The ArabEyes project does just what it sounds like it ought to do: Builds support for Arabic in the Unix/Linux environment. Yesterday, the project celebrated its 2nd anniversary with an "Evolution Party" that started at 2. p.m. UTC on irc.freenode.net's #arabeyes channel.
'Party' in this case really meant 'work session.' And not all participants were from Arab-speaking countries. To show you what was discussed, instead of trying to describe it all, we're presenting a lightly edited transcript of the session. Remember, this was IRC and off-the-cuff, and although the chat was in English, most of the participants are not native English speakers, so please make allowances for both the nature of IRC and language difficulties before you criticize anyone's spelling or grammar.
<mostafah> Hello all
<walid|afk> ya hala
<Stanislav> Hi all!
* walid|afk is now known as wal
<yuyu> party in 2minutes.........................
* minami has joined #arabeyes
* isam_is_dead is now known as isam
<isam> Stanislav: how are you going ? great to have you here
<Stanislav> Hi there!
<Stanislav> I'm quite busy at work now, I'm glad to be here.
<isam> Stanislav: your work in Kababel, and responsivness with Arabic
issue is something that we really value here
<yuyu> hi all
<yuyu> hmm, welcome to arabeyes evolution party !
<yuyu> we'll start by a quicjkie presentation of everybody
<Stanislav> Isam: Thanks! I'm not an expert in Arabic, but I want
KBabel to be the best tool available
* karaki has joined #arabeyes
<yuyu> Stanislav Visnovsky
<Stanislav> yuyu: yes?
<yuyu> Stanislav ?
<yuyu> ok - right
<isam> I am Isam Bayazidi, a member in Arabeyes team, was coordinator
for the translation team some time ago, and currently trying-to-be
active member in Arabeyes :)
* roozbeh has joined #arabeyes
<Stanislav> I'm Stanislav Visnovsky, maintainer of KBabel. I see a lot
of familiar names here :-))
<yuyu> minami, MINAMI Hirokazu , mlterm developer ?
* ssato has joined #arabeyes
<ssato> Hi everyone
<yuyu> ssato hi
<minami> yes. Hi all
<karaki> Hi, I am Ken Araki, a developer of mlterm:D
<yuyu> ok, I am Youcef Rabah Rahal - a memeber of the core team of
* roozbeh has quit IRC (Client Quit
<ninjadude> ninjadude, Ross Golder, gtranslator maintainer
* roozbeh has joined #arabeyes
<mostafah> roozbeh: Mr. Pournader. Salam?
<yuyu> welcome everyone ;) we would like to make this party an
occasion to talk about technical details/issues of arabization/unicod
<roozbeh> mostafah: salam.
<mostafah> roozbeh: I've been following your efforts on the standard
Hijri calendar as well as the timezone contribution to the Olson
<roozbeh> mostafah: nice to meet you. where are you from?
* kamal is an mlterm and fribidi user. thank you guys for doing a
great job and making it possible.
<mostafah> roozbeh: We even had some contact a while ago about about
Mozilla Calendar, remember?
<yuyu> so, who would like to start ? depending on everyone schedule ?
<yuyu> btw, other core-team members will be here in a few minues
<yuyu> kuru: Mohamed Elzubeir
<yuyu> nadim: Nadim Shaikli
<roozbeh> mostafah: oeone?
<yuyu> and anmar: Anmar Oueja
<mostafah> roozbeh:Yes. You know Masoud there?
<yuyu> isam has been for a loong time a core member of arabeyes
<yuyu> roozbeh, nice to meet you and talk with you in real ;)
<yuyu> roozbeh, where is behdad ?
<yuyu> roozbeh, will he come ?
<roozbeh> yuyu: behdad's currently in Windows mode, with no IRC
client. I'm chatting with him on YM.
<isam> Stanislav: if I may ask, where is Kbabel heading ?
<isam> Stanislav: will we see the DB2 dependency go off along the road
* kubota has joined #arabeyes
<yuyu> kubota: hello !
<roozbeh> mostafah: yes, masoud has been among the guys who helped me
learn linux six or seven years ago.
<yuyu> kubota: you are tomohiro ? ;)
<kubota> In Japan, it is about 23:00 but I am now working....
<yuyu> kubota, mlterm developer ?
<yuyu> kubota, welcome to our party
<isam> ninjadude: although I really like the look n feel of
gtranslator, the lack of rough translation, and soring strings in
the DB makes gtranslator not efficiant for a full time translator
<kubota> well, a kind of.
<Stanislav> isam: KBabel in KDE 3.1.1 and newer uses DB 4, KDE 3.1.3
supports even 4.1 (they changed the interface again).
<isam> ninjadude: the suggested translation thing is important
<yuyu> btw, if some of you want to know about arabeyes project and its
history, please take a look at www.arabeyes.org/history.php
<kubota> I am responsible to mlconfig. I would like to hear about
GTK2 migration of mlconfig.
<ninjadude> isam: cool - if there's a bugzilla report about it, i'll
get round to looking at it :)
<ninjadude> to be honest, i don't even use gtranslator a great deal
<Stanislav> isam: KBabel reaches its limits as a translation editor,
so the planned features are project support and also better rough
<ninjadude> i just don't have time - i'm just doing what i can to keep
it maintained at the moment
<Stanislav> Do you know FreeCATS project?
<yuyu> ninjadude, that would a real plus ;)
<roozbeh> isam: gtranslator is dead, isn't it?
<ninjadude> in about two months time, i'll be able to work on it
full-time for a couple of weeks, and get all the bugs/feature
requests cleared down
<Stanislav> gtranslator is alive again AFAIK
<ninjadude> roozbeh: no, I'm looking after it now, and Fatih is
keeping his hand in a bit too
<isam> roozbeh: ninjadude is the gtranslator maintainer .. ask him :)
<yuyu> ninjadude, we will report those bugs then ;)
<roozbeh> ninjadude: my ignorance then, sorry.
<ninjadude> please do, i can't promise a quick reaction force, but i
can promise an eventual fix
<yuyu> ninjadude, but the rogh translation is something that should be
thought as a priority IMHO
<isam> Stanislav: what would be great to see in Kbabel is a PREVIEW
feature.. to see the application and dialog in the freshly
translated file .. but I guess technicaly that is difficult
<Stanislav> isam: a bit, at least in KDE. AFAIK you cannot specify
where is the .mo file located
* gloonie has joined #arabeyes
<yuyu> gloonie, hi, welcome to our party
<yuyu> gloonie, you are ?
<isam> Stanislav: I guess the next step for Kbabel is to be more then
just a personal program .. to have it possible that translation
teams work together on Kbabel.. that require the use of a Database
server instead of the DB
<gloonie> Hi, Just wanted to wish you all congratulations on excellent
work and celebrate!
<yuyu> gloonie, thank you
<isam> gloonie :) thank you, welcome in ..
<isam> gloonie: are you gloonie at earthlink ?
<gloonie> yep, that's me!
* Uniball has joined #arabeyes
<yuyu> Uniball, salam
<gloonie> unibal, salam
<Stanislav> isam: Yes, there are some solutions for this already.
KBabel should be a client for such servers then.
<yuyu> Uniball is the author of the katoob editor http://www.arabeyes.o
<Uniball> yuyu/ gloonie/ kubota salam
<isam> gloonie: I am the one that google new people when they arrive :)
<roozbeh> yuyu is the host. ;)
* behdad has joined #arabeyes
<Stanislav> isam: As to translation memory, the FreeCATS project seem
to be quite interesting in this regard, since these are professional
translators and developers of CAT tools.
<yuyu> behdad, welcome welcome
<gloonie> isam, good idea. I usually ask stupid questions, like the
meaning of the Logo. Working on a Postgresql-based rule engine for
<isam> FreeCATS.. I will look into that indeed
<roozbeh> to correct, yuyu is the microphone!
<isam> gloonie: I know that already.. google gave me the postgress,
and VIM postings .. :)
<behdad> hi all.
<isam> hi behdad
<gloonie> sorry, folks, plumber just arrived, must leave!
* gloonie has quit IRC ("Client exiting"
<behdad> well, what's going on?
<Stanislav> ninjadude: Do you plan any support for external
dictionaries and translation memory in GTranslator?
<isam> behdad: we were asking our selves where is fribidi heading ...
will it stay in the maintainance loop for ever, or we will see
something new soon :)
* Arabic Shaping
<kubota> I have a question on FriBidi. There seem several similar
<kubota> Is FriBidi the de-facto standard in Arab world?
<roozbeh> isam: behdad and me are supposed to deliver some fribidi
extensions by early April due to a contract. And that includes
<behdad> Well, actually I'm not so lazy. We need the time to see what
the Unicode cons. is going to decide about that.\
<behdad> Well, tell me what do you have in mind.
<behdad> The only other implementation comparable to FriBidi is ICU,
which is Huge.
<roozbeh> kubota: several similar implementations of Unicode bidi?
<roozbeh> kubota: i know of only two which are both compliant and free
* Osax has joined #Arabeyes
<isam> Stanislav: https://savannah.nongnu.org/projects/freecats/
<Stanislav> isam: yes, that's it.
<isam> Stanislav: looks that it will need a long time to be something
<isam> Stanislav: still pre-alpha
<kubota> roozbeh: for example, http://crl.nmsu.edu/~mleisher/ucdata.htm
l shows several implementations which have small differences one
<ninjadude> Stanislav: yes, those two are in the pipeline
<Stanislav> isam: I don't really know. They miss developers. But the
knowledge is there
<isam> Stanislav: there is no translation memory applications or
systems that are Open Source, right ? besides FReeCATS
<Stanislav> isam: CATs? No, I'm not aware of any
* kuru has joined #arabeyes
<kuru> salam all
<Stanislav> ninjadude: Maybe we should agree on some reasonable
interoperability then, not to reinvent the wheel
<isam> Salam Kuru
<behdad> kubota: most of them are not standalone libraries, like
Java, IE, KDE, ...
<Osax> salam kuru
<kuru> behdad: hey, you made it ;)
<Uniball> kuru: salam
<behdad> kuru: yep. but I should leave and be back in a few hours.
<kuru> Uniball: salam -- I'm not going to ask about the duali deb pkgs
<isam> bb in seconds
* isam has quit IRC ("ChatZilla 0.8.31 [Mozilla rv:1.4/20030727]"
<kubota> behdad: I see. However, Markus Kuhn strongly opposed to
support bidi in xterm because there are no simple standard to handle
bidi on terminal....
<kuru> kuru: Markus opposes a lot of things.
<roozbeh> kubota: that NMSU page is very outdated first of all, and
then then only new thing it mentions is PGBA, which is not unicode
* isam has joined #arabeyes
<behdad> kubota: I know. This is why I stopped working on that too.
<Uniball> kuru: i'm not a DD, I have to find a sponsor, though i
didn't try yet ;)
<Osax> glad to see all of you here.. though have to go to my class..
:) wish you a very nice time and see you later!
* isam has left #arabeyes ("Client exiting"
* isam has joined #arabeyes
<kuru> kubota, you are one of the mlterm guys, no?
* Osax has quit IRC ("Client exiting"
<kuru> kubota: what are your thoughts on cmd-line interactions with
<roozbeh> kubota: Markus Kuhn is probably very right. Have you tried
to read and implement that ECMA standard on bidi terminals? That's
the only standard on the matter.
<kubota> kuru: yes, I am now responsible to mlconfig.
<Stanislav> isam: I've found out how to do a preview for glibc apps,
now I need to convince Stephan Kulow to do it for KDE.
<behdad> kubota: Oh, I just remembered who you are :D
<kubota> kuru; mlterm's policy is being practical, so it supports
bidi. Is that right? >karaki
<kuru> kubota: well, t the cmd-line level, plain bidi can be
<isam> Stanislav: well, I guess it is a system thing..
<kuru> kubota: for example: I want to do $ mv file1 file2 .. if both
file1 and file2 names are in RTL chars, it will look like: $ mv
<kubota> kuru: maybe... How Arab people think about this?
<Stanislav> isam: you can set LANGUAGE to contain a path to your own
<karaki> kubota: yes.
<kuru> kubota: and I'm not sure if this is oething that shells should
worry about or if it's something that mlterm shoul worry about.
<roozbeh> kuru: so, you should really go hack bash.
<kuru> kubota: if the cmd 'mv' was an Arabic word, then that's fine..
but since it's not.. it looks wrong.
* Mohammed has joined #arabeyes
<kuru> roozbeh: you think it's a bash thing?
<karaki> kuru: Nadim also points such a problem to me.
< mv file1 file2
<roozbeh> kuru: i think if you want an arabic terminal, you should go
hack almost everything.
<kuru> karaki: yes, it's something nadim and I have talked about
<behdad> I think the most important thing to hack in terminal is
ncurses bidi support.
<behdad> Hi Mohammad
<kuru> roozbeh: you know we don't want hacks, we want permanent
<behdad> I never found out why no one is gonna do that.
<kuru> behdad: cause ncurses is way too hairy? ;)
<behdad> kuru: hack means to change to be as you want it to be!
<behdad> kuru: but is the right way.
<karaki> It is very difficult to solve by not only terminal emulators
but console applications' support.
<kuru> behdad: bleh.. I was just joking.
<behdad> kuru: ncurses is much more easier that yet another terminal
<roozbeh> kuru: permanent solutions are called standards. go read the
ECMA standard and discuss it to the minor details. then go write a
new standard if you don't like it and submit it as an IETF
Internet-Draft. that's the way to do it.
<kubota> roozbeh: Somebody doing it?
<roozbeh> kuru: everything is too hairy. proper-bidi-so-the-user-feels-
at-home is fairy itself. it is that makes bidi support in everything
<kuru> roozbeh: you're probably right.. but the user can feel at home
by adaptation, no?
<roozbeh> kubota: nobody's doing that. i know a few israeli experts
who were involved in development of that ECMA standard, but that's
<roozbeh> kuru: i'm not getting what you mean by adaptation.
<kuru> So you're saying this cmd-line issue is a bash/zsh/tcsh issue..
and nothing to be done on the terminal emulator's side?
<kuru> roozbeh: you walk into a room.. it's very humid.. you spend 2
days in there, and it feels natural.
<kuru> so this could end up like a crusade on all shell programs.
<isam> Stanislav: but there is certain complications with preview.. as
sometime one dialog have strings from different POs
<kuru> there's got to be an easier way.
<roozbeh> kuru: i'm saying that you can't trust every terminal to have
bidi support. also, you need some level of support at the
<behdad> kuru: this is the microsoft way.
<isam> Stanislav: for instance strings of kdelibs are everywhere
<kubota> Future terminal should support not only bidi but also Indic,
Thai, CJK, and so on, which makes it much difficult.
<kuru> kubota: agreed.
<roozbeh> kuru: you already need to do a crusade on many different
things, say, changing digits to the Hindi ones for the Asian Arab
<Stanislav> isam: yes, they are, you can only preview applications.
It's not possible to preview libraries reasonably.
<Stanislav> isam: but even trivial support would help a lot IMHO
<ninjadude> Stanislav: yes, we probably should work on sharing some
<ninjadude> or interoperability
<kuru> roozbeh: yes I realize that.. but I am specifically talking
about cmd-line renderings on terminal emulators.. and mlterm even
<kuru> I tend to think of mlterm as the 'good' example.
<roozbeh> kubota: or you can only ask the future terminal to support
level 1 ISO 10646 (which is legible to some degree, at least after
some "adaptation") and forward the user to graphical apps.
<Stanislav> Ninjadude: KBabel currently strongly depends on Qt, but
the dictionaries will be probably rewritten soon
<Stanislav> Ninjadude: Did you notice that gettext now contains a
library to read PO files?
<ninjadude> no, i hadn't noticed that
<roozbeh> kuru: i'm not opposed to bidi terminals, as Markus is, I
just want a proper specification. the problem of bidi terminals is
that nobody knows what it is.
<behdad> Stanislav: really?
<isam> Stanislav: using that lib will ease your string parsing work
<kuru> roozbeh: We're not in disagreement there.
<roozbeh> kuru: to me, mlterm's arabic is just a first try at the
subject. something to play with to see how horrible the goal really
<kubota> roozbeh: agree. I am afraid that, when a clear specification
will be established in future, mlterm will maybe become "illegal"
<behdad> Any volunteers for ncurses?
<behdad> Mohammad? you? ncurses bidi support.
<kuru> roozbeh: which is why I am interested in how to go about fixing
one of the few issues I have found in mlterm.
<roozbeh> kuru: you'll need, say, at least a dedicated month from
someone interested in both bidi and terminals to somehow approach
the bidi terminal specification problem. can you find that? i guess
<behdad> where is nadim?
<roozbeh> kuru: the issues will keep coming and coming, and you'll
finally encounter a case where there are two conflicting issues.
<kubota> kuru: standard should not be defined by a certain
implementation (like mlterm).
<roozbeh> kuru: you need to choose one of them, but how in hell?
<yuyu> behdad ,he'll be here in a few minutes
<roozbeh> kubota: you're more than right!
<kuru> ok, fine, so back to non-existing standards then ;)
<behdad> kuru: It is impossible to have a terminal that handles
whatever bidi you need on command line. so bash should do its own
<behdad> kuru: the difference is that with bidi terminal, bash needs
to just put some bidi marks, instead of reordering. which is a
great improvement indeed.
<kuru> I would be interested in playing with bash if I know that it's
worth the trouble.
<roozbeh> behdad: but it will somehow become a kuru-term. user
expectations from a bidi terminal will be different for each user.
<kubota> and bash has to know how terminal behaves.
<kuru> but roozbeh certainly doesn't make it sound palettable ;)
<Stanislav> isam, behdad: Yes, see gettext-po.h in gettext 0.12
<kuru> kubota: that's one of the reasons why I think the terminal
emulator should do the work here.
<behdad> roozbeh: I dont think so.
* Uniball has quit IRC (Read error: 60 (Operation timed out)
<behdad> roozbeh: an ltr after quotation marks, pdf before the
matching one, things like that.
<Stanislav> ninjadude: I'll contact you by e-mail, OK?
<kuru> behdad: exactly.
<roozbeh> behdad: well, i just had your first line when i was writing
that. i get what you mean now.
<roozbeh> behdad: i agree. but still we need old escape codes to turn
bidi on and off and such things.
<kubota> Speaking about ncurses, I am now discussing with Thomas
Dickey on doublewidth and multibyte characters. It is difficult
<roozbeh> behdad: who's going to define those, and fight the path in
IETF, or god forbid, ISO?
* tumbak has joined #arabeyes
* isam has quit IRC (Remote closed the connection
<behdad> hopefully we would not need that :D
<roozbeh> kubota: btw, update the width data with unicode 4.0 data
files, while you're at it.
* ChanServ sets mode: +o kuru
<roozbeh> behdad: we want sun and ms and others in the game, don't we?
* isam has joined #arabeyes
* kuru changes topic to 'arabeyes.org - Evolution Party - 14:00-02:00
<kubota> roozbeh: which width file do you mean?
<behdad> roozbeh: no!
<karaki> Congratulations on arabeye's 2nd anniversary:) Bye...
* karaki has quit IRC ("irchat exiting..."
* isam has quit IRC (Remote closed the connection
<behdad> roozbeh: putty is there.
<roozbeh> kubota: Unicode character widths. 0, 1, 2, stuff. wcwidth?
<roozbeh> behdad: putty's where?!
<Mohammed> congrats to all of you :-)
<kuru> putty does support utf-8, etc.
<kuru> just no bidi.
* Mohammed is now known as Mohammed_Yousif
<behdad> open source community
<kuru> horrible code too
* isam has joined #arabeyes
<kubota> roozbeh: well, mlterm may have its own width table.... (I
don't understand most part of mlterm source code.)
<roozbeh> behdad: i don't like the ghetto game of course. invent
something, don't publish the specs, use it only in arabeyes. doesn't
look to have much future.
<kuru> roozbeh: the putty project on arabeyes is dead afaik
<kuru> roozbeh: and it is never meant to stay in arabeyes only.
<behdad> roozbeh: why would we need MS support?
<behdad> roozbeh: for their dumb ternial?
<behdad> roozbeh: do publish the specs. spreed it around, kde,
gnome, xterm, ....
<roozbeh> behdad: why not IETF then? since they may have valid
<roozbeh> behdad: ms was just an example. but we need guys like sun
and ibm. they ship lots of terminals and stuff to the arab world.
global interoperablity and so.
<kuru> ibm's ixterm work is very itneresting.. though no one seems to
want to look under the hood there.
<behdad> roozbeh: because we need a practical spec.
<behdad> roozbeh: I'm not saying "not IETF". not its not needed.
<behdad> what happens if we start our own way in, say, linux-based
open source, say GPLed terminal emulators?
<behdad> roozbeh: what happens is that it gets into everywhere. this
is how linux was developed.
<roozbeh> behdad: you're right, but we're talking about specs, not
software. let's see where are the linux-only standards. how many do
you know? lanana? :)
<kamal> does any one know about an implementation based on ECMA ?
<roozbeh> behdad: if we want something that stays, we should go the
hard way. IETF, that is.
<behdad> kamal: we looked for it, never found.
<roozbeh> kamal: no. i don't even know anyone who has read it and is
still a coder.
<behdad> roozbeh: how? where are the experts?
* Zaki has joined #Arabeyes
<kamal> roozbeh: hmm interesting. I read it and later on lost my job.
now that you mention it , I think it may be related ;)
<behdad> got to leave now. see you all later (a few hours maybe).
<roozbeh> behdad: in ibm egypt, in sun, in ms, in cisco, in IETF in short. if you want to create a standard, IETF is actually the easiest way.
-->| kefah (~email@example.com) has joined #arabeyes
<kefah> Hello there.
<kuru> kefah: salam
<kuru> behdad: later
<kamal> behdad: the linux-open-source way is for someone that really needs a bidi terminal would implement one. if it is good , others will pick it up and use it
<behdad> roozbeh: so lets do it!
<behdad> I'm in
<--| behdad has left #arabeyes
<isam> kefah: Salam
<yuyu> kefah, salam, welcome
<kuru> roozbeh: looks like the IETF is a lot easier to get involved in
<roozbeh> kamal: lol on the job thing....
=-= roozbeh is now known as roozbeh_away
<kefah> Isam, yuyu.
<kuru> kefah: I don't believe we have met.. but I understand you are working on wine?
<kefah> Yes, Shadi Mari, one of my staff is. But apparently, he wont able to join us today. :-(
<kamal> kefah: salam. which part of wine ?
<kefah> What have been done so far is testing the arabic support on wine, fribidi and shaping.
<yuyu> kefah, is it part of the univeristy prohject ?
<kuru> kefah: is there somewhere we can track progress, help if possible, etc?
=-= ssato is now known as ssato_zzZ
<kefah> No, this is a commercial project. Shadi is a lead developer.
<kefah> Yes, so we tested fribidi and arabic letters. Shaping is not working yet.
<kuru> kefah: commercial but open source no?
<yuyu> kefah, you mean it is closed source ? ;)
-->| khamis (~firstname.lastname@example.org) has joined #arabeyes
<kuru> yuyu: wine is open source last I checked.
<kefah> Ofcourse ;-) I am an Open Source source advocate.
<yuyu> khamis, salam, welcome in
<kuru> kefah: then where can we track progress ?; )
<khamis> salam, how r u all?
<kuru> khamis: good, how are you?
<kefah> It took shadi some time until he got to that piont (around three days).
<yuyu> kuru, i know i know ;)
<kuru> puts the spot light on kefah, turns off the other lights.. starts the interrogation
<kefah> Because he faced issues with keyboard, X, and RH.
<kefah> He needed to modifiy the sourcecode a little and recompile.
<kuru> kefah: so when do you anticipate to have something to show-case.
<khamis> i am fine
<khamis> i remeber he faced some problems
<kefah> After that we stopped working on wine, because we've got other urgent stuff. :-(
<kuru> kefah: we are very very interested in this wine arabization project.
<isam> siksik: welcome
<khamis> hala isam
<yuyu> khamis: 'he'
<kuru> kefah: well, if you guys are no longer working on it actively, maybe Arabeyes can help?
<kefah> BTW, kuru who r u? :-)
<kuru> kefah: Mohammed Elzubeir
<kefah> Oh. Sure.
<khamis> yes we need some help there
<kefah> So, wine carries a huge potential.
<kuru> kefah: but we can save a lot of time by exchanging information so we don't have to go through the same things Shadi went though, for example.
<kefah> And if you are interested in continuing where shadi stopped, I can have him contact you directly and pass you his findings.
<yuyu> khamis, are you working on wine too ?
<kuru> kefah: I was thinking of a more 'open' form of communication ;)
<isam> yuyu: khalis is your man for all types of wine :p
<yuyu> isam: hehehe
<khamis> i helped in the initial stage only
=-= tumbak is now known as zaid
<kuru> khamis, is it true? ;)
<kefah> Shadi, also held some discussion with Shachar Shemesh and they reached the point of knowing what to change to support arabic shapping (what function, how, ...etc).
<yuyu> isam: we need champagne for the party ;)
<kefah> kuru, shoot.
<khamis> actually the expert is Shadi
<kuru> yuyu: I'll take a tequila shot ;)
<isam> will change irc client again
<khamis> i just helped in the first stages
<khamis> then i was involved in other things
<kuru> khamis: well, we need Shadi to get on the developer list on arabeyes and display his full findings ;)
<khamis> while Shadi continued the work
<kuru> so at least we have a reference point.
-->| isam (~Isam@126.96.36.199) has joined #arabeyes
<yuyu> kefah: is there a webpage or something ?
<yuyu> kefah: for reference/doc puropose
<kefah> No, all the work shadi did was internal in freesoft. The discussion he had with shemesh might be on wine mailing list.
<kuru> kefah: do you consider the work/findings so far confidential to freesoft?
<kuru> kefah: if so, no biggie..
<kefah> kuru, man, I am an opensource lover. No I would not do that. :-)
<khamis> 100 100 kefah
<isam> kuru: we had hard time convinving Kefah that BSD is fine .. as he loves GPL :)
<kuru> loving open source doesn't mean you can't consider the initial development stage confidential ;)
<khamis> i don't think that kefah is convinced
<kuru> kefah: my point is, we need a public location of tracking progress and documenting important things.
<kuru> kefah: either freesoft should have such a page, or we (arabeyes) can help in that.
<yuyu> kefahmaybe this can be done at arabeyes.org
<kuru> kefah: the important thing is that it is available to the public so if someone says, "hey, I can help move this on" then they can start right off.
<kefah> kurur: I see your point.
<isam> kefah: having it in the Arabeyes means that other paries may be invovled in the development, I don't guess this is something that you would mind
-->| Net (~email@example.com) has joined #arabeyes
<kefah> I believe for the time being, it would be best host by arabeyes. I can have shadi prepare a quick howto on configuring and running arabic on wine.
<kuru> kefah: being an open source business, I think that would possibly be one of your biggest advantages (ie. free labor) ;)
<isam> Welcome Hamza .. another Freesofter :)
<kuru> Net: salam -- you probably want to get on irc via regular user (not 'root')
<Net> hi isam , yah , ithink u right
<kuru> wow.. this is a FreeSoft world ;)
<kefah> kuru, isam: I agree. this is what open souce is about. my company is making money out of the services I provide to others. so there is no conflict of interest here. :-)
<yuyu> Net: salam - welcome in
<yuyu> Net: you better un-root ;)
<-- Net has left freenode (Client Quit)
<kuru> kefah: what other projects are you working on ?
<isam> yuyu threatend NET !
<kefah> kuru: My main goal is delivering linux to the mainstream. Making linux a comodity operating platform. freesoft has created freeENTERPRISE. Which is a collection of open source based linux solutions.
<yuyu> kefah: what is freesofts website ? is it freesoft.org ?
<isam> yuyu: freesoft.jo
<kefah> kuru: (... continued) aimed at the enterprise. freeGATEWAY, freeEXCHANGE, freeSERVER, and freeDESKTOP.
<kuru> kefah: I can see 'freeDESKTOP'.. what do you do in terms of 'polished' arabic?
<kuru> kefah: so you are ostly in the 'support/solutions' business.
<kefah> kuru: freeDESKTOP is where arabization effort mainly lays.
<kefah> kuru: yes.
<Stanislav> I need to go. It was a great pleasure to be here. Congratulations and keep up the great work!
<kuru> Stanislav: thanks for coming!
<ninjadude> before you go, have you got a jabber id?
<kuru> Stanislav: and keeup the great work on KBabel
<isam> Stanislav: thank you for passing by.. we will be here for another 10:30 hours :)
<yuyu> Stanislav: fell free to come back later
<yuyu> Stanislav: or tomorrow, or next week or ... ;)
<kuru> kefah: so can we set a time-table for when Shadi would be able to deliver some documentation?
<kuru> yuyu is always here
<Stanislav> Thank all of you, I know we will meet again soon!
<Stanislav> Enjoy the party....
<khamis> thanks for you
<-- Stanislav has left freenode (Remote closed the connection)
<khamis> kefah is away and will be back soon
<kuru> khamis: yell at him for me ;)
<kuru> khamis: "your computer jusst caught fire"
<khamis> i cannot he is my boss
<kuru> khamis: not if you're trying to save his computer ;)
<kamal> khamis: wow, IRCing with your boss ? It's like smoking in front of your father ;)
<isam> kamal: LOL :) what an anology !
<khamis> will we work in freesoft to have a freeWORLD
<kuru> kamal: tell me about it.. I want that boss ;)
<khamis> no you cannot have him
<khamis> he is my boss
<kuru> khamis: I will be that boss then ;)
<kuru> khamis: heheh.. so, what are you working on now?
<kuru> khamis: anything of special interest?
<isam> kuru: that is clasified info .. you have to be shot after hearing what he will say
<khamis> i am the freeSERVER and freeGATEWAY guy
<khamis> i helped in freeDESKTOP
<yuyu> isam: i heard that too ;)
-->| Net (~firstname.lastname@example.org) has joined #arabeyes
<isam> khamis is multi talanted
<khamis> sorry not freeGATEWAY -> freeEXCHANGE
<yuyu> Net: welcome back
<Net> Net Worker Is back !!
<khamis> not in the gateway and firewall things
<kuru> Net: sorry for scaring ya off ;)
<khamis> i need somethime to catch up there
<Net> thanks yuyu
<isam> khamis: are you following the Open GroupWare thing ?
<kuru> khamis: that fun stuff.. I've been struggling with that all day today at work ;)
<yuyu> Net: i hope we diddnt frighten you ;)
<khamis> no, but i can look into it
<kefah> kuru: what do you do for living? I learned from nadeem that you came back from the US.
<khamis> i love everything in opensource even if it is bad
-->| lars (~email@example.com) has joined #arabeyes
<yuyu> lars hi
<kefah> Hello Lars.
<isam> Lars: hi
<yuyu> lars welcome in :)
<khamis> hi lars
<kuru> kefah: I am the 'IT Manager' as they cal it at this trading company.. I manage all their computer systems (spanning several countries)
<yuyu> lars: a bit early no ? couldn't resist to come and say hello ? ;)
<kuru> kefah: I'm based in dubai now.
<kefah> kuru: Oh. That is interesting.
<kuru> lars: nice of you to come by
<lars> yuyu: not really. I said between 5 and 6pm Oslo time ;-)
<kuru> kefah: I just starte sticking linux everywhere ;)
<lars> kuru: thanks, I'm happy to come :)
<isam> Lars: I don't know if you remember.. but I used to nag on the Arabic shaping breaking issue after some flow point numbers
<kefah> Lars, welcome. I would like thank you for the magnificent work on QT. great job.
<lars> isam: Sorry, but it's a bit hard to relate IRC nicknames to some bugreports for me ;-)
<kuru> lars: so where do we stand on the harakat with Qt?
<lars> kefah: thanks :)
<isam> lars: we invivted you to start complaining about Arabic issues if you do not mind
<kuru> kefah: so, like I was saying, can we set a time-table as to when we can expect to get some docuemtnation on Shadi's work?
<lars> isam: I was afraid of that ;)
<yuyu> lars: i am Youcef Rahal. we had some exchnges on qt bugs and qt translation
<isam> lars: one of the still hanging issues is the Shaping with HARAKAT or TASHKEEL .. wait .. I will find the definition for it in the UNICODE pages
<kuru> lars: the good news is, we don't have any snipers in Oslo ;)
<Mohammed_Yousif> lars: They set-up a trap for you :-)
<kefah> kuru: no time table, but a specific date. Let me work it out with shadi. feel free to send me you contanct info, so I pass the doc to you.
<lars> kuru: what's your problem with them? They are supposed to work.
<kuru> kefah: firstname.lastname@example.org would be a good place to send anything related to his work.
<kuru> kefah: but my personal email is email@example.com
<kefah> kuru: will do ensha'allah.
<lars> kuru: Seems I'm lucky I'm living at the end of the world ;-)
<-- zaid has left freenode ("Client exiting")
<kuru> kefah: what's Shadi's contact address? (so I can harass him personally) ;)
<kuru> kefah: should'a guessed ;)
* yuyu has quit IRC (Read error: 104 (Connection reset by peer)
<lars> isam: as I said after my tests Harkat seem to work. tashkeels
are only used for justifying and are currently not implemented.
* yuyu2 checks his list for qt issues ;)
* yuyu2 is now known as yuyu
<isam> lars: when you draw one of the HARAKAT you are supposed to have
one Haraka on a letter at a time
* ChanServ sets mode: +o yuyu
* zaid has joined #arabeyes
<kefah> lars : what about the arabic indic digits? any work on that?
<isam> lars: but currently, I can draw 3 different types of Harakat on
the Same letter
<kefah> lars: In the arab world we use inidic digits most of the time.
<lars> isam: ok. if that's the only issue I can probably fix this. I
never bothered because I up to now thought I rely on the user to
input meaningful text.
<lars> kefah: that will have to wait until after Qt 3.2. The whole
issue of localized digits is rather involved (there are many other
languages using their own digit shapes).
<kefah> lars: is this a matter of rendering. A global digit rendering
option which would display in either indic or arabic? as we should
not break the applications that depend on numerals.
<isam> lars: what other Arabic shaping engines do is that they ignore
the input of a HARAKA if a HARAKA is already there
<lars> kefah: they have their own code positions in Unicode. In the
long term I want to be able to make conversions to localized digit
<isam> kefah: I think there is a need to have a system setting for
that .. to select whether use "Arabic, Indic, System, or Context"
<kuru> lars: note that not all of the 22 arab speaking countries use
indic numerals by default.
<yuyu> kefah: not most of the time ;)
<yuyu> kuru: yeas, excatly
<yuyu> lars: in the western arabic countriues, arabic digits are used
<Mohammed_Yousif> kuru: The major ones do use them
<kefah> kuru, yuyu: you are right. not
<lars> isam: possible. The other possibility is to use the approach
used for indic languages: You show the additional haraka on top
(better around) of a dotted circle
<kuru> Mohammed_Yousif: that's irrelevant. The point is, it should be
something that can be set via locale settings.
<lars> yuyu, kuru: I know all that. It's just that we have to find a
solution that will also fit all other languages having localized
<kefah> isam: I fully agree. My point here is, numerals (digits) must
be on the rendering level. i.e. actual characters must always be the
<lars> all indic languages have them aswell for example.
<kefah> isam,lars: this is to allow all applications that use numerals
to act properly even with indic digits. I believe this is the case
* shadi has joined #arabeyes
<lars> kefah: yes, but the problem is that sometimes you don't want
localized shapes. Think about your web rbowser and you looking at an
english web page.
<kuru> lars: so how involved is this issue? ie. from Qt's point.
<Net> ÃâÃÂ¨ÃÂ¤Ã?ÃÂ¢ÃÂ©ÃâÃÂ£ÃÂ¯Ãâ¹ÃÂ©ÃÂ»>>ÃÅ ÃÂ¨Ã¢ËÂ«ÃËÃÂ®ÃÂ´,ÃâÃÂ ÃÂ©Ã?ÃÂ®ÃÂ®Ã¢â¬Â°Ã¢ËÂ«ÃÂµÃâÃÂ¦Ãâ
<shadi> hello everyone
<kuru> in a perfect world, all digits should be Arabic numerals ;)
<khamis> here is the man
<yuyu> shadi: hello - welcome in
<khamis> Shadi is here guys
<kefah> lars: unlike arabic / english, where you should be able to
type both arabic and english, in numerals you should only choose
<kuru> shadi: hello there.
<yuyu> kuru: hehehe - right ;)
<shadi> hi khamis
<shadi> whats up guys?
<kefah> kuru, shadi. shadi, kuru.
<kuru> shadi: kefah was telling me earlier that your job depended on
you mailing us some documentation on your wine research/work in the
next 12 hours ;)
<lars> kefah: other people have different opinions about this. We are
thinking of providing some settings where you can choose your
preference. But as I said this is not going to happen whithin Qt
<shadi> kefah, you are killing me man, no way
<yuyu> lars: i have a few questions on qt trannslation
<kefah> lars : I see.
<shadi> what have you been discussing guys?
<kuru> shadi: he also said something about it being in SGML format
<kuru> shadi: everything.. but wine came up for awhile.
<kefah> kuru : :-D
<yuyu> lars: has the arabic translation been integrated into qt ?
<shadi> SGML, back to the old days, i thought its dead ;)
<lars> yuyu: for qt itself and the demo example, yes. not sure about
the other translation files.
<kuru> shadi: hey! all arabeyes documentation is in sgml ;)
<isam> shadi: you use the Web ? PDF ? PS ? it is all SGML :)
<shadi> Kefah, what do you have for us today?
<yuyu> lars: great ;)
<shadi> absolutely, but how about writing SGML
<isam> or maybe not :)
<yuyu> lars: how can we coordinate to have the .ts files beign
updated for current versions of qt ?
<kuru> shadi: seriously though, when do you think you can put together
a comprehensive document on everything you've done related to wine?
<yuyu> lars: 3.2 for example ?
<kefah> lars: any news on reversing the tree control? I wonder how
hard is that.
<lars> yuyu: the easiest way is to send a mail to firstname.lastname@example.org
whenever you have an updated version and ask for inclusion into the
<yuyu> lars: we cant include a ts file before the release ? ;)
<kuru> lars: btw, (a bit off-topic).. do you know if Trolltech has any
intentions of putting out a later version of Qt for windows?
<shadi> kuru: well, i havent yet thought about putting a document, we
need to arrange for this, i assume you are aware of what i have been
doing on wine!
<lars> back to harakat for a second: could you have a look at
tell me if the description in there is correct. I could implement
this rather easily.
<kuru> shadi: I am only aware that you are the most intimitely
involved in it..
<lars> kuru: 3.2 for Windows exists ;-)
<kuru> lars: free?
<kuru> I must be dyslexic.. but I read that 2.3
<lars> kuru: no. I was joking. There is an ongoing discussion inside
Trolltech about licensing for the windows version, that's all I can
say for now.
<lars> kuru: even in arabic numbers are written left to right ;-)
<kuru> shadi: it's always good to document as you go along so you
don't get hit like now with the need for a comprehensive document ;)
<kuru> lars: heh.. that must be it ;)
<isam> lars: reading it..
<shadi> kuru: you are right, i can list all the steps i did and all
the colusions i reached to
<lars> yuyu: it's a bit difficult, but for x.y.0 releases it should be
possible if you check the betas. for patch level releases you
usually don't get any new strings.
<kuru> shadi: I understand from kefah that you are at the point where
you know all the entry points in the wine code that need to be
<kuru> shadi: I think that is the most crucial
<yuyu> lars: ok - i understand
<isam> lars: the MS link that you gave me describe the probelm
properly, as for the solution .. you know better .. I am not
familiar with Dotted_circle .. but if it is an empty space to base a
Haraka on, then this is what we need
<shadi> kuru: thats true, and im going to document this very soon
<lars> isam: look a bit up on the same web page. They have an image of
the dotted circle on the page ;-)
<lars> isam: MS requires that every arabic true type font has one of
<isam> lars: you see that dotted_circle while you are typing the way
you see it in the image ?! I thought that is an imaginary circle :)
<kuru> shadi: great. I'll email you and see when we can get that going.
<kuru> shadi: I'm very interested in wine and any development that can
head it toward proper arabic support.
<shadi> kuru: will be waiting from you :)
* kuru sends an email now
<lars> isam: you see it the way you see it on the image. The harakat
is placed somewhere around the circle (above or below for arabic).
<kuru> the dotted circle is just a transparent glyph
* roozbeh_away is now known as roozbeh
<lars> isam: I can show you how it looks for hindi. One second.
<shadi> kuru: this is crusial, it will be a big hit.
<isam> lars: so that will indicate that there is an error in input..
but is that circle a Joiner ? meaning will it break shaping ?
<isam> kuru: is it zero width ?
<lars> isam: shaping is not defined for it. The circle is only for
rendering. It won't be in the string you input.
<isam> lars: so it is like the ZWJ but the difference that it can be a
base for a HARAKA
<isam> lars: but how will this fix the issue of double haraka ?
* roblimo has joined #arabeyes
<lars> isam: I'm not sure what kind of joining properties it should
<yuyu> roblimo: hi- welcome
<lars> isam: in that you'll see immediately when you made an error in
<roblimo> yuyu: Same to you, fellow! :)
<kuru> isam: it has no effect on shaping.
<kuru> roblimo: hey.. you're the one from newsforge
<isam> Roblimo: Welcome :) nice that you showed up
<kefah> roblimo, hello. We met at the OSS conference in Amman.
* Net has quit IRC ("using sirc version 2.211+KSIRC/1.2.4"
<roblimo> kuru: Right now from Slashdot; sending interview questions
to Bruce Perens.
<kuru> roblimo: oh.. /. it is.
<roblimo> Hi, isam - kefah - how's life.
<roblimo> kuru: NewsForge again in an hour.
* Net_Worker has joined #arabeyes
<shadi> lars: hi, i got a question for you. After i took a look at
part of the source code of QT, especially the part that deals with
glyphs, i noticed that you only dealing with standard arabic glyphs
and excludes a certain range. Is that true?
<roblimo> Might as well bang out a fast story about ArabEyes' 2nd
<isam> roblimo: things are fine with me .. thank you for showing up
<roblimo> isam: You're in SA these days?
<isam> roblimo: that would be great :)
<yuyu> roblimo: that would be _really_ great ;)
<isam> roblimo: no .. I am working with an SA company from Amman :)
<roblimo> I've moved to a small country called Florida, just south of
<lars> isam: that's how this looks for indic. The last character input
was a non spacing mark (same as a harakat) that doesn't belong at
this place: http://trolls.troll.no/lars/wrong_input.png
<isam> roblimo: :) I will move south soon too
<isam> lars: so you see the dotted circle in real .. now is that
dotted cicle defined in UNICODE ?
<lars> shadi: not really. I did for a long time limit Qts shaping to
the so called "presentation forms B". Using open type shaping this
is no longer the case in Qt 3.2.
<lars> isam: yes it is. 0x25cc in Unicode.
<isam> lars: so, we should expect all UNICODE fonts to have that
<isam> lars: I guess that is good to indicate the error made ..
<lars> shadi: the limitation to presentation forms B is still present
if you use server side fonts (xlfd fonts), or if your font doesn't
have open type tables.
<lars> isam: if the circle is not there, you'll usually get the harkat
drawn around the missing glyph character (an empty box).
<isam> roblimo: by the way, about 15 Jordanian are thankful for the
Mandrake CDs :) and those 15 probably passed them to more pwoplw
<shadi> lars: oooh i see
<lars> isam: but it should get common practise to include the glyph in
all arabic fonts.
<kuru> isam: the "otted circle" is a neurtral.
<isam> lars: no no harm done either ways.. I guess the implementation
fixes this error
<kefah> lars: I got microsoft fonts (tahoma and the others), I do not
know what xlfd, bud xfs is turned off. but still formB is not there.
<isam> lars: but there is a need to define the behavior of that
circle, if ti is a joiner or not
<khamis> hi roblimo
<kefah> roblimo: This world is very small. here we are meeting again.
* anmar has joined #arabeyes
<yuyu> ha anmar !
<kuru> anmar, you made it ;)
<anmar> Good morning people......
<anmar> let the party begins :)
<isam> kefah: log to slashnet and see roblimo more :)
* ChanServ sets mode: +o anmar
<lars> kefah: X11 has to font APIs: Xft (you probably heard about that
one) and the old one: xlfd.
<Net_Worker> hi anmar
<lars> If your fonts are antialiased you're using Xft.
<anmar> Net_Worker: hello there...
<yuyu> anmar: we didnt wait for you - we finished the cakes and the
<kuru> lars: has Keith gotten to a stable API for Xft by now?
<anmar> yuyu: daoo..... I knew it
<lars> isam: IMO it should be non joining.
* systems has joined #arabeyes
<lars> kuru: sort of. He changed some stuff from Xft1 to Xft2, but Qt
supports both now.
<anmar> Net_Worker: so what brings you here...
<kefah> lars: so what does it take to have FormB on QT / X11 ( I use
<Net_Worker> the open source world
<anmar> Net_Worker: hehe....:) good
<Net_Worker> i'm a freesoft[er] now !
<kuru> anmar: you know each other?
<anmar> kuru: you mean Net_Worker ? No
<Net_Worker> no , i don't think so kuru , but i like the nick name
<kefah> lars: yes, my fonts are anialiased.
<anmar> Net_Worker: it is my real name :)
<Net_Worker> ok , nice name
<lars> kefah: so you're using Xft. It all depends on the font you
have. If it is an open type font, you'll get most of the advanced
shaping. If it is an older font, you'll get only presentation forms
B (which is usually a subset of what the font has to offer).
<anmar> Net_Worker: hehe...
<anmar> hello khamis
* lars has quit IRC ("using sirc version 2.211+KSIRC/1.3.7"
<khamis> hello anmar
<kefah> lars: Microsoft's Tahoma, Arial, Traditional Arabic ...etc.
Already got formB. I am not sure if they are open type compliant.
are you a ware of any?
<anmar> ninjadude: love the name... been watching jackie chan movies
* lars has joined #arabeyes
<kefah> lars: Microsoft's Tahoma, Arial, Traditional Arabic ...etc.
Already got formB. I am not sure if they are open type compliant.
are you a ware
<kefah> of any?
<anmar> khamis: I did not have the honour of meeting you yet ?
<lars> kefah: depends on their age. The lates versions of the fonts
have open type tables.
<kefah> lars: I got the latest. So, do I need to check any thing else,
if I can not see formB?
<khamis> me too, but i heard about you and i hope we will have the
chance to meet each other soon
<anmar> khamis: it is all lies.. do not believe a word you heard :)
<lars> kefah: you should always get presentation forms B (they define
the basic connected forms), forms A you might not get.
<khamis> ok thats better
<kefah> lars: BTW, I was going through the link: I believe DIAC8 is
wrong, it should be Arabic Hamzah not Maddah.
* Net_Worker has quit IRC ("using sirc version 2.211+KSIRC/1.2.4"
<lars> kefah: right. the Unicode points afterwards U+0654 and U+0655
<lars> so the table is right, but the description is wrong.
<kefah> lars: oh. My appologies. I meant form A not B through all the
<lars> kefah: there are ways to get forms A for non open type fonts
aswell, but it's a significant amount of work, as most fonts only
define subsets of forms A.
<isam> kefah: that Unicode 3.0 Bok came handy today :)
<lars> kefah: I guessed so after a while ;-)
<lars> isam: it's lying next to me ;-)
<lars> kefah: how important is it for you to get presentation forms A?
<roozbeh> Ã¢?âÃÂ©ÃÂ¿Ã¢â°Â¥Ã¢âÂ¬ÃÂ¥ ÃÂ¸ÃâÃ¢âÂ¬ÃÂ¥Ã¢â¬Å¡ÃâÃÂ¥ÃÂ¿Ã¢âÂ¢ÃÂ¸ÃÂ ÃÂ¸ÃÅÃÂ¸ÃÂ¡ ÃÂ¿ÃÅ¸Ã¢âÂ¬ÃÂ¥ÃÂ¸ÃÅÃÂ¸ÃÂ ÃÂ¿ÃÂ®ÃÂ¿Ãâ ÃÂ¸ÃÂ ÃÂ¸ÃÅÃÂ¸ÃÂ¡ÃÂ¿ÃÂ¼
* Shachar has joined #arabeyes
* khamis has quit IRC ("using sirc version 2.211+KSIRC/1.2.4"
<Shachar> happy annual evolution party
<isam> Shachar: thank you .. welcome in
<Shachar> so someone
* alive :-)
<zaid> i am
<kuru> Shachar: hello there
<zaid> its just they r talking way too much technical for me
<shadi> hi Shachar, how are you doing?
<isam> Shachar: minutes ago Wine was the main issue .. not we are on
<Shachar> zaid: So far, I don't see them talking at all
<Shachar> shadi: greate, thanks
<Shachar> isam: So basically, I just missed my opertunity ;-)
<zaid> Shachar: they just stopped talking :p
<Shachar> zaid: which means that they are no longer too technical
<Shachar> always look on the bright side
<roozbeh> isam: while you're at it, try preordering the 4.0 book:
* wal has left #arabeyes
* wal has joined #arabeyes
<isam> Shachar: Shadi, Kefah, Kuru, and others are stil here :) so you
will have your turn :)
<Shachar> roozbeh: I bought the 3.0 one, and was disappointed
<shadi> Shachar, what r u up to those days?
<roozbeh> Shachar: how dissapointed? the binding came loose too soon?
<Shachar> an excellent paper weight, but not up to par with the PDFs
* Mohammed_Yousif has quit IRC (Read error: 60 (Operation timed out)
<roozbeh> Well, you can't read the PDFs in bed :D
<Shachar> shadi: in case you havn't heard, we opened an NPO for
promoting free and open source software in Israel.
<isam> roozbeh: so you are a bed reader .. I see
* A`WaRe has joined #arabeyes
<kamal> hello Shachar
<kefah> lars: (i was a way) Form A adds a lot to the beaty of arabic
scripts. Before that let me continue going over your link.
<isam> Shachar: like a LUG ?:)
<Shachar> shadi: takes most of my time, but puts MS on edge, so it
must be a good thing
<kefah> lars: DIAC7, and DIAC6 are ok.
<Shachar> isam: no, a formal legally recognizable body
<Shachar> we have lots of LUGs already
<kamal> Shachar: you mean hamakor ?
<kefah> lars: DIAC5 ok.
<Shachar> we even put a small word in about arabeyes, though I rather
doubt you saw anything out of that
<kuru> Shachar: url?
* A`WaRe is now known as jAckAl[S]
<Shachar> what were the Wine discussions?
<kefah> lars: in DIAC 4, 0610-14 are reserved?
<Shachar> kuru: http://www.hamakor.org.il, but most of it is in
Hebrew. There is a solitary English page somewhere there.
* java2arab has joined #arabeyes
* ibrahim has joined #arabeyes
<kuru> Shachar: we were trying to coerce shadi to write up some
documentation on everything he's done/found on wine/bidi so we can
have more people involved
<ibrahim> salam alikoum
<java2arab> Salaam Alukem
<ibrahim> arabeyes is full today:)
<lars> kefah: not sure what they are. Unicode 4.0?
<roozbeh> kefah: what is DIAC 4?!
<shadi> Shachar, what are the latest major improvements have been done
on wine from your side?
<isam> FYI: for those used to full channels, Arabeyes channel had
reached it's full capacity today
<lars> kefah: about shaping: trolls.troll.no/lars/arabic.png . This
uses fonts from MS.
<Shachar> shadi: I've got ICU integrated
<isam> so, wine droped fribidi and moved to ICU ?
<Shachar> wine never had fribidi
<Shachar> I was waiting for Behdad to move it to the new interface, so
I can put in the UTF-16 support
<shadi> how would it make a difference from the code was put in there
<Shachar> He actually sounded relieved to hear I moved ahead with ICU
<Shachar> shadi: I didn't understand the question
<kefah> lars: DIAC4 : 0610-0614, and 0659 are reserved. The rest are
OK, execpt that they are missing 06E1.
* systems has quit IRC (Read error: 60 (Operation timed out)
* zaid has quit IRC ("Client exiting"
<lars> kefah: for which group?
<shadi> wine didnt support fribidi right? and the code to implement
lets say simi-bidi was there i think was dealing with UTF-8 right?
<lars> kefah: DIAC4 aswell?
<kefah> lars: DIAC4 : Arabic Quranic Marks above.
<Shachar> shadi: wine never used fribidi. It originally used a minimal
two-rules algorithm I wrote so it will have something
<roozbeh> kefah: but what is DIAC4?!
<lars> kefah: ok. That's no problem to add.
<lars> roozbeh: http://www.microsoft.com/typography/otfntdev/arabicot/o
<Shachar> Also, wine doesn't work with UTF-8
<kefah> lars: DIAC3 OK. sorry roozbeh, I was engaged.
<Shachar> mainly because Windows doesn't
* jAckAl[S] is now known as jAckAl[Busy]
<Shachar> The major missing things are Bidi Edit control and keyboard
<lars> kefah: 0x610-0x614 are defined in Unicode 4.0.
<roozbeh> lars, kefah: those DIAC things are an MS invention. You
can't generate that automatically from Unicode data...
<shadi> so Shachar, the latest code that deals with ICU, has it been
<lars> roozbeh: I'm not talking about creating that automatically.
It's rather easy to add a short lookup table to Qt for that.
<kefah> lars: oh . I see.
<roozbeh> lars, kefah: actually, there are new characters in the
Arabic block in 4.0 that don't fit into that model.
<Shachar> I didn't know Unicode 4 changed anything as far as bidi was
<Shachar> shadi: already in CVS
<roozbeh> specifically, U+0600 and the gang.
* java2arab has quit IRC ("Client exiting"
<kefah> lars: DIAC 2 (0656) is not in Unicode 3.0
<roozbeh> lars, kefah: specifically, U+0600 and the gang.
<Shachar> for edit control it will probably have to be seperated into
a different DLL, though
<Shachar> there was great opposition to doing that before it was
necessary by Alexandre, so I waited
<shadi> Shachar, there was an idea to seperate the code into another
DLL as for bidi, is it the case now?
<lars> roozbeh: hmmm... the main question for me is how to handle them
in an input widget.
<Shachar> read my reply :-) not yet
* nadim has joined #arabeyes
* ChanServ sets mode: +o nadim
<isam> Salam Nadim
<lars> roozbeh: kefah and some others were asking me to visually
indicate invalid harakat on input.
<nadim> isam: how are things shaking :-)
<Shachar> shadi: i havn't seen anything from you. I was expecting the
<yuyu> nadim: salam
<lars> s/harakat/combinations of harakat/
<nadim> someone keeping a log ?
<ibrahim> nadim wa alikoum al salam
<yuyu> nadim: i do,
<kefah> lars: DIAC 1 should move 06E1 to DIAC4. and 0657 is not
defined in Unicode 3.0.
<nadim> yuyu: great
<anmar> nadim: i opened logging on irssi
<yuyu> nadim and anmar too i think
<roozbeh> lars: i'll say wait for 4.0's arabic chapter to arrive, then
see what it is recommending about the new chars. don't support
unicode 4.0 chars until you really understand what kind of devil
<nadim> sorry I'm late..
<yuyu> nadim: but save it fro myour side too please
<isam> nadim: things are fine ..
<nadim> yuyu: will do
<roozbeh> lars: just keep with what ms is already saying, at most.
don't extend it to 4.0.
<lars> kefah: have a look at http://www.unicode.org/charts/ for the
unicode 4.0 chart