Re: mark the timestamptz variant of date_bin() as stable

On Wed, Sep 1, 2021 at 3:25 PM Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>
> John Naylor <john(dot)naylor(at)enterprisedb(dot)com> writes:
> > On Wed, Sep 1, 2021 at 2:44 PM Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> >> I see that these two answers are both exactly multiples of 24 hours
away
> >> from the given origin. But if I'm binning on the basis of "days" or
> >> larger units, I would sort of expect to get local midnight, and I'm not
> >> getting that once I cross a DST boundary.
>
> > Hmm, that's seems like a reasonable expectation. I can get local
midnight
> > if I recast to timestamp:
>
> > # select date_bin('1 day', '2021-11-10 00:00
+00'::timestamptz::timestamp,
> > '2021-09-01 00:00 -04'::timestamptz::timestamp);
> > date_bin
> > ---------------------
> > 2021-11-09 00:00:00
> > (1 row)
>
> Yeah, and then back to timestamptz if that's what you really need :-(
>
> > It's a bit unintuitive, though.
>
> Agreed. If we keep it like this, adding some documentation around
> the point would be a good idea I think.

Having heard no votes on changing this behavior (and it would be a bit of
work), I'll start on a documentation patch. And I'll go ahead and re-mark
the function as immutable tomorrow barring objections.

--
John Naylor
EDB: http://www.enterprisedb.com