Re: [gtk-vnc-devel] gtk-vnc trouble

[Anthony Liguori <anthony codemonkey ws>]

On Wed, 2007-08-22 at 17:30 +0100, Daniel P. Berrange wrote:
> On Wed, Aug 22, 2007 at 06:25:11PM +0200, Gerd Hoffmann wrote:
> > >> Oooh - are you running against a QEMU vnc server perhaps ?  QEMU supports
> > >> a 'relative mouse' RFB protocol extension where the motion coords on the
> > >> wire are treated as relative instead of absolute. It looks like we always
> > >> enable support for that extension if the server accepts it. And if this
> > >> is activated, then the pointer will be grabbed no matter what. 
> > > 
> > > Yes, I think that should be the behavior too otherwise the extension
> > > isn't terribly useful :-)  It probably makes sense to be able to disable
> > > the extension though.
> > 
> > Also qemu should probably *not* provide that extention when started with
> > "-usb -usbdevice tablet", right?
> 
> Yeah, or made configurable. With the series of patches I sent for adding
> password auth, and TLS encrption the -vnc arg takes a list of comma 
> separated options, eg
> 
>     -vnc localhost:1,passwd,x509

The problem isn't the extension on the server.  I think we just have to
adjust what we're doing on the client.  The extension just provides more
information to the client.  It's better to just do the right thing on
the client than to hack the server such that it doesn't give the client
the extra information to prevent it from doing the wrong thing.

Regards,

Anthony Liguori

> We could easily add another option to enable relative mode or not. So if
> the app  launching qemu was providing some form of absolute pointer, or
> was intending to add one via the monitor, then it could skip the relative
> pointer.
> 
> The other possibilty would be some way to hook into the virtual drivers, 
> so that the VNC server could detect / query which pointer was activated
> by the guest OS. This would help in the case of the guest OS activating
> the VMWare Mouse, vs the plain PS2 mouse - both are always available in
> recent QEMU.  Whether we can detect this is an open question I can't 
> answer though....
> 
> > Oh, and being able to query absolute vs. relative coordinates mode would
> > be nice too (for feedback in the gui).
> 
> Yep, easy enough to add.
> 
> Dan