On 2013-08-26 21:13:49 +0200, Dimitri Fontaine wrote:
> Andres Freund <andres@2ndquadrant.com> writes:
> > If anything but the proposed feature, it should be an SRF - passing in
> > text parameters isn't very discoverable. And I can't imagine where an
> > SRF would have too high overhead.
>
> The two existing extensions doing that are SRFs, yes. One of them has
> been made by our committer Joe Conway. Let's please have a look how to
> best include existing code into core.
I think the discussion so far is about whether we want to include
something like it into core (tentative yes so far?) and what the
userinterface should look like. Whether we use the proposed patch,
include the Joe's code or write something new seems to be the smaller
part.
Last time round there were some arguments made *against* including all
the fields: http://www.postgresql.org/message-id/7590.1267824977@sss.pgh.pa.us
Personally I don't see much problem with either exposing all of
pg_control or exposing the individual fields as individual functions.
Greetings,
Andres Freund
-- Andres Freund http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services