Re: Extension Templates S03E11

Hi,

Jeff Davis <pgsql(at)j-davis(dot)com> writes:
> In the CF app, this is marked "Ready for Committer". That's a bit vague
> here, considering Dimitri, you, Peter, and Alvaro are all committers.
> Who is this patch waiting on? Is the discussion concluding, or does it
> need another round of review?

Thanks for the confusion I guess, but I'm no committer here ;-)

This patch has received extensive review in July and I think it now
properly implements the design proposed by Tom and Heikki in 9.3/CF4.

As the path didn't make it already, yes it needs another (final) round
of review. The main difficulty in reviewing is understanding the design
and the relation in between our current model of extensions and what
this patch offers.

You might find the discussions we had with Markus Wanner quite useful in
this light. The current situation is that I believe the patch to
implement the same “template” model as the on-disk extensions, down to
dependency tracking.

IIRC I left only one differing behavior, which is that you're not
allowed to DROP an Extension Template when it's needed for a dump and
restore cycle, where you could be doing that at the file system level of
course (and pg_restore on a new system would depend on other files).

Regards,
--
Dimitri Fontaine
http://2ndQuadrant.fr PostgreSQL : Expertise, Formation et Support