Re: Arabic on Zaurus/Ipaq

[Abdulhaq Lynch <al-arabeyes at alinsyria dot fsnet dot co dot uk>]

On Saturday 26 February 2005 07:52, Nadim Shaikli wrote:
> --- Abdulhaq Lynch <al-arabeyes at alinsyria dot fsnet dot co dot uk> wrote:
> > On Friday 25 February 2005 15:28, Arafat Medini wrote:
> > > Nice stuff! Would you want to import this to our CVS servers? And
> > > continue devon it? Someone of core, what do you think ? I think
> > > importing such an interesting experiment would be really sweet!
> >
> > It would certainly be helpful to put it on the CVS servers and also I'd
> > like to put the OpenEmbedded font package somewhere too (easily
> > ftp-able).
>
> If you are willing to maintain it, by all means - simply send an email to
> 'contact' and it will get taken care of at 'core's next meeting.  The point
> that needs to be made is that we really need to make sure that anything we
> import has an owner who is responsible for it (dead or fermenting projects
> simply don't appeal to anyone unless they are completely done).  We have a
> serious issue with keeping up with our various projects and making sure
> that people upkeep 'em.
>

I agree that maintenance is vital. KPrayertime is very stable now so I don't 
think maintenance of this new work will be too burdensome timewise.

Well the main thing is hosting the font packages for now - there is very 
little code on the Qt/E side (a couple of small patches for tashkeel - the 
hard part was finding out where to put the patches). 

I think it might be useful to host the vocabulary program (which is in OK 
working order on normal PCs). In order to run the arabic vocabulary program 
on the zaurus I ported it from Postgres to SQLite. It is now much more 
'installable' on other people's machines so I am fairly sure there would be 
interest in this program (running on a normal PC) with arabic learners. In 
fact the program was originally written for me to create an arabic vocabulary 
book but I soon realised that it could be useful for revision also. Currently 
there are about 350 words in it and I am aiming for 5001 of the most common 
words.

> >
> > I had a look at FriBidi for the bidi side and it seems to do an awful lot
> > of things that I would have thought would be better left in the parent
> > application - such as mapping mouse coordinates to characters, and
> > encoding conversions. Is there a cut-down version that just reorders
> > unicode characters? Memory space is at a very big premium for handhelds.
>
> MiniBidi [1] is an option (it will soon ISA get shaping embedded as well).
> As for a stand-alone shaping/tashkeel widget you can take a look at this
> perl script [2] which was later ported to C [3].

Minibidi looks very interesting.

>
> BTW: you should most definitely get your work adopted by the upstream
> project (Qt/E).  Any patches/work/efforts we create are as good as the
> current release and to make sure that your hard work is not wasted for
> future releases it is best to get the application developers to accept your
> work (even if it is #ifdef'ed out).

Yes, I am currently trying to do that but it may take some time. I also agree 
about the #ifdefs. One obstacle is that the OpenEmbedded development 
environment is a bit of a beast and I am still getting used to it and of 
course the maintainers have to get used to me.

wassalaam
abdulhaq