Re: about P1 part of BIDI alogrithm

From: Eli Zaretskii <>
Date: Tue, 11 Oct 2011 08:26:53 -0400

> Date: Tue, 11 Oct 2011 12:39:57 +0200
> From: Kent Karlsson <>
> CC: <>,
> <>
> Well, I think there is a silent (but reasonable, I would say) assumption
> that mirroring does not change the width of a glyph... I would think that if
> a font does not fulfill that, then you have a font problem (or mix of fonts
> problem), not a bidi problem.

It would be considered as a misfeature if Emacs would fail to display
a character just because it is from a different font. Of course, if
the mirroring glyph comes from a very different font, the result will
be ugly, but it is still legible, unlike if you just paint a hollow
box with some hex number, or display only a part of the glyph because
it didn't fit the window.

IOW, the Emacs implementation of the UBA is an integral part of its
display engine, and so it cannot say "it's a font problem, not a bidi
problem", because eventually it becomes a display engine problem...
Received on Tue Oct 11 2011 - 07:29:48 CDT

This archive was generated by hypermail 2.2.0 : Tue Oct 11 2011 - 07:29:49 CDT