From: | Andrew Dunstan <andrew(at)dunslane(dot)net> |
---|---|
To: | "Jim C(dot) Nasby" <jnasby(at)pervasive(dot)com> |
Cc: | Pavel Stehule <pavel(dot)stehule(at)hotmail(dot)com>, a_dursun(at)hotmail(dot)com, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: proposal for PL packages for 8.3. |
Date: | 2006-08-09 17:15:13 |
Message-ID: | 44DA1821.9050606@dunslane.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Jim C. Nasby wrote:
>>> 4. Syntax must be as closer as plpgsql (declaration, assingment etc)
>>> rather than any syntax that we have to learn :-)
>>>
>> PostgreSQL support other languages than PL/pgSQL. We need universal syntax
>> for plperl and others too
>>
>
> Why? Don't those other languages have support of their own for this?
>
> If we try and make this completely cross-language I fear we'll end up
> with something so watered down and obtuse that it'll be useless. I think
> it makes much more sense to design something for plpgsql and only
> commonize whatever it makes sense to.
>
plperl and pltcl at least have support for now for shared non-table
session data. The trouble is that it is shared ONLY inside the
interpreter. That means there is no sharing between, say, a plperl func
and a pltcl func. Now it would make far more sense if session objects
could be shared between interpreters, especially if they are namespace
scoped. So I think you need to give a good reason for NOT sharing.
cheers
andrew
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Treat | 2006-08-09 17:20:41 | Re: how to determine which types take a length argument |
Previous Message | Tom Lane | 2006-08-09 17:07:30 | Re: [HACKERS] [PATCHES] Values list-of-targetlists patch for comments (was Re: |