XInput 2: the big picture, relationship to toolkits

[Daniel Stone]

On Thu, Aug 14, 2008 at 10:48:24PM +0930, Peter Hutterer wrote:
> On Thu, Aug 14, 2008 at 02:59:46PM +0300, Daniel Stone wrote:
> > On Thu, Aug 14, 2008 at 11:00:42AM +0930, Peter Hutterer wrote:
> > > Either way, it removes the need for tracking the XEventClass in the
> > > client, making things much saner and closer to core protocol event tracking
> > > from a toolkit/app perspective.
> > 
> > Oh yeah, and while I'm here, did anyone ever find a use for multiple
> > device classes? I haven't yet seen one; just seems like ridiculous
> > overengineering that makes our API more complicated than South Ossetia.
> 
> Assuming you're referring to having two device classes of the same type?

Yep.  BUT MY KEYBOARD HAS TWO BELLS.

> The only thing I can think of here is having two axis classes, one of which is
> absolute, one being relative. But then again, I don't think the server
> supports it, and even on  protocol level it wouldn't work.
> You can't tell which class an event would be coming from, so you'd lose all
> the benefits.

Yeah.  I get the feeling we want to tag axis data as abs/rel anyway,
which would make DGA go away with zero effort[0].

Cheers,
Daniel

[0]: Porting all DGA clients to Xi.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
URL: <http://lists.x.org/archives/xorg/attachments/20080814/4c63a530/attachment.pgp>