On Tue, Nov 3, 2015 at 2:19 PM, Peter Geoghegan <pg@heroku.com> wrote:
> On Tue, Nov 3, 2015 at 11:15 AM, Robert Haas <robertmhaas@gmail.com> wrote:
>> OK, I see. Fixing comments in the back-branches is not always a
>> productive use of time, and in general I might like it if you pushed
>> for such things less frequently. But I've done it anyway in this
>> instance.
>
> I guess I favor doing it where the comment is actually wrong, which
> does apply here -- we don't *rely* on anything. We could very well
> abort when state is TSS_BUILDRUNS, but we elect not too, since
> TSS_BUILDRUNS is taken to mean that it's too late for aborting to be
> worth it.
Fair point.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company