On 11/20/23 12:24, Robert Haas wrote:
> On Mon, Nov 20, 2023 at 5:35 AM Laurenz Albe <laurenz.albe@cybertec.at> wrote:
>> I can accept that adding log messages to back branches is ok.
>> Perhaps I am too nervous about things like that, because as an extension
>> developer I have been bitten too often by ABI breaks in minor releases
>> in the past.
>
> I think that adding a log message to the back branches would probably
> make my life better not worse, because when people do strange things
> and then send me the log messages to figure out what the heck
> happened, it would be there, and I'd have a clue. However, the world
> doesn't revolve around me. I can imagine users getting spooked if a
> new message that they've never seen before, and I think that risk
> should be considered. There are good reasons for keeping the
> back-branches stable, and as you said before, this isn't a bug fix.
Personally I think that the value of the information outweighs the
weirdness of a new message appearing.
> I do also think it is worth considering how this proposal interacts
> with the proposal to remove backup_label. If that proposal goes
> through, then this proposal is obsolete, I believe.
Not at all. I don't even think the messages will need to be reworded, or
not much since they don't mention backup_label.
> But if this is a
> good idea, does that mean that's not a good idea? Or would we try to
> make the pg_control which that patch would drop in place have some
> internal difference which we could use to drive a similar log message?
The recovery in pg_control patch has all the same recovery info stored,
so similar (or the same) log message would still be appropriate.
> Maybe we should, because knowing whether or not the user followed the
> backup procedure correctly would indeed be a big help and it would be
> regrettable to gain that capability only to lose it again...
The info is certainly valuable and we wouldn't lose it, unless there is
something I'm not getting.
Regards,
-David