Re: Automatic paragraph direction and BiDirection layouts in Gtk+

[Behdad Esfahbod <behdad cs toronto edu>]

Hi Jonathan,

Nice to see someone's interested :).  BTW, as Dov already
mentioned, my document was only a draft to start with, and it has
known flaws.  More below:

On Thu, 18 Mar 2004, Jonathan Ben Avraham wrote:
> Hi Dov,
> See inlines below...
>
> On Thu, 18 Mar 2004 dov imagic weizmann ac il wrote:
> > Hi Jonathan,
> >
> > Regarding Behdad's spec, note that there has been no decision to
> > implemented it into gtk. I expect it will have to be changed in
> > several ways until it may be implemented. Look at my comments
> > regarding what Behdad wrote below. E.g.  the widget alignment
> > today is assigned with a floating point number between 0.0 and
> > 1.0. Adding "inherit" and "against" to such a scheme might require
> > an API extension.
>
> How did widget alignment ever get to be a floating point number? How
> difficult would it be to change this to an enum or even an integer?

That's actually really nice that they are float.  That's not
going to be a major problem.  For example, I have already
proposed a solution in the same thread.

> Actually, not being familiar with the code, I shouldn't be commenting on
> this.
>
> > Being familiar with the gtk-sources I feel that I can quite
> > safely say that doing the actual implementation is quite easy
> > once the spec has been written. The tough part was the
> > automatic propagation of the base direction that is already
> > in Gtk-2.4. All that remains is book-keeping of all the settings
> > and then extracting and overriding the basedirection and the
> > alignment from the agreed upon source.
> >
> > Thus most of the work is reaching an agreement of what should be
> > done, and there Owen Taylor has the final word. Thus I believe
> > that we should first use the mailing list to reach an agreement,
> > and then I'll be happy to implement it.
>
> Would it be possible to implement it in a test workspace so that we could
> try it?

The idea is to start a discussion and then go for an
implementation that eventually will be provided to Owen...  I'm a
bit busy this month, so early April is fine with me.

>  - yba

--behdad
  behdad.org