Re: pgsql-server: > Please find enclose a submission to - Mailing list pgsql-committers

From Tom Lane
Subject Re: pgsql-server: > Please find enclose a submission to
Date
Msg-id 27804.1093841589@sss.pgh.pa.us
Whole thread Raw
In response to Re: pgsql-server: > Please find enclose a submission to  (Peter Eisentraut <peter_e@gmx.net>)
Responses Re: pgsql-server: > Please find enclose a submission
List pgsql-committers
Peter Eisentraut <peter_e@gmx.net> writes:
> I've said several times before that I did not particularly like the
> functionality added by that patch (building non-server modules, and
> building contrib modules outside the normal build system).  Therefore,
> I didn't put it high in the to-look-at queue.  It might help if someone
> else would comment on whether we want this.

Well, I think it's important to be able to build both server-side and
client-side modules without access to the original build tree.  In an
RPM-style environment you should only need the files installed by the
"foo-devel" RPM to build packages that depend on package foo.  We are
clearly not there as of current releases, and I was hoping that pgxs
would fix this.

As for contrib in particular, it is not *necessary* to be able to build
the contrib tree outside the original build tree --- but if we don't
support that then we don't have an easy part-of-the-distro way to test
that the pgxs functionality really works.  So I think it makes sense to
be able to do that.  I am not any keener than you on this particular
approach ... the nesting of ifdefs in the makefiles seems mighty ugly.
But I'd like to find a way to clean it up, not just claim that we don't
need to do it.

            regards, tom lane

pgsql-committers by date:

Previous
From: tgl@svr1.postgresql.org (Tom Lane)
Date:
Subject: pgsql-server: Tweak md.c logic to cope with the situation where WAL
Next
From: dpage@pgfoundry.org (User Dpage)
Date:
Subject: pginstaller - pginst: Zero memory prior to use.