On Wed, Nov 14, 2007 at 01:34:05PM -0400, Marc G. Fournier wrote:
> For all the discussions on why doing this so quickly was such a bad idea, do
> you realize that *so far*, there have been a whole three *active* projects that
> hadn't been moved over? pgweb, pljava and pgjdbc ...
That is completely irrelevant. My point is a simple one: we're a mature
project, and we should act like grown ups with our infrastructure. That
means rather more _specific_ notice to users about when things will go away.
The way it happened, it looked like someone woke up one morning and said, "I
think I'll take down gborg today." If we want people to trust our software
with their critical data, we have to act as though predictability is a good
thing.
Nobody is suggesting that it was ok to have gborg linger the way it did.
All I'm saying is that the next time we shut something down, we need to tell
_everybody_ well in advance exactly _when_ we think things will go away
(emergencies are, of course, excepted). This doesn't mean four-hour
"maintenance windows" at midnight or any of that. But it does mean that,
some weeks in advance of something going away, there should be some evidence
that the changes are planned.
A
--
Andrew Sullivan
Old sigs will return after re-constitution of blue smoke