On May 7, 2007, at 7:47 AM, Zdenek Kotala wrote:
> Jim Nasby wrote:
> And you describe current processes based on email communication.
> But if we setup some tracker some process will be changed. I think
> first step is determine what we really want and after we can
> discuss how to reach it.
If we lived in an ideal world I'd agree with you 100%. But we live in
PostgreSQL-community-world. :) There is a *lot* of resistance in the
development community to going to any kind of a tracker, even if it
would mean essentially zero change to how the development has to
work. If you don't believe me go look in the archives; I believe this
debate happens about twice a year, and every time the result is the
same: lots of emails, zero change.
> Create own tracker is reinvent a wheel and waste a time. There are
> a lot of trackers and I believe that one of them fit postgres
> requirements. I agree with your idea to try one and if it will be
> necessary we can add some functionality. But I think that there are
> not clear requirements and I also afraid that there is not unified
> view of core team on this.
Yes, when it comes to doing a full-blown tracker it would be a huge
amount of wheel reinvention. But that's not the case with a simple
patch tracker.
Let's take the baby step of a patch tracker first and see what we
learn from it.
--
Jim Nasby jim@nasby.net
EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)