Re: clearing opfuncid vs. parallel query

On Thu, Sep 24, 2015 at 11:54 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Andres Freund <andres(at)anarazel(dot)de> writes:
>> On 2015-09-23 17:29:50 -0400, Robert Haas wrote:
>>> Well, I can't vouch for what any human being on earth has thought
>>> about over a twenty-year period. It's not intrinsically unreasonable
>>> in my mind to want to alter an operator to point at a different
>>> procedure.
>
>> Wouldn't we use plan invalidation to deal with that anyway?
>
> Plan invalidation wouldn't help, because the obsolete data exists
> on-disk in stored rules. You'd have to run through the pg_rewrite
> entries and update them.
>
> To my mind though, the lack of an ALTER OPERATOR SET FUNCTION command
> is on par with our very limited ability to alter the contents of
> an operator class. In principle it would be nice, but the practical
> value is so small that it's not surprising it hasn't been done ---
> and we shouldn't continue to hold the door open for a simple way of
> implementing it when there are significant costs to doing so.

Also, it's not like this change couldn't be UN-done at a future point.
I mean, Tom didn't like the flag I added aesthetically, but if we
needed it, we could have it. Or we could engineer something else.

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