[ANNOUNCE] predictable pointer acceleration

[Simon Thum]

Tiago Vignatti wrote:
> Simon Thum escreveu:
>> I'm pretty sure it's not - I used it to test my accel's behaviour on 
>> blocking. But ATM it just sucks the server down :( But in general it's 
>> true that X block on a lot less occasions than before say 2 years. 
>> Let's hope you'll get the rest.
> 
> Funny, nothing happened in this venue in the past 2 years (or decade?).
Well it's just personal experience modulo system factors. It doesn't 
even need to be Xorg itself, e.g. scheduler/kernel improvements, IDE DMA 
(I didn't have it for a while due to chipset/kernel issues), gfx drivers 
all play a role here. And xorg's design isn't particularly suited to 
avoid those pitfalls, at least as it is now.
So if I can strike out the side effects of such things on acceleration I 
do it - no need to ask who's the culprit.