Re: interesting article on XFree86 speed

On Mon, Oct 28, 2002 at 12:59:59PM -0500, Owen Taylor wrote:
| 
| Where you get into trouble are apps like Acroread, Mozilla, a paint
| program with a lot of layers, a word processor ... where there are
| significant computations involved in the bits displayed on the screen,
| and when resizing, where a resize can involve a lot of recomputation.
| ...
| If you imagine something like an opaque resize of a text widget, you can 
| see why 10-20ms can be very demanding ... resizing a window 1 pixel
| wider might involve a complete relayout of a megabyte of text.

I don't see that there's much we can do (other than at the application
level) to help in cases like that.  Not even full backing store for
top-level windows does the trick, since whatever the backing-store
previously contained will be inappropriate for the new window size.

Backing store could help in limited circumstances, when windows change
visibility without being reconfigured.  But maintaining the backing
store has costs that are potentially large (especially if graphics
memory becomes overcommitted), and is subject to interesting race
conditions of its own.  Identifying bad practices and teaching app
developers to provide an intelligent response to exposure events seems
like a more robust solution.  The protocol dumper that Jim mentioned
sounds like a good starting point.

Allen