[PATCH wayland-protocols 1/2] Introduce wp_relative_pointer interface

[Derek Foreman]

On 17/11/15 04:09 AM, Jonas Ådahl wrote:
> The wp_relative_pointer interface is an extension to the wl_pointer
> interface created from wl_seat. It has the same focus, but adds the
> functionality of sending relative pointer motions unhindered by
> constraints such as monitor edges or other barriers. It also contains
> unaccelerated pointer motion information.
> 
> Signed-off-by: Jonas Ådahl <jadahl at gmail.com>
> ---
> 
> There has alreay been RBs on this protocol, but I'd like to get them
> explicitly re-RB:ed considering that the protocol has changed some, even
> though the changes have not been major ones.
> 
> 
> Changes since the last version sent as weston patches:
> 
> Interfaces and protocol renamed according to wayland-protocols naming
> conventions.
> 
> Accelerated and unaccelerated motion deltas have the data type
> wl_fixed_t now again. 64 bit fixed point deltas was deemed unnecessary,
> and wl_double_fixed_t was dropped.
> 
> 64 bit timestamp argument name suffix changed from msb/lsb to hi/lo.
> 
> Various other minor changes.
> 
> 
> Jonas
> 
> 
>  Makefile.am                                        |   1 +
>  unstable/relative-pointer/README                   |   4 +
>  .../relative-pointer-unstable-v1.xml               | 122 +++++++++++++++++++++
>  3 files changed, 127 insertions(+)
>  create mode 100644 unstable/relative-pointer/README
>  create mode 100644 unstable/relative-pointer/relative-pointer-unstable-v1.xml
> 
> diff --git a/Makefile.am b/Makefile.am
> index f1bac16..2db6bd5 100644
> --- a/Makefile.am
> +++ b/Makefile.am
> @@ -5,6 +5,7 @@ nobase_dist_pkgdata_DATA =							\
>  	unstable/text-input/text-input-unstable-v1.xml				\
>  	unstable/input-method/input-method-unstable-v1.xml			\
>  	unstable/xdg-shell/xdg-shell-unstable-v5.xml				\
> +	unstable/relative-pointer/relative-pointer-unstable-v1.xml		\
>  	$(NULL)
>  
>  pkgconfigdir = $(libdir)/pkgconfig
> diff --git a/unstable/relative-pointer/README b/unstable/relative-pointer/README
> new file mode 100644
> index 0000000..64c42a1
> --- /dev/null
> +++ b/unstable/relative-pointer/README
> @@ -0,0 +1,4 @@
> +Relative pointer protocol
> +
> +Maintainers:
> +Jonas Ådahl <jadahl at gmail.com>
> diff --git a/unstable/relative-pointer/relative-pointer-unstable-v1.xml b/unstable/relative-pointer/relative-pointer-unstable-v1.xml
> new file mode 100644
> index 0000000..0411d1f
> --- /dev/null
> +++ b/unstable/relative-pointer/relative-pointer-unstable-v1.xml
> @@ -0,0 +1,122 @@
> +<?xml version="1.0" encoding="UTF-8"?>
> +<protocol name="relative_pointer_unstable_v1">
> +
> +  <copyright>
> +    Copyright © 2014      Jonas Ådahl
> +    Copyright © 2015      Red Hat Inc.
> +
> +    Permission is hereby granted, free of charge, to any person obtaining a
> +    copy of this software and associated documentation files (the "Software"),
> +    to deal in the Software without restriction, including without limitation
> +    the rights to use, copy, modify, merge, publish, distribute, sublicense,
> +    and/or sell copies of the Software, and to permit persons to whom the
> +    Software is furnished to do so, subject to the following conditions:
> +
> +    The above copyright notice and this permission notice (including the next
> +    paragraph) shall be included in all copies or substantial portions of the
> +    Software.
> +
> +    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> +    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> +    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
> +    THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> +    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
> +    FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
> +    DEALINGS IN THE SOFTWARE.
> +  </copyright>
> +
> +  <interface name="zwp_relative_pointer_manager_v1" version="1">
> +    <description summary="get relative pointer objects">
> +      A global interface used for getting the relative pointer object for a
> +      given seat.

for a given pointer?  (I don't see any place here where we're passing a
seat...)

> +
> +      Warning! The protocol described in this file is experimental and backward
> +      incompatible changes may be made. Backward compatible changes may be added
> +      together with the corresponding interface version bump.  Backward
> +      incompatible changes are done by bumping the version number in the
> +      protocol and interface names and resetting the interface version.  Once
> +      the protocol is to be declared stable, the 'z' prefix and the version
> +      number in the protocol and interface names are removed and the interface
> +      version number is reset.
> +    </description>
> +
> +    <request name="destroy" type="destructor">
> +      <description summary="destroy the relative pointer manager object">
> +	Used by the client to notify the server that it will no longer use this
> +	relative pointer manager object.
> +      </description>
> +    </request>
> +
> +    <request name="get_relative_pointer">
> +      <description summary="get a relative pointer object">
> +	Create a relative pointer interface given a wl_pointer object. See the
> +	wp_relative_pointer interface for more details.
> +      </description>
> +
> +      <arg name="id" type="new_id" interface="zwp_relative_pointer_v1"/>
> +      <arg name="pointer" type="object" interface="wl_pointer"/>
> +    </request>
> +  </interface>
> +
> +  <interface name="zwp_relative_pointer_v1" version="1">
> +    <description summary="relative pointer object">
> +      A wp_relative_pointer object is an extension to the wl_pointer interface
> +      used for emitting relative pointer events. It shares the same focus as
> +      wl_pointer objects of the same seat and will only emit events when it has
> +      focus.
> +    </description>
> +
> +    <request name="destroy" type="destructor">
> +      <description summary="release the relative pointer object"/>
> +    </request>
> +
> +    <event name="relative_motion">
> +      <description summary="relative pointer motion">
> +	Relative x/y pointer motion from the pointer of the seat associated with
> +	this object.
> +
> +	A relative motion is in the same dimension as regular wl_pointer motion
> +	events, except they do not represent an absolute position. For example,
> +	moving a pointer from (x, y) to (x', y') would have the equivalent
> +	relative motion (x' - x, y' - y). If a pointer motion caused the
> +	absolute pointer position to be clipped by for example the edge of the
> +	monitor, the relative motion is unaffected by the clipping and will
> +	represent the unclipped motion.
> +
> +	This event also contains non-accelerated motion deltas. The
> +	non-accelerated delta is, when applicable, the regular pointer motion
> +	delta as it was before having applied motion acceleration and other
> +	transformations such as normalization.
> +
> +	Note that the non-accelerated delta does not represent 'raw' events as
> +	they were read from some device. Pointer motion acceleration is device-
> +	and configuration-specific and non-accelerated deltas and accelerated
> +	deltas may have the same value on some devices.
> +
> +	Relative motions are not coupled to wl_pointer.motion events, and can be
> +	sent in combination with such events, but also independently. There may
> +	also be scenarious where wl_pointer.motion is sent, but there is no
> +	relative motion. The order of an absolute and relative motion event
> +	originating from the same physical motion is not guaranteed.

It still makes me a little grumpy that we're potentially going to post a
pile of absolute events (potentially 1000+ per second with a ridiculous
gamer mouse) that a client ignores, but still get sent and trigger a
callback.  I think it would be nice to have a way to disable the
absolute motion events, possibly as a side effect of creating a relative
pointer.

> +
> +	If the client needs button events or focus state, it can receive them
> +	from a wl_pointer object of the same seat that the wp_relative_pointer
> +	object is associated with.

button presses are still strictly ordered though, right?  That is, I
might see

rel
abs
button

or:
abs
rel
button

but I'll never see
abs
button
rel

for a situation where the button occurred after the rel motion?

I liked the discussion about framing rel pointer events inside
wl_pointer.frame... I think that's something that can be bolted on later
without breaking any backwards compatibility?

I guess a stream where wl_pointer didn't generate anything and
wp_relative_pointer did could potentially look weird to an application
not expecting it.

I hope there won't be any problems stemming from button presses coming
with 32 bit timestamps while rel motion comes with 64 bit.  These could
appear to be non-monotonic to a client?

> +      </description>
> +
> +      <arg name="utime_hi" type="uint"
> +	   summary="high 32 bits of a 64 bit timestamp with microsecond granularity"/>
> +      <arg name="utime_lo" type="uint"
> +	   summary="low 32 bits of a 64 bit timestamp with microsecond granularity"/>
> +      <arg name="dx" type="fixed"
> +	   summary="the x component of the motion vector"/>
> +      <arg name="dy" type="fixed"
> +	   summary="the y component of the motion vector"/>
> +      <arg name="dx_unaccel" type="fixed"
> +	   summary="the x component of the unaccelerated motion vector"/>
> +      <arg name="dy_unaccel" type="fixed"
> +	   summary="the y component of the unaccelerated motion vector"/>

Utterly worthless bike shed painting:
Personally, I'd like to see the unaccelerated deltas posted as "dx" and
"dy" and the accelerated ones as "dx_accel" or "dy_accel", since if
you're using the rel pointer interface there's a good chance you're
writing a first person game where accelerated deltas are terrible.

Qualifying all the names would be fine too, and have nothing simply
called "dx" and "dy", then developers are more likely to think for 5
seconds instead of just grabbing dx and dy.

Anyway, low priority whining aside, this looks really good to me, and
closes a functional hole we've had for far too long.

Reviewed-By: Derek Foreman <derekf at osg.samsung.com>


> +    </event>
> +  </interface>
> +
> +</protocol>
>