Re: pg_xlogdump - Mailing list pgsql-hackers

From Peter Geoghegan
Subject Re: pg_xlogdump
Date
Msg-id CAEYLb_XaEQC1tzar1CmPQPUU=JpgQZoe2SyYQXhcRDRS3CUtGQ@mail.gmail.com
Whole thread Raw
In response to Re: pg_xlogdump  (Peter Eisentraut <peter_e@gmx.net>)
List pgsql-hackers
On 26 February 2013 17:02, Peter Eisentraut <peter_e@gmx.net> wrote:
> Well, this is exactly the problem.  Because of this skeleton idea, most
> external extension modules do not build unless you set USE_PGXS=1 before
> building, because they think that they live in contrib by default, which
> is completely bizarre and user-unfriendly.

repmgr is a popular external extension module. The README actually
suggests moving the entire source tree into /contrib as an alternative
to setting USE_PGXS=1. That does seem kind of weird, and yet I can
understand the train of thought.

My advice to others working on external modules would not be to
generalise from the example of /contrib, but to generalise from the
example of popular external modules. This is particularly important
when targeting multiple Postgres versions.


-- 
Regards,
Peter Geoghegan



pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: Re: "COPY foo FROM STDOUT" and ecpg
Next
From: Heikki Linnakangas
Date:
Subject: Re: "COPY foo FROM STDOUT" and ecpg