2010/9/9 Tom Lane <tgl@sss.pgh.pa.us>:
> Pavel Stehule <pavel.stehule@gmail.com> writes:
>> 2010/9/9 Robert Haas <robertmhaas@gmail.com>:
>>> That seems like a lot of complexity for no real benefit, to me.
>
>> no, you can to specify a expected result type - it's very for some
>> convert or import functions. So we expect so out procedures will
>> supports to OUT parameters, then implementation of this mechanism has
>> minimal overhead to current implementation. Just to add types of OUT
>> parameters to searching algorithm.
>
> It's *not* trivial, not at all. You are ignoring all of the semantic
> implications. Should foo(IN x int, OUT y int) be considered different
> from, and thus allowed to exist at the same time as, foo(IN x int,
> OUT y float)? If so, how do you represent that in the catalogs?
> Possibly more to the point, any such decision means that it'll be
> impossible to call any stored procedure without fully specifying the
> types of output arguments as well as input arguments, else the system
> can't tell which procedure you meant to call. That doesn't sound like
> a notational improvement to me.
it can be a foo(int, int) and foo(int, float) in catalog. These lists
are distinct so there are not a problem.
>
> I'm with Robert: this would be a huge extra complication for a
> remarkably small amount of benefit.
>
I don't agree - you see a procedure like some void function, and it's not true
Regards
Pavel
> regards, tom lane
>