Re: [PATCH] EnableDisableTrigger Cleanup & Questions

On Thu, Nov 6, 2008 at 10:08 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> I have no objection to cleaning up the backend internals, but system
> catalog definitions are client-visible.  I don't think we should thrash
> the catalog definitions for minor aesthetic improvements.  Since 8.3 is
> already out, that means client-side code (like pg_dump and psql, and
> probably other programs we don't control) is going to have to deal with
> the existing definition for the foreseeable future.  Dealing with this
> definition *and* a slightly cleaner one isn't a net improvement from the
> client standpoint.

Well, it didn't seem like anyone had an issue changing the definition
at 8.3 time.  As for pg_dump/psql, those changes are fairly simple.
And, there aren't that many PG utilities out there.  PGAdmin looks
like it would require a 1-3 line change (depending on coding
preferences) and I don't see anything that checks it in Slony.

I'm fine with cleaning up the internal-side, I just don't think
there's that much relying on tgenabled.  In fact, Google code search
seems to show more things relying on a boolean tgenabled rather than
the current implementation.

Oh well, it was just a thought.

-- 
Jonah H. Harris, Senior DBA
myYearbook.com