On Tue, 2011-03-29 at 21:28 +0300, Tiago Vignatti wrote:
> On 03/29/2011 08:40 PM, ext Aaron Plattner wrote:
> > I disagree.  mi is a toolkit of machine-independent driver support
> > routines.  While I understand that X needs to advance forward and gain
> > features, I don't think it should turn its back on its loyal professional
> > workstation heritage either.
> 
> we wouldn't be turning back, Aaron.
> 
> This is a nice opportunity to remove some big piece of code and 
> eventually open space for even more clean up.

What do you think removing mioverlay will let you fix?  Who does that
benefit, and who does it hurt?

Honestly the overlay support is _not_ simply confined to that one file.
Once you remove it you'll notice that basically every place in dix that
says 'pLayerWin' can be simplified, and if you do so you'll be ripping
out so much that nvidia _won't_ be able to maintain overlay support.  

I think we're coming to the point where it's useful to be able to build
out that kind of subsystem in the name of size or whatever, but the way
you do that is the way XACE handles being disabled.

I know this sounds odd coming from someone who's so actively
pro-deletion, but I've tried to stay close to the rule that whatever I'm
deleting has to either be superfluous or have a functional replacement.
cfb* had fb.  mibank had shadow.  mioverlay has... well, it has
cult-and-paste into the workstation drivers and then hope that we don't
break one of the assumptions it's relying on.  That's less compelling.

- ajax

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
[email protected]: X.Org development
Archives: http://lists.x.org/archives/xorg-devel
Info: http://lists.x.org/mailman/listinfo/xorg-devel

Reply via email to