Re: Collations and Replication; Next Steps

On Wed, Sep 17, 2014 at 7:46 AM, Greg Stark <stark(at)mit(dot)edu> wrote:
> You could have a problem if you have an expression index on (timestamp
> AT TIME ZONE '...'). I may have the expression slightly wrong but I
> believe it is posisble to write an immutable expression that depends
> on the tzdata data as long as it doesn't depend on not the user's
> current time zone (which would be stable but not immutable). The
> actual likelihood of that situation might be much lower and the
> ability to avoid it higher but in theory I think Peter's right that
> it's the same class of problem.

Really? If you have an "IMMUTABLE" expression predicated on something
that isn't immutable, then it isn't actually immutable, regardless of
the ability of Postgres to determine that. It's that simple. So, if
that's true, then it sounds like a bug to me. I didn't check, but I
think that AT TIME ZONE just changes the display format, and the
relevant output function is only stable, and so this shouldn't be a
problem.

> Peter G, could go into more detail about collation versioning? What
> would the implications be for Postgres?

Well, I think one implication might be that Postgres won't start, just
because you updated the OS (we have no way to get back results
consistent with the old collation file, I would think). At least your
database isn't corrupt, but that's still a really bad loss of
availability. This makes me lean towards pursuing ICU support as part
of any versioning scheme. There is a reason why everyone else does
something similar. Apparently DB2 uses ICU.

--
Peter Geoghegan