Re: [Savannah-hackers] Re-register all new projects?

[Elfyn McBratney]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tuesday 04 May 2004 21:06, Rudy Gevaert wrote:
> On Tue, May 04, 2004 at 04:08:10PM -0400, Richard Stallman wrote:
> > Rudy suggested that, when Savannah was capable of starting
> > new projects again, it would be a good idea to ask everyone
> > who wanted a new project to register it again.
> >
> > Now Savannah can once again accept new projects.  Would you
> > rather ask people to re-register, or would you rather operate
> > on the old registrations?  We can do it whichever way you
> > find easiest.
>
> Another way to handle is: mail everyone who has a project in the
>       queue, and ask them to send a mail to address@hidden if
>       they want there project on savannah.  If they don't, let them
>       mail to address@hidden  Whe then serve people who mail to
>       x1 on a fifo basis.  After a couple of weeks we clear all
>       projects who haven't answered.
>
> I would prefere to clear everything as this will be the easiest for
> all projects.
>
> When a project gets submitted, we supply an url so they can reregister
> easily.  We should be able to retrieve that url so they can reregister
> very easy.

I'd prefer that (clearing the queue now), too.  Thinking about it, ~50% of the 
projects in the queue would probably have gone elsewhere by now (some of the 
projects have been there for ~4-5 months).

If we were to collect all email addresses from the pending queue, clear the 
queue and then send emails to all registrants welcoming them to re-register 
their project if they so choose.  How does that sound?

However, I think we still have some things that needs polishing before we 
start accepting new projects.  A few things come to mind:

  - web front-end to manage projects' CVSROOT/
  - web front-end to manage projects' download area
  - update FAQ

Also, one thing that bothers me about starting to accept new projects again is 
that the savannah.gnu.org machine is very heavily loaded, and, IMO, isn't 
scaling very well for it's current projects.  There are far too many anoncvs 
users (in terms of resources) right now, which if limited might even bring 
back some of that real-time feeling ;)

I've also found it very hard investigating why branch-accesses on both the 
automake and gcc trees are broken as the machine blocks (sometimes for 
lengthy periods) constantly.

Elfyn

- -- 
Elfyn McBratney, EMCB
mailto:address@hidden
http://www.emcb.co.uk/

PGP Key ID: 0x456548B4
PGP Key Fingerprint:
  29D5 91BB 8748 7CC9 650F 31FE 6888 0C2A 4565 48B4
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAmBVdaIgMKkVlSLQRAkxLAJ9zGRe8wejmUojqPdGsvbp23cbErwCeLicH
WRm/3jeWJpYS2ww/TlKb5AM=
=Mq7N
-----END PGP SIGNATURE-----