Re: Draft of Proposal for the GNOME Foundation.

[Maciej Stachowiak <mjs eazel com>]

Nat Friedman <nat@helixcode.com> writes:

> Maciej Stachowiak <mjs@eazel.com> writes:
> 
> > Some comments below.
> > 
> > Nat Friedman <nat@helixcode.com> writes:
> > 
> > >        In the past, being a part of the GNOME project has simply meant
> > >        "I wrote  some code" or "I hang  out  on the mailing  lists and
> > >        build the  thing  from  CVS  frenetically every  three  hours."
> > >        There is no reason to change this.
> > 
> > Yes there is. Can you honestly say that having a debate on gnome-list
> > is a useful decision-making process?
> > 
> > Anyway, I think more than not capturing the consensus, the idea that
> > anyone who wants to should have a voice in decision-making is pretty
> > opposite to most opinions expressed so far.
> 
> Yeah, I agree that it has to be a bit more than "I want in."  But the
> bar needs to be pretty low.  Raise it above gnome-list if you want,
> that's fine.
> 
> What I described was inspired by what Jim mentioned about the IETF,
> and by the knowledge that an "exclusive" membership is a Bad Thing.

The IETF is a structure where totally open membership has more or less
worked - although I have heard complaints of some working groups being
stacked by companies that can pay large numbers of people to advocate
their view. However, more restrictive models of membership have worked
for other organizations, such as the ASF. It seemed to me many people
were more interested in an ASF-like model than IETF-like.

> > Does that mean that no module is part of GNOME until it becomes part
> > of a GNOME release (and that therefore people working on it are not
> > per se working on GNOME)?
> 
> I answered this question in detail in my reply to Havoc.

I saw your answer, where you said that something being not
release-worthy should not exculde it from being part of
GNOME. However, this removes the property that "making a GNOME
release" == "defining GNOME", since things not yet in any release or
even necessarily planned for one would still be part of GNOME.
 
> > "virtual global entity" == "no actual legal existence"?
> 
> Right.  This was what Bart proposed.  It would be represented locally
> by national GNOME foundation points of presence.  These would be used
> whenever a legal entity in a given jurisdiction is needed: for the
> transfer of funds, for example.
> 
> It does sound a little crazy, I know.  Bart?

Yeah. I too think it's crazy, in part because it means it would be
impossible for the board to, e.g. manage trademarks or hold
copyrights.

> > I don't think "any contribution" is the right place to set the
> > threshold. When I get a 5-line patch I don't want to have to worry
> > about whether I'm giving that person full GNOME voting priveleges
> > automatically if I check it in.
> 
> I think maybe one problem you're experiencing here is that there's no
> distinction between 'belonging' and 'voting.'  You don't want to
> exclude people from belonging because you don't want to alienate them.
> But then again, you don't want uninformed people to vote.
> 
> This is a classic democratic problem, isn't it?  The only really
> important decision the voting proletariat will make is the slate they
> choose for the board.  The board will make all the other decisions.  A
> substantial amount of organization would have to occur in order for a
> referendum to actually be issued and pass.  The idea is that referenda
> are only used rarely, in the case where the public is outraged, or
> something, and their feeling is not harmonious with that of the board.
> 
> So I don't think there's anything really bad about having a board
> voting base.

I don't think everyone who participates needs an official stamp of
`belonging'. GCC has many contributors who are not on the steering
committee or any other official power structure. Apache has many
contributors who are not ASF members (though with consistent valued
contributions over time they become members). I don't see a problem
with granting priveleges only after trust is earned.

> > >     Miguel will be the chairman and will preside  over all meetings of
> > >     the  board, unless he  is declared legally  insane and  "fit to be
> > >     tied" by the UN or the Pope.
> > 
> > I hate to appear to speak against Miguel (since he's the reason we're
> > here more than anyone else) but I think this clause is lame system
> > design. We didn't make George Washington President for Life.
> 
> Actually, they tried to.  But he refused.

And that was a very wise decision on his part, I hope we will all agree.

> So what are we doing for Miguel, then?

I expect that the natural process of electing a board will result in
boards that include Miguel. I'd be surprised if a slate that did not
include him could ever be elected. If it was, then it could only be
because he had lost his mandate to lead in some flagrant way, in which
case any kind of permanent position would not be appropriate. I think
it's unlikely that will come up, however.

> > >     In order for a referendum  to pass, 1/3rd  of the total membership
> > >     must  participate, and 2/3rds   of the participating members  must
> > >     approve.  There will be a mailing list for all of the members, and
> > >     all  referenda   must be announced  to the   list by the initiator
> > >     before they are opened on the voting system.   At least three days
> > >     must pass before  the referendum is closed,  and no referendum can
> > >     remain open for  longer than seven  days.  [ These numbers are, of
> > >     course, eminently debatable. ]
> > 
> > I don't think we should allow any referendums other than recall of the
> > board.
> 
> What if the board unfairly decides to exclude a piece of software from
> a release?  A recall of the board at that point -- right before a
> release -- could be disasterous for the whole project (not that the
> kind of lack of unity/harmony required to reach this point isn't!).  A
> referendum to force the inclusion of the package would be a lot
> cleaner.

If the board were reversed on a major decision, it would seriously
undermine their ability to have any kind of voice of authority. It's
also necessary for the board to sometimes make individual decisions
that many may disapprove of, even though their judgement as a whole is
still trusted. This is the essence of representative
democracy. Allowing the membership at large to micromanage every
decision would result in endless debate and indecisiveness over every
decision, which is exactly what having a board is designed to avoid.

If the board no longer represents the will of the people, let it be
removed, even if that means delays, rather than crippled by constant
reversals of it's decisions.

> >      
> > > V. Release Engineering / Defining GNOME
> > > 
> > >   The  board of directors   will  be responsible for   authorizing the
> > >   release of a new version of GNOME.  The board will determine the set
> > >   of  modules which will  make  up the release   at  least 60 days  in
> > >   advance of the  release date, subject  to unanimous  approval of the
> > >   module maintainers.
> > 
> > Requiring unanimous approval for anything is not good. Setting fixed
> > date limits is also not good, this should be up to the judgement of
> > the board and the release coordinators (what if we decided to have
> > releases 30 days apart)? Also writing stuff like this into the bylaws
> > might make trouble later. For example, if we wanted to drop a module
> > from the release, would that mean we'd have to delay is 60 days?
> 
> Yeah, but how do you do a release if the module maintainers aren't
> going to cooperate?

Of course we must develop consensus around a schedule, but that's very
different from requiring unanimous approval.
 
> > In addition to the specific objections, I don't think this proposal is
> > very much in the spirit of even the very rough consensus we do have
> > now.
> 
> Really?  Why not?  Here's what I see, from reading the list:
> 
>     * Havoc and Maciej are worried about giving power to too many
>       people, and having an inclusive, large body in the foundation.

Other people have expressed similar concerns, though perhaps not on
this list. There have also been even more anti-democratic proposals
made, such as the board appointing itself initially and appointing
it's own successors. 

I think having a not completely trivial standard for membership
strikes a good balance.

 - Maciej