Re: pgsql: Avoid archiving XLOG_RUNNING_XACTS on idle server

On Wed, Apr 6, 2016 at 4:18 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>> FWIW, I vote also for reverting this patch. This has been committed
>> without any real discussions..
>
> Michael, its a shame to hear you say that, so let me give full context.
>
> The patches under review in the CF are too invasive and not worth the
> trouble for such a minor problem. After full review, I would simply reject
> those patches (already discussed on list).
>
> Rather than take that option, I went to the trouble of writing a patch that
> does the same thing but simpler, less invasive and more maintainable.
> Primarily, I did that for you, to avoid you having wasted your time and to
> allow you to backpatch a solution.
>
> We can, if you wish, revert this patch. If we do, we will have nothing,
> since I object to the other patch(es).

I don't think you have an absolute veto over other patches, though you
certainly have the right to object, and you certainly don't have to
commit them yourself. But even more than that, the fact that you
don't like those other patches does not mean that you can commit
something without discussion. Even if every argument you are making
here is correct, which I'm not sure about, other people obviously
don't think so. That stuff should be worked out, as far as possible,
in pre-commit review, which is only possible when you post the patch
before committing it. I think it is fine to commit things
occasionally without posting them ahead of time if they are obviously
uncontroversial, but that isn't the case here.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company