Re: pg_stat_statements in core

ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp> writes:
>> Now I'm working on storing statistics into disks on server
>> shutdown. If it is impossible unless the module is in core,
>> I would change my policy...

I'm really not happy with a proposal to put such a feature in core.
Once it's in core we'll have pretty strong backwards-compatibility
constraints to meet, and I don't think you are anywhere near being
able to demonstrate that you have a solid API that won't require
changes. It needs to be a contrib or pgfoundry package for awhile,
to shake out feature issues in a context where users will understand
the API is subject to change. (As an example of why I'm skittish
about this: just a few days ago someone was complaining about the
plans to get rid of pg_autovacuum, despite the fact that it's been
clearly documented as subject to change or removal since day one.
People expect stability in core features.)

It seems to me that all you're really missing is a shutdown hook
someplace, which would be a reasonable core addition.

regards, tom lane