[PATCH wayland-protocols v3] Introduce pointer locking and confinement protocol

[Jonas Ådahl]

On Thu, Jan 14, 2016 at 04:25:00PM +0000, Daniel Stone wrote:
> Hi,
> 
> On 13 January 2016 at 02:14, Jonas Ådahl <jadahl at gmail.com> wrote:
> > This patch introduces a new protocol for locking and confining a
> > pointer. It consists of a new global object with two requests; one for
> > locking the surface to a position, one for confining the pointer to a
> > given region.
> >
> > Signed-off-by: Jonas Ådahl <jadahl at gmail.com>
> > Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net>
> > Reviewed-by: Derek Foreman <derekf at osg.samsung.com>
> 
> Hm, seems like we're dangerously close to agreement after however long
> it's taken ...
> 
> > Changes since v2:
> >
> > Added a "lifetime" enum which is passed to the locking/confining requests. It
> > is used to specify whether the constraints should be oneshot or reoccurring.
> > Oneshot and reoccurring both has race conditions when they are deactivated, and
> > this enables the client to choose what race condition it prefers.
> 
> ... so best bikeshed something. Please change 'reoccurring' to either
> recurring, or perhaps 'persistent' as even better.

"Recurring" is according to the dictionary incorrect here, while
"reoccurring" being better suited. But with "persistent" we avoid that
confusion all together, so lets go with that.

> 
> > +      <arg name="lifetime" type="uint" summary="lock lifetime"/>
> 
> Missing enum tag?

Nja. There is no wayland-scanner release which supports that feature.
After wayland-1.10 we should add both this, and the wl_pointer.frame
framing to the relative pointer protocol.

> 
> The rest looks good to me, so given these are minor enough that they
> can be fixed when pushing:
> Reviewed-by: Daniel Stone <daniels at collabora.com>

Thanks; adding your stamp of approval.


Jonas

> 
> Cheers,
> Daniel