Re: superuser() shortcuts - Mailing list pgsql-hackers

From Stephen Frost
Subject Re: superuser() shortcuts
Date
Msg-id 20141204224548.GI25679@tamriel.snowman.net
Whole thread Raw
In response to Re: superuser() shortcuts  (Alvaro Herrera <alvherre@2ndquadrant.com>)
List pgsql-hackers
Alvaro,

* Alvaro Herrera (alvherre@2ndquadrant.com) wrote:
> Several dozens messages ago in this thread I would have dropped this
> item, TBH.

I'm getting to that point, though it's quite frustrating.  I didn't much
care initially but it's gotten to the point where the current situation
just strikes me as quite 'wrong'.

> *Maybe* I would consider changing the specific messages
> around the specific code that's being tinkered with, but probably not
> other ones.

If I thought that was a way to move forward, then I'd be happy with it.
I have no problem with this being a policy of "make it match the policy
when you change it, but don't just go changing everything right away as
it'll cause too much disruption."  I'm mostly argueing about what the
policy should be here and less about the specific minor changes in this
patch.

I'm pretty sure that's not acceptable to Robert or Andres though, as I
think they're also argueing policy.  I can certainly understand a
position which is "it's not wrong and we shouldn't change it for the
sake of changing it."

> I haven't looked at this patch recently, but if it's changing 1% of the
> error messages in the backend, I doubt it's a good idea.

This just confuses me- you said above that you'd consider changing just
the specific messages around the specific code being tinkered with
(which would be less than 1%..) and then say it's not a good idea to
change just 1%?  The actual patch is a mostly unrelated (and,
seemingly, not terribly controversial) change, as compared to this
discussion about error messages.

> More in general, if I see strong, well-rationalized opposition to some
> non-essential idea of mine, I tend to drop it because I don't see the
> value in arguing endlessly.  Life's already way too short.

I'm not sure that I see it as well-rationalized, but it certainly seems
to be consistent and the error message changes are certainly
non-essential.  Perhaps it'd be better discussed in Ottawa.
Thanks,
    Stephen

pgsql-hackers by date:

Previous
From: Peter Geoghegan
Date:
Subject: Minor documentation tweak to pg_stat_all_tables view description
Next
From: Andres Freund
Date:
Subject: Re: Minor documentation tweak to pg_stat_all_tables view description