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

Re: Patch to remove presentation forms from Arabic xkeyboard layout



On Fri, Jan 04, 2008 at 12:00:00AM +0200, Mohammed Sameer wrote:
> On Thu, Jan 03, 2008 at 03:35:42PM +0200, Khaled Hosny wrote:
> > On Thu, Jan 03, 2008 at 01:32:55PM +0200, Mohammed Sameer wrote:
> > > On Wed, Jan 02, 2008 at 04:58:02PM +0200, Khaled Hosny wrote:
> > > > I just submitted a patch[1] that removes the various Lam-Alef presentation
> > > > forms from Arabic keyboard layout, this should temporally fix the
> > > > well-known lam-alef problem[2]. To avoid confusing the user, the b,
> > > > shoft-b, shitf-g and shift-t keys in the Arabic keyboard (should produce
> > > > لا, لآ, لأ and لإ respectively) will now produce nothing.
> > > > 
> > > > 1. https://bugs.freedesktop.org/show_bug.cgi?id=13894
> > > > 2. https://bugs.freedesktop.org/show_bug.cgi?id=8195
> > > > 
> > > 
> > > No. I'm sure they are there for a reason even if we don't know what it is.
> > > Isam Baiazidi was the one who added them. I'd rather ask him about the consequences
> > > first before removing them. I know they are causing problems and I'm always implementing
> > > workarounds for them but we need to know why were they added.
> > > 
> > For whatever the reason, this shouldn't ever happened, presentation
> > forms are here for computability with legacy systems and shouldn't be
> > used to encode text, this is even discouraged[1] by Unicode standard.
> > 
> > Even if we agreed on necessity of using them, then we will need 12 keys
> > for them; 4 letters × 3 contextual forms, cause we are now just using
> > one contextual form (isolated) for all cases, with is IMHO non-sense and
> > is the source of the rendering problem.
> > 
> > Having a separate key for Lam-Alef belongs to metal typesetting era and,
> > for no good reason, it survived in the digital era. 
> > 
> > I can't see any good use for them, however, I had no idea who added them
> > to ask.
> > 
> > 1. http://unicode.org/faq/middleeast.html#0
> 
> It'll still be a regression. The proper fix is to fix xkb itself.

Regression from what? I agree it is a regression from the expected
behaviour, but not from the current, definitely broken, behaviour. A
proper fix should involve xkb for sure, but I don't have the knowledge
to do that myself and no body is welling to fix it.

-- 
 Khaled

Attachment: signature.asc
Description: Digital signature