[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gnumed-devel] CVS guidelines (or: governance of gnumed)
From: |
Ian Haywood |
Subject: |
Re: [Gnumed-devel] CVS guidelines (or: governance of gnumed) |
Date: |
Tue, 18 Nov 2003 20:30:17 +1100 |
On Mon, 17 Nov 2003 21:31:40 +0100
Sebastian Hilbert <address@hidden> wrote:
> To all !!!!! developers,
>
> Today something happened that never should have.
> Large pieces of code were moved into the main trunk without informing the
> project admins and/or this list.
I agree is this a bad idea, although a I am loathe to "cast the first stone"
(one will find plenty of empty log entries under my name in CVS!)
(As an aside: I am also guilty of spaces/tabs admixture, probably as a
resultof my editor. (Emacs)
I am happy to change: what editors to people suggest?)
I agree with Hilmar gnumed is too early in its development to have a
Linux-style process. What we need to get to that stage is:
1. a "benelovent dictator" or at least "presidium" to co-ordinate things
2. a clearly established codebase with documented internal APIs
3. a clear design endpoint
Prior to 2. Torvalds worked largely in isolation. The *BSDs started in the
pre-internet era with Berkeley students working largely face-to-face.
( and again with the AT&T codebase to work with) Both initally had commerical
Unices to aim for.
Horst is the logical choice for a Linus type role, but AFAIK he is too busy.
So I would propose a "gnumed junta"
1. Karsten: backend database schemas, bootstrapping of same
2. Myself: middleware layer
3. Syan: wxPython frontend
4. Sebastian: packaging
All major commits on these areas must either be by this person or with their
approval.
I am trying not to be prescriptive or exclusionary here: volunteers for other
domains/alternative divisions are welcome
naturally we should still discuss major design decisions on this list.
Ian
--
PGP public key E750652E at wwwkeys.pgp.net
9BF0 67B7 F84F F7EE 0C42 C063 28FC BC52 E750 652E