[haiku-appserver] Re: new drawing bug

Axel Dörfler wrote:

It doesn't have to be related - one looks like a clipping problem, the other one seems to be related to coordinate conversions; maybe it's even both.

Yes, I don't know yet.

I think all these problems have something in common, and I will investigate these problems in the next days until I find out what the hell is going on.
Please excuse me for these problems, I'm just trying to make all code easier to understand and use. Sorry, again.

That effort is surely appreciated, and maybe the drawing bugs are also related to many people working on the same component; it would be nice if this could be avoided, but as long as it doesn't take over, I guess we have to live with it.
Sooner or later the app_server code will be more stable,

I too wait for that moment. Also, I really want to be done with this clipping code. Although I like working at it (it's interesting really :) ), I'm sick of giving you problems. ATM all my efforts will be concentrated into getting this thing done!


BTW after the font stuff, I plan to work on the event handling, working my way through to RootLayer.

You mean the cursor thread and system keyboard shortcuts? You you like to explain a bit? Thanks.


        You know what I'll be doing: clipping and redrawing stuff! :)

On that subject... Testing Haiku in QEmu left me a bad impression. If I move a window over another one, that one from the back won't update until I finished moving the front window. I don't like that, and that all the fault of RootLayer locking.
I think we should somehow make a region-based lock or at least make a drawing queue(in HWInterface) in which we can push drawing instructions which don't intersect with the region-selected-for-rebuild.
What do you think?



bye, Adi.

Other related posts: