(Apaku: sorry to hijack your post, but this is where I started writing.)

First off: people still seeing the background change, please rebuild 
kdeui, this should be disabled as of Friday.

Pino: be reasonable. I developed on my computer. I tested on my computer 
and was satisfied with the results. I proposed merging my results to 
trunk (providing the diff of what I was merging). *No one objected*. At 
what point exactly does this process break down? If people can't be 
bothered to apply a ten-line patch, what makes you think they are going 
check out and build *an entire separate branch*? The end result would 
have been the same.

This is a two-way street; if the only way I can get people's opinions is 
to shove something down their throat, well then don't be surprised when 
I do that, and when I get irritated that people get all hostile when 
I've *asked* for opinions before doing anything. Especially when I find 
out *second hand* that people are annoyed. (IOW, if you have a 
complaint, /post it to the bloody thread instead of bitching about it 
behind my back/.)

I still think my only regret is not making a change sooner. Pinheiro 
asked for it, and since I saw nothing wrong with it, and since it seems 
people can't be bothered to look at something until it is forced upon 
them, I don't see what could have gone differently... except that if I'd 
waited until the kcm was done to make the change, we would be having 
this conversation *in November*.


(And, changing gears again...)

I do think there is some bikeshedding going on here also. Some of the 
arguments I am seeing seem, as pinheiro noted, to be overlooking the 
"for testing" part of this whole thing. Yes, I did something extreme; 
the idea was to say "hey, people are going to be able to do this, so you 
might want to keep it in mind". Part of the idea was to expose bugs, and 
it did that... including some in Qt!

When all is said and done, I don't really care what the default is, 
because it's *the default*, and I don't have to use it if I don't want 
to. People concerned about performance can disable (or, more likely, not 
enable) the effects. People saying "this is kwin's job", well I've tried 
to mention where kwin falls down but I've also tried to explain that /I 
want kwin to pick up the workload where it can/. (I'm not sure if this 
will happen in 4.0 but I don't foresee any problems hacking on it for 4.1.)

(Pino: I don't have access to my development machine right now, but I'll 
try to take a look at what is going on with Okular when I can.)

-- 
Mathew
(sorry, .sig file is on the other computer)