Hi,
> Hm, would be a relief to have suport for the cursor in the library, as
> this is something most of the developers would want to have but have to
> code on their own each time;
Yeah. We'll try to take care for it soon.
> Ah. but there could be a software emulation for those cards that don't
> provide it;
This needs application help.
> > > if ( ggiEventRead( visual, &theEvent, emAll ) == 0 )
> > Um - why do you use emAll here ?
> To get all pending events out of the queue; also those I'm not interested
> in, they simply get thrown away;
This is possible, but if you do so, this will cause no problem.
> this is what I wont to have; why should I wont to have the queue fill up
> constantly?
> But move events will fill up in the queue;
> So I'll get an overflow of the queue after some time?
No. Actually there are multiple queues, one for each type, and they fill up
independently. It is correct, that the move events will overflow their queue
at some time, but nothing bad will happen. They will just get thrown away.
> No, nothing was drawn, only the cursor position was saved in two
> variables;
> Would make me wonder if there is no hardware cursor, because the Matrox
> Millenium card is one of the best suported cards in the X server.
Yes. That shouldn't be. I have just done a little test:
monitest - Flat panel check draws a solid retangle that moves with the mouse
cursor. I had a very hard time to get the X server (XGGI, which should be
worse, as it uses a software cursor) to 1% CPU here on a PII-300.
Please send me the code, to see if I can find a problem with it.
CU, Andy
--
= Andreas Beck | Email : <[EMAIL PROTECTED]> =
