Re: running script on server shutdown (TODO)

On Sun, 2005-12-11 at 23:54 -0500, Bruce Momjian wrote:

> If I am wrong, please someone tell me why I am wrong. Don't way we
> would like to have X too as well as what we already have, but not supply
> any more manpower. You might as well say you want to live forever, but
> not give any way of accomplishing it.

I am very impressed with the way things run around here. I've observed
at close hand, but externally to, one RDBMS dev organisation and the
problems we have are like nothing in comparison. I've also sat on
various company architecture boards and PostgreSQL would be rated within
the top quartile of those.

Keeping things open makes the whole process less impersonal and better
decisions are made as a result. Keeping things flexible is important too
and we could easily lose that for no gain.

Bruce has expanded the TODO list and does that very even-handedly. A
bugtracker would not make that happen any better, probably worse.

There is one area where we are lacking and that is having an organised
programme for helping and encouraging new developers. My fairly
broad-brush gut feel would be that about 50% of new developers fall foul
of the process somehow and never finish their patches as a result. I
don't believe this thread is an isolated example. In the short term we
could justify that based solely on the time savings for major
developers, though in the long term this could pay off handsomely for
everybody. Regular "Developer Training" by phone, please. Caller pays
their own costs. Something like WebEx...

Best Regards, Simon Riggs