Re: Proposal: TABLE functions

>
>"Pavel Stehule" <pavel.stehule@hotmail.com> writes:
> > it can by more simple than I though. I need only one flag, and if its 
>true
> > then I don't create language variables for OUT params. But I need one 
>next
> > column in pg_proc.
>
>I thought you said this was just syntactic sugar for capabilities we
>already had?
>

My mistake. I am sorry. I have to store somewhere flag. One bit, which 
signalise "don't use OUT arguments as function's parameters". Other is only 
game in parser.

> > Currently a lot of columns in pg_proc is bool. What about one binary 
>columns
> > for other options? I hope so next versions can support autonomous
> > transaction, which need flag too.
>
>I think stored procedures of that sort aren't functions at all, and
>probably don't belong in pg_proc.
>

Why not? Some people use "ugly" implementation of it in plperlu and DBI. 
pg_proc and related infrastructure works well. It miss only little bit 
bigger adaptability. I thing so can be interesting one general option byte, 
and one byte reservated for language handlers.

Regards
Pavel Stehule

_________________________________________________________________
Najdete si svou lasku a nove pratele na Match.com. http://www.msn.cz/