Re: X atoms rework

[Jan Djärv]

I think only the fourth part of the patch is worth it.  The atoms that are not 
gotten at startup are not in any critical path, rather the functions that use 
them are called very seldom.  Some atoms are only used once.

        Jan D.


Julien Danjou skrev 2010-11-10 14.49:
> I've been hacking in xterm.c this days, and saw some things I did not
> like that much about the handling of X atoms.
>
> Therefore, I rewrote this part, and here's a set of patch that enhance
> things a bit.  It should make X initialization faster over the
> network, at least, and stop requesting atoms that Emacs should already
> have locally.