Re: Summary of bidi branch?

> Cc: [email protected], [email protected]
> From: David Kastrup <[email protected]>
> Date: Fri, 30 Mar 2007 14:50:57 +0200
> 
> During the Emacs 21 pretest phase, I submitted dozens of bug reports
> about display engine problems in connection with preview-latex.  I
> think that quite a number of optimizations got killed because of that:

Yes, but the most important (those I mentioned and a few similar ones)
are still there.  Gerd told me back then that he actually tried the
display engine without any optimizations, and found that it was
unusably slow.

> IIRC, emacs-bidi was supposed to be a rather clean design implementing
> the specs quite straightly, correct?

If you mean the UAX#9, then yes, the intent was to do exactly what it
says.

> So one will probably have more
> cleanup remaining outside of the patch itself than inside it, right?

I don't know.  There are a few gray areas in UAX#9, which we would
need to interpret as best for Emacs.

> I am just trying to get some feeling about the situation, to get a
> guess how much work might be entailed before it is realistic to think
> about getting it into some release in the future.

I think we will not have any idea about this until Someone(tm) will
spend some time using and debugging the display engine with bidi
turned on.