wayland-protocols scope and governance

[Simon Ser]

> Thank you all for the work you did on this document. In general I'm in
> favor of the current draft. I have the following rather technical
> concerns, but I guess they can be dealt with quickly:

Thanks for chiming in!

> New members are proposed by current members but no initial members
> have been defined. I think we need that in a single sentence,
> otherwise how could the first members join? I would propose the
> initial members to be the compositors (in alphabetic order): KWin,
> mutter, weston, wlroots/Sway.

I guess we need to maintain a list of members in the repo. We need an
initial list and I also had these 4 compositors in mind. Maybe also GTK
and Qt? But after all, the goal is to just bootstrap the members list.
We shouldn't try to build an exhaustive list, it's easy to add new
members.

Points of contact I had in mind:

- KWin: Roman Gilg <subdiff at gmail.com>
- Mutter: Jonas Ådahl <jadahl at gmail.com>
- Weston: Pekka Paalanen <pekka.paalanen at collabora.com>,
  Daniel Stone <daniel at fooishbar.org>
- wlroots: Drew DeVault <sir at cmpwn.com>, Simon Ser <contact at emersion.fr>

I think it would make sense to include this initial list in a MEMBERS
file in the upcoming wayland-protocols patch series that adds this
governance document.

Thoughts?

> We have defined projects and per project point persons of which there
> are 1 or 2. In case of 2 and these persons do not agree on the same
> vote there should be an automatic fallback defined (for example to
> NOOP in the case of 3.1e). Also the process how these point persons
> are set is not defined. It says the projects choose them, but how to
> confirm that easily? I.e. is there an official way to set them?
> Through a merge request again? Or through an official email address
> per project? The first one is problematic if there is not yet a point
> person defined or the only one can not merge anymore for some reason
> like accident.

Yeah, I guess it doesn't hurt to consider the case where two
point-of-contacts disagree:

    d. During a vote, if two points-of-contact for the same member
       disagree, the member's vote is considered blank.

I think the point-of-contact(s) should come in the merge request that
adds a new member. Regarding the authenticity of the new member
request, would it be enough to rely on common sense and except current
members to reject a request that comes from a non-maintainer?

> On another note on IRC some days ago I also raised concerns in regards
> to the question if there is a quorum defined per vote (member
> inclusion/expulsion and website, document amendments). It was then
> pointed out to me that the described vote majorities always are
> defined in regards to all existing members and not just the voting
> ones what implicitly defines a quorum. I thought this was not clearly
> described in the document, but reading it carefully again I think it
> actually is fine how these votes are defined.

Nevertheless, I added some clarifications to the text. I'll send them
in the next version.