Re: [HACKERS] Proposal for Grand Unified Configuration scheme - Mailing list pgsql-hackers

From Tom Lane
Subject Re: [HACKERS] Proposal for Grand Unified Configuration scheme
Date
Msg-id 19385.952416931@sss.pgh.pa.us
Whole thread Raw
In response to Proposal for Grand Unified Configuration scheme  (Peter Eisentraut <peter_e@gmx.net>)
List pgsql-hackers
Peter Eisentraut <peter_e@gmx.net> writes:
> The idea here is to unify all the various configuration settings into one
> coherent scheme.

A good goal.  Your sketch seems reasonable, but one comment:

> ... For
> example it would be nice if I could start the postmaster with the -F
> option and it would look that up in the grand unified options table (see
> above) and say "ah, that's a per-backend option" and pass it on to the
> backend.

In fact -F is *not* a per-backend option, and certainly we dare not
change it on-the-fly via SET.  The setting is useless and even dangerous
unless all backends are behaving the same way (see pghackers archives if
you've forgotten why).  More generally, some options are reasonable to
set at any time and some aren't; your mechanism needs to deal with that.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [HACKERS] Optimizer badness in 7.0 beta
Next
From: Patrick Welche
Date:
Subject: alter_table.sql