Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> My feeling is that if it's as flakey and unreliable as it currently
> is, we shouldn't ship it. Removing it from CVS doesn't mean "you
> can't use this any more"; this is open source. It just means people
> will have to go and get an old copy out of CVS and presumably in so
> doing they will be aware that we've removed it for a reason. We have
> a well-deserved reputation for quality and I would like to see us
> preserve that.
[ shrug... ] It is not any more flaky than it's been since it was put in.
The people who have been depending on it presumably have use-patterns
for which it doesn't fail, and we're not going to be doing them a
service by ripping out functionality for which we can't offer a
As for the "quality" argument, contrib modules are not guaranteed
to be of the same standard as the core code; anyone who thinks they are
should disabuse themselves of the notion by reading some of that code.
regards, tom lane
In response to
pgsql-hackers by date
|Next:||From: Robert Haas||Date: 2010-02-01 18:40:06|
|Subject: Re: [BUG?] strange behavior in ALTER TABLE ... RENAME TO on inherited columns|
|Previous:||From: Robert Haas||Date: 2010-02-01 18:37:57|
|Subject: Re: plpython3|