NGO
with you three (though I'd have to check what the legalese is in order
to be a tax exempted NGO in the US), or better look for a different one
that will be closer to these goals. Anyway, for a start, 1) and 2) are
immediately feasible.

- AraBasic
As part of the previous TYKA effort, an Arabic programming language for

- AraML
an XML based formatting language for kids in Arabic, part of the TYKA

- Unix Tools Arabization
Wrapping and localizing tools such as ls, man, cp, rm, etc... as well as
a minimal shell in Arabic as part of the TYKA effort.

- Debian Website Translation
Yes, GNOME is more important of course, but this is a project I started
a few years ago when very few people even cared about Arabic in the Unix
world. The goal was that, through a distribution as widespread as
Debian, we would attract more interest to Arabization. It is the lack of
manpower that killed the project at the time. Though we have a good hand
of translators today, we still want more, and we definitely have a cruel
shortage of developers. Please check the http://www.debian.org website
and look at the bottom of the page to see how the l10n is done, and how
so many languages are already there as part of such an effort.
Internally, it is through a CVS repository and a nightly build.

- Reverse Bidi Specs and implementation in Fribidi or in a standalone
This is REQUIRED for reading monodirectional Arabic text in bidi
applications. I will have to go through testing ICU, using it as an
inspiration for specifying a clean Reverse Bidi algorithm, and suggest
an implementation as part of Fribidi or as a standalone library.

There, you have the ideas more or less that I originally had in
priority... If you think my developing effort would better be spent
elsewhere, just tell me what other project you think is important, or
tell me what you think is the most important in what I mentionned.

Isam, Mohamed, still waiting for your keyboard layouts (standard pc105
US QWERTY?), I will assume you are using consolechars since it's the
font setter that has been used for some time now. Don't worry, it'll
only take you a few minutes to test...