[Gdbheads] Proposed poll on proposals

[Ian Lance Taylor]

The gdbheads discussion seems to have slowed down over the weekend.  I
hope that has given everybody some time to reflect.

Robert Dewar mentioned using straw polls to find out how people feel
about an issue.  I propose doing one now.  I read through all the
e-mail, and found six specific proposals, from different people, for
changes to gdb maintenance.  They are listed at the bottom of the
e-mail message.

I suggest that all interested parties vote yes or no on each of these
proposals.  They are somewhat though not wholly independent.  I
suggest that you vote for the proposals you really want to see
happen--don't vote yes for the ones which you would like to see happen
only if others don't happen.

I suggest that people send votes either to the mailing list or
directly to me.  I will gather all the votes, and report total counts,
and also a list of all the people who voted.  I won't specifically say
how people voted for those votes sent directly to me.  I won't vote
myself.  Posting the list of people who voted will also indicate those
people who did not vote, which will help see how seriously we should
take the results.

If anybody objects to me as a vote counter, I expect that we can find
somebody more objective and uninvolved.

This poll is obviously completely non-binding.  Indeed, as I said
before, I don't think voting is a particularly good way to make
maintainership decisions.  The only purpose of the poll is to discover
the overall sentiment, if we can.  Therefore, anybody reading this is
allowed to vote; I won't make any attempt to restrict the poll to just
active gdb maintainers, or steering committee members, or anything
like that.  For that matter, don't feel that your vote commits you to
anything; you can change your mind at any time.

I don't know whether this will tell us anything.  But I think it is an
interesting experiment, and it's possible that the results will be
helpful in refining the ideas.

I'll be in meetings all today tomorrow, and will post a summary of any
results late tomorrow (Tuesday) evening (EST).

Ian

Here are the proposals, in the order in which I found them by reading
the discussions with a threaded e-mail reader:

1) Global maintainers are permitted to approve patches to any part of
   gdb, even parts which have a specific area maintainer.

2) If maintainers disagree on a patch, and can not resolve the
   disagreement by discussion, the disagreement may be resolved by a
   vote of all global maintainers and relevant area maintainers.

3) Assign more area maintainers to areas for which patches are not
   being reviewed fast enough.

4) Track patches using an automated system; permit global maintainers
   to approve a patch if the relevant area maintainers have not acted
   for N days, for some value of N.

5) Create an ombudsman for gdb, who is responsible for listening to
   people's reports of problems, and then resolving them first through
   the global maintainers, and then, if that fails, through the
   steering committee.

6) Reconstitute the steering committee to make it more active and to
   make it better reflect the gdb community.  Include people from the
   gcc and binutils projects.  Perhaps do not include current gdb
   maintainers, to permit more objectivity.