Re: Bytea network traffic: binary vs text result format

On 6/4/07, Markus Schiltknecht <markus(at)bluegap(dot)ch> wrote:
>
>
> There already are. Not in libpq, though. Most (if not all) internal
> types have those functions.

Well, I don't think it would be puzzling that someone will ask for them even
if those functions already exist in PostgreSQL because, as you stated, it's
not in libpq. People will actually have to dig in the source to pinpoint
where in the source tree those functions exist and how to include them in
the program, which, I assume, is less ideal than when it's already included
in libpq in the first place. There was a post before on a user who got
disappointed because of the "crappy" support of libpq (not PostgreSQL) for
binary formats.

In line with this, I'm aware of the issues that this may pose:
http://archives.postgresql.org/pgsql-hackers/1999-08/msg00374.php this is
already very old (8 years
ago!). And they discuss about the issue that the representation might
change from version to version so it's not done. There have been plans
to incorporate CORBA,
which IMHO is an overkill,
to solve this problem but I don't think it's done yet because it's
probably too complex (?). I'd rather recommend them to use
ASN.1 (if that's feasible..).

--
Life is too short for dial-up.