X startup times: Jaunty vs. Moblin

[Timo Aaltonen]

On Thu, 12 Mar 2009, Bryce Harrington wrote:

> Another note of possible relevance is that X depends on hal now, for
> input device stuff.  I notice in the white log, that hald has completed
> its work before X begins (looks like it sleeps until hald is done?)
> Whereas in the black case, there seems to be hald activity going on in
> parallel with the X startup.  Not sure what it means, but it looks
> interesting.  I notice in the moblin patches is one that indicates some
> sort of hald synchronization.  (Scott, mind tossing the patches up
> somewhere for us to poke through?)

xserver 1.6.0 contains commit 6b8bbbd21b39181e9 which should do the same:

http://cgit.freedesktop.org/xorg/xserver/commit/?h=server-1.6-branch&id=6b8bbbd21b39181e9dd3175fa73c05008c361ba2

trying without xorg.conf would rule out the difference in setting up the 
default devices. Btw, the initial startup is already a lot slower. By the 
time the pci bus has been probed, we've lost 0.3 seconds.

t