[PATCH v4 wayland-protocols] virtual-keyboard: Add new virtual keyboard protocol

[Simon Ser]

On July 29, 2018 6:55 PM, Dorota Czaplejewicz <dorota.czaplejewicz at puri.sm> wrote:
> On Tue, 24 Jul 2018 15:10:29 +0200
> Dorota Czaplejewicz <dorota.czaplejewicz at puri.sm> wrote:
>
> > On Thu, 12 Jul 2018 18:15:32 -0400
> > Simon Ser <contact at emersion.fr> wrote:
> >
> > > Hi,
> > >
> > > Sorry for the delay.
> > >
> > > I'm not sure I like this new design.
> > >
> > > Finally, I'm not even sure this security mechanism belongs here. I think adding
> > > this mechanism to potentially all privileged protocols will result in duplicated
> > > code and "pollutes" protocols. Here are some other solutions:
> > >
>
> --- snip ---
>
> > > * Allow other clients to use this interface too, and use another protocol to
> > >   manage authorizations. Basically the idea would be not to expose this
> > >   interface, and require the client to request access to this privileged
> > >   interface through an authorizer protocol. Someone already mentioned it, this
> > >   is Orbital's approach [2]. This allows the compositor to spawn a dialog asking
> > >   to the user "do you want to allow <client> to create a virtual keyboard?".
> > >
> > > What do you think?
> > >
> > I'm in favor of option 3, which will be useful for other protocols as well, which are not necessarily in position to be kept in a permanent privileged process, e.g. screen recording.
> >
> > It would also make me happier as a protocol author, because the protocol becomes more compact and easier to maintain without explicit managing.
> >
>
> Thinking a bit more about the topic, this patch with the authentication features removed would end up exactly the same as patch v3. Patch v3 is already implemented in wlroots too.
>
> Hereby I withdraw this update. Please refer to 20180622152045.10563-1-dorota.czaplejewicz at puri.sm which can be read here: https://lists.freedesktop.org/archives/wayland-devel/2018-June/038629.html

Hi,

While the interfaces, requests and events are the same, there are some wording
changes from v3 to v4 and some enum reference fixes:

- added clarification to the destructor
- namespaced enum references

Is it possible to backport these changes to v3?

Thanks,

Simon

> Regards,
> Dorota Czaplejewicz