Re: Infinite Interval

On Mon, 30 Oct 2023 at 10:01, Ashutosh Bapat
<ashutosh(dot)bapat(dot)oss(at)gmail(dot)com> wrote:
>
> Thanks. I went through the whole patch again and am quite fine with it.
>
> Here's my version of commit message
>

Going over this again, I noticed a pre-existing integer overflow
problem in interval_time(), which patch 0001 attached fixes.

I squashed the other 2 patches (main patch + new BRIN tests) together
into 0002, and did another copy-editing pass over it. I found one
other issue, which is that overflow checking in interval_um() had gone
missing, and there didn't seem to be any regression test coverage for
that, so I added some.

I also changed the error message in interval_time to "cannot convert
infinite interval to time", which is slightly more informative, and
more consistent with the nearby error messages in time_pl_interval()
and time_mi_interval().

Finally, I rewrote the commit message in slightly higher-level terms,
but that's really up to the committer to decide on.

I'm marking this as ready-for-committer. I'll probably pick it up
myself in a few days, unless another committer claims it first.

Regards,
Dean