Re: SetQuerySnapshot, once again

Tom Lane wrote:
>
> Hiroshi Inoue <Inoue(at)tpf(dot)co(dot)jp> writes:
> >> I do not like the idea of treating the first select in a function
> >> differently from the rest. And such a rule wouldn't let you build
> >> guaranteed-stable functions anyway;
>
> > AFAIK there has been no analysis where we can get *stable*
> > functions. As far as I see, we can expect SELECT-only functions
> > to be *stable* if and only if they are surrounded by SELECT-only
> > *stable* functions.

Oops I was wrong. The last *stable* isn't needed.

> This idea might be a bit off-the-wall,

Probably I mentioned once long before.
We can't expect reasonable result for
select fn1(..), fn2(..), ... from ... ;
if there are some fnx()-s with strong side effect.

> but how about:
>
> 1. If a plpgsql function is declared immutable or stable, then all its
> queries run with the same snapshot *and* CommandCounterId as prevail
> in the calling query.

IMHO it's impossible to handle anything with one concept.
Functions could be *immutable*(? deterministic in SQL99)
or *stable* even though they have strong side effect.

regards,
Hiroshi Inoue
http://w2422.nsk.ne.jp/~inoue/