[haiku-appserver] Re: BWindowScreen / pageflipper once more

  • From: "Rudolf" <drivers.be-hold@xxxxxxxxxxxx>
  • To: haiku-appserver@xxxxxxxxxxxxx
  • Date: Sat, 06 May 2006 23:33:58 +0200 CEST

Hi :)

I saw the commit. So did you also verify pageflipper now works 
correctly? (just curious :)
Tomorrow I'll update my copy of SVN and have a look :)


BTW About the cursor: I did this in a testapp:
-setup BwindowScreen: all modes are 60Hz however.
-once it shows, call BSCreen::Setmode to set 70Hz. At this point the 
cursor shows, which I think should not have been influenced by the 
BScreen call.
(app server could check of the call was done within a BWindowScreen 
screen, and dont' touch the cursor then?)

I think this kind of use of Bscreen::Setmode inside on a workspace 
that's occupied with fullscreen BWindowScreen is a valid use. Apart 
from the cursor it indeed works as expected in R5/dano. Any thoughts 
about that?



> "Rudolf" <drivers.be-hold@xxxxxxxxxxxx> wrote:
> > > The mess can easily be drawn on screen when moving the mouse 
> > > cursor 
> > > around.
> > > This one probably shouldn't be drawn in that case (uh, soft 
> > > cursor) 
> > > :-)
> > Ah, good one. Though I don't know about the cursor exactly. I mean, 
> > it 
> > should probably be shut-off indeed, but I could imagine it's 
> > functions 
> > should be available to apps using BWwindowScreen: i.e. an app 
> > should 
> > be 
> > able to re-enable it and use it? I don't know.
> The cursor is now hidden just once, so that an application can easily 
> show it again if it needs it.
> > That thought crossed my mind as well (looking at the distorted 
> > curled 
> > lines). I didn't see any proof of that looking at the driver log 
> > though, or it must be a missing palette in the driver: that could 
> > be 
> > it, the system should probably set it. Pageflipper uses 8 bit mode 
> > AFAIK.
> In case you've missed the commit message: it's fixed now :-)
> Bye,
>    Axel.

Other related posts: