[torquedev] [patch] bind to ip on multihomed pbs_servers
adi at drcomp.erfurt.thur.de
Tue Feb 12 09:13:05 MST 2008
On Mon, Feb 11, 2008 at 12:58:31PM -0800, Garrick Staples wrote:
> > I'd say let it go into a regularly synced branch torque-ipv6.
> Agreed. If you really want to do this, we'll branch trunk (again) get you
> write access.
I'm not sure if this (alone) would change anything. Matti already has
his own repository (git, not subversion).
It's probably one or two days of work to re-integrate the patch into a
current branch. And then? Falling behind, again?
He definitively needs support, has to negotiate how changes should be
done, hints why things are implemented the way they are (instead of
chasing the rationale through the code, which is time consuming)
>From my point of view (absolutely unimportant, I'm external to the whole
project), it could be better to bring it right to the trunk. This avoids
multiple branches with all the conflicts, synchronization problems
a.s.o. In addition, it forces every developer to work for the new
(global) goal: clean, modern communication code.
Nobody could write pbs_net_t anymore. That's an important question: is
there consensus for such a leap, do other feel the same that things
really need to be changed? It's your community decision ;)
As mentioned earlier, I'd prefer a broken trunk for some weeks to the
mess of parallel branches.
Just my $0.02
mail: adi at thur.de http://adi.thur.de PGP/GPG: key via keyserver
Nur wer die Arbeit kennt, weiß was ich meide
More information about the torquedev