Re: PROPOSAL FE/BE extension to handle IN/OUT parameters

From: Dave Cramer <pg(at)fastcrypt(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: PROPOSAL FE/BE extension to handle IN/OUT parameters
Date: 2005-06-21 21:24:30
Message-ID: 153F2ED3-C5B4-46F9-B74D-C381E09A3373@fastcrypt.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Yeah, I think that might work if I understand it correctly.

Assuming I would be able to prepare, and bind all the parameters, and
the OUT parameters
would be ignored.

FWIW, I proposed adding to the protocol, not modifying the existing
messages, so it would be backward compatible and not break existing
clients.

Dave
On 21-Jun-05, at 5:14 PM, Tom Lane wrote:

> Dave Cramer <pg(at)fastcrypt(dot)com> writes:
>
>> The current situation with IN/OUT parameters requires that
>> considerable juggling is required on the client end to not pass the
>> OUT parameters in the query. This could be alleviated by adding two
>> messages for stored procedure calls
>> 1) PrepareCall which sent the types, and direction of the parameters
>> 2) BindCall which sends the binds the parameters to the above
>>
>
>
>> Is it possible to get this into 8.1, or is this a total non-starter
>>
>
> Changing the protocol is a nonstarter at this late date in the release
> cycle. I previously offered you a hack that would accomplish the same
> thing (or at least it looks like the same thing to me): ignore
> parameters of type VOID when looking up a function. Is that unusable
> from your end?
>
> regards, tom lane
>
>

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2005-06-21 21:26:40 Re: PROPOSAL FE/BE extension to handle IN/OUT parameters
Previous Message Magnus Hagander 2005-06-21 21:22:54 pg_terminate_backend idea