Re: updated hstore patch

>>>>> "David" == David E Wheeler <david(at)kineticode(dot)com> writes:

>> The only open question I can see is what delete(hs,$1) resolves to
>> when $1 is an unknown-type placeholder; this is probably an
>> incompatibility with the old version if anyone is relying on that
>> (but I don't see why they would be).

David> Given your examples, I think it probably should resolve to
David> text as it does, as deleting a single key is likely to be a
David> common case. It should otherwise be cast.

I think you're missing the point here; I can't control what it resolves
to, since that's the job of the function overload resolution code.

But I checked, and delete(hstore,$1) still resolves to
delete(hstore,text) when the type of $1 is not specified, so there's
no compatibility issue there that I can see. (I'm not sure I
understand _why_ it resolves to that rather than being ambiguous...)

>> The overhead is possibly non-negligible for reading old values,
>> but old values can be promoted to new ones fairly simply
>> (e.g. using ALTER TABLE).

David> So then it's negligible for new values?

Yes. (One bit test, done inline)

--
Andrew.