Re: Strange results with date/interval arithmetic

Tom Lane wrote:
> Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> > At least it is before final. Can we give people on hackers an SQL
> > script to run in every database to fix this? That is how we have
> > handled this in the past.
>
> We have? I don't really know how one would adjust catversion without
> an initdb. (Bear in mind it's inside a binary, CRC-protected control
> file; couldn't be done without a special-purpose C program AFAICS.)
>

Yes, that seems like a problem. Also, will this affect regression tests
for people who don't apply the patch? That makes it extra important we
make sure the patch is applied which gives more weight to the catversion
bump.

> If you wanted to *not* bump the catversion then we could let people run
> a script to fix the two pg_proc entries, but I think that way is likely
> to do more harm than good in the long run. Too much chance of someone
> carrying the wrong entries into production and not noticing their wrong
> answers for a long time.

Yes.

> The ground rules for beta testers have always been "you may have to
> initdb before final", and I think that's where we are now, annoying
> as it is.
>
> (Too bad we don't have a working pg_upgrade...)

I could probably get it working tomorrow if people want it.

--
Bruce Momjian | http://candle.pha.pa.us
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026