Re: antialiasing for emacs

[Miles Bader]

Hi, I tried your patch, which seemed to work very well at first -- my
normal emacs fonts were indeed anti-aliased, and looked very nice.  The
patch seems impressively small.

I had to use the following extra patch to get things to compile on my
system (Debian unstable), apparently due to bogosities in <png.h>.
I'm not sure why your change tickled it; maybe due to including <Xft.h>
in dispextern.h.

emacs-xfns-png-include.patch
Description: Patch to include <png.h> before emacs-specific includes

When I tried to go into info (which uses a fair number of fonts), emacs
hung, with an error on standard output like:

   Xlib: unexpected async reply (sequence 0x1248)!

I then tried to start emacs with a more standard configuration, and
_that_ made it hang immediately with a similar error message:

   (emacs) xrdb -remove
   (emacs) src/emacs -q -fn -*-courier-medium-r-*--12-*-*-*-*-*-iso8859-1
   Xlib: unexpected async reply (sequence 0x545)!

[I had to use the `-fn' option because the standard font emacs uses
doesn't seem to exist on my system!]

I guess this must be one of the small bugs you were talking about... :-)

> [1] The slowness is because I can't trust the attributes of a face not
> to change.  If there was a different face for the cursor than the rest
> of the text (instead of simply changing the GC), the story would be
> different. 

I'm confused by this -- isn't there a separate face for the cursor (`cursor')?

-miles
-- 
If you can't beat them, arrange to have them beaten.  [George Carlin]