[GRASS5] Proposal: RFC 1: Project Steering Committee Guidelines

[Frank Warmerdam]

Radim Blazek wrote:
> Hi all,
> 
> I think that all contributors should have a possibility to vote not only
> PSC members. I don't see any point in delegating votes from
> about 10 persons to some 5 PSC members. Why people doing
> a lot of work for GRASS who don't have time for or don't want to become
> PSC members should not participate in decision process?
>
> My proposal is that PSC prepares only a question for voting
> and all 'contributors with vote' can vote about it. The proposal
> passes if majority of 'contributors with vote' votes for the
> proposal (clear majority, no vetos).

Radim,

Well, "contributors that can vote" *are* the PSC.

I would caution against setting things up where you need to wait for
a quorum.  That is, where you depend on actually hearing back from
a majority of members.  In my experience this can cause "stallage"
unless folks are quite diligent about responding to each proposal.  The
approach defined with two +1's in two business days is called lazy
consensus by Apache.  Basically folks need to object if they have an
issue otherwise things proceed normally.  I think for most matters this
is much smoother than requiring a full vote.

You mention "no vetoes".  The intention of the veto rule is that if
someone feels strongly enough to vote again a proposal it is likely
worth trying to revise the proposal to address their concerns.  But there
is always the option of a second vote to override the veto, this
requiring an absolute majority of eligible votes.  The intent is to aim
for consensus, but with a mechanism to settle things if that cannot be
achieved.

Please note that I am just trying to explain the rationale why things
are done the way they are currently in MapServer (and in most/all Apache
projects as I understand it).  GRASS is well within it's rights to
use it's own voting mechanism.

> 'Contributors with vote' will be all contributors (not only developers)
> who did substantial contributions to the project. Initialy it can be persons
> mentioned as candidates for PSC. New 'contributors with vote'
> can be accepted by voting.

I don't see any reason not to name all existing committers, and other
substantive contributors to the PSC as long as they are willing/interested.
As long as you don't need full votes very often a large-ish PSC is not a
bad thing (IMHO).  I think PSC's in Apache are literally all the developers
with commit privileges.

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGF, http://osgeo.org