Tom Lane wrote:
> I think that what this discussion is really leading up to is that we
> are going to decide to apply the same principle to performance. The
> out-of-the-box settings ought to give reasonable performance, and if
> your system can't handle it, you should have to take explicit action
> to acknowledge the fact that you aren't going to get reasonable
> performance.
What I don't understand is why this is such a huge issue. Set it to a
reasonable level (be it 4M or whatever the concensus is) & let the
packagers worry about it if that's not appropriate. Isn't it their job
to have a good out-of-the-package experience? Won't they have better
knowledge of what the system limits are for the packages they develop
for? Worst case, couldn't they have a standard conf package & a special
"high-performance" conf package in addition to all the base packages?
After all, it's the users of the RPMs that are the real problem, not
usually the people that compile it on their own. If you were having
problems with the "compile-it-yourself" audience, couldn't you just hit
them over the head three or four times (configure, install, initdb &
failed startup to name a few) reminding them to change it if it wasn't
appropriate. What more can you really do? At some point, the end user
has to bear some responsibility...
--
Jeff Hoffmann
PropertyKey.com
