Re: Patches that don't apply or don't compile: 2017-09-12

On Fri, Sep 15, 2017 at 8:13 AM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Wed, Sep 13, 2017 at 5:18 AM, Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org> wrote:
>> Thinking further, I think changing patch status automatically may never
>> be a good idea; there's always going to be some amount of common sense
>> applied beforehand (such as if a conflict is just an Oid catalog
>> collision, or a typo fix in a recent commit, etc). Such conflicts will
>> always raise errors with a tool, but would never cause a (decent) human
>> being from changing the patch status to WoA.
>
> Well it would perhaps be fine if sending an updated patch bounced it
> right back to Needs Review. But if things are only being auto-flipped
> in one direction that's going to get tedious.
>
> Or one could imagine having the CF app show the CI status alongside
> the existing status column.

FWIW that last thing is the idea that I've been discussing with
Magnus. That web page I made wouldn't exist, and the "apply" and
"build" badges would appear directly on the CF app and you'd be able
to sort and filter by them etc. The main submission status would
still be managed by humans and the new apply and build statuses would
be managed by faithful robot servants.

How exactly that would work, I don't know. One idea is that a future
cfbot, rewritten in better Python and adopted into the pginfra
universe, pokes CF via an HTTPS endpoint so that the CF app finishes
up with the information in its database.

--
Thomas Munro
http://www.enterprisedb.com