Re: Extension Templates S03E11

* Dimitri Fontaine (dimitri(at)2ndQuadrant(dot)fr) wrote:
> Stephen Frost <sfrost(at)snowman(dot)net> writes:
> > Clearly we need the information from the extension package (the scripts
> > which are on the PG server's filesystem today, but need not be in the
> > future) but that doesn't mean we need to keep those text blobs in the
> > catalog.
>
> So, I guess it would have been good to hear about that about a year ago:
>
> http://www.postgresql.org/message-id/13481.1354743758@sss.pgh.pa.us
> http://www.postgresql.org/message-id/6466.1354817682@sss.pgh.pa.us
>
> We could have CREATE TEMPLATE FOR EXTENSION store the scripts into some
> files in PGDATA instead of the catalogs, but really I don't see the
> point.

Yeah, I don't particularly like that idea either, but especially if it's
going to be per-database again. I can kinda, sorta see the point if
this was done cluster-wide but you don't like that idea and I'm not a
big fan of pushing these files out onto the filesystem anyway.

What I don't entirely follow is the argument against having
non-file-backed extensions be dump'd through pg_dump/restore. Even in
that thread, Tom appears to agree that they'd have to be dumped out in
some fashion, even if they're stored as files under PGDATA, because
otherwise you're not going to be able to restore the DB..

On the other hand, I can appreciate the concern that we don't really
want a dump/restore to include the extension definition when it's
already on the filesystem. That said, it amazes me that we don't
include the version # of the extension in pg_dump's 'CREATE EXTENSION'
command.. How is that not a problem?

Thanks,

Stephen