Re: meeting

[Tom Tromey <tromey at redhat dot com>]

Tom> I reset the bridge using a trick to try to work around its
Tom> timezone woes.  I used this successfully once last week -- I hope
Tom> it works again tomorrow.

It did, yay.

Tom> My agenda items are:
[...]

Some outcomes:

* Thiago is happy with the decision.  As far as we can tell he was the
  only non-RH person on the call.

* There is some confusion about our relationship to gdb.  This is
  understandable, IMO, since it is a bit vague.  Basically I think the
  best way to think about this project is that it is a development
  branch with reasonably specific goals (see the earlier threads).

* Speaking of the goals, an action item for everyone is to look at the
  roadmap and see (1) if anything is missing, and (2) what you are
  interested in working on.

  For #1, Sami asked about the state of non-stop multi-thread
  debugging.  There are patches on the list.  Andrew asked about the
  multi-process work, but we don't really know enough about it yet --
  the discussion on the gdb list seems to be preliminary
  investigation.

* There was general consensus that we should not reuse the frysk list
  for this work.  So, we will set up a new list.  Project name ideas
  that I remember:

  gdb--

  Hmm, that is the only one I wrote down, but I know Phil had another
  one.

  "--" seems a bit dismal to me, but "++" seems a bit arrogant :)
  How about "Project Pelican"?

  Or anything else.  Please.

  I'd like to settle this today, so...  respond.

* Where to host?  Lots of hosting choices out there, but sourceware
  seems like the default.  We all have accounts, we have access, etc.
  I'd like to get things set up ASAP, say today.

* Talked about source control some.  Jim Meyering is setting up a git
  mirror of gdb CVS.  We'll use this as our upstream and have our own
  git repository.

  Andrew brought up gdb's eventual move to svn.  We can revisit our
  choices if/when that happens.

* We talked about our planned process.

  The basic change is introducing universal patch review.  The scratch
  idea is:

  - All patches must be reviewed by someone other than the author.
  - I forgot to mention this, but Apache-like, a strong objection
    should stall a patch until a rough consensus is reached.
  - Anybody "in the project" can review a patch.  In fact, I think it
    is pretty important that everybody do reviews.
  - Proposed patch review guidelines:
    * Does it have internal documentation (comments)?
    * Does it follow upstream coding style?
    * Does it have external documentation, if needed?
    * Does it have a test case, if needed?
    * Is it clear/complete/etc?

  Andrew asked about how we will decide to accept new contributors
  into the fold.  I think we'll solve this when it comes up.

* Action items, for me:
  - Set up hosting, mailing list
  - Send consolidated roadmap to the new list
    (Probably process stuff too)
  - Send announcement to gdb list

Tom