Re: Unable to identify an ordering operator '<' for type 'smallint[]' - Mailing list pgsql-sql

From Mauricio Fernandez A.
Subject Re: Unable to identify an ordering operator '<' for type 'smallint[]'
Date
Msg-id KPEAKEDEHEPGJLOLEDMBOEPCCFAA.mfacontacto@ono.com
Whole thread Raw
In response to Re: Unable to identify an ordering operator '<' for type 'smallint[]'  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-sql
Thanks Jaime, I know, you are right : it´s a very bad idea to develop in
version that is superior to the
one in production, but, mmmm....  I don`t know, you win ;).

And thanks to Tom too, because he told the key phrase "GROUP BY
ac.potencialpacientes", as I can skip that clause, now I get the result
expected

Thanks for your help

Mauricio Fernández A.
Ingeniero de Sistemas
U. Autónoma de Manizales

-----Mensaje original-----
De: Tom Lane [mailto:tgl@sss.pgh.pa.us]
Enviado el: miércoles, 11 enero, 2006 19:25
Para: Mauricio Fernandez A.
CC: pgsql-sql@postgresql.org
Asunto: Re: [SQL] Unable to identify an ordering operator '<' for type
'smallint[]'


"Mauricio Fernandez A." <mfacontacto@ono.com> writes:
> The only strange thing in the query is that ac.potencialpacientes is
> smallint[] (in fact is the only thing with sense I can see in the error
> message) but I am not using the '<' operator as it suggest.

No, but you're trying to GROUP BY ac.potencialpacientes, and in 7.3 the
only way to do grouping is sort/unique.  So you have to be able to sort
the datatype, and 7.3 doesn't have code to do comparison of arrays.

> Can somebody help me?.

Upgrade.  7.3 is pretty ancient anyway.  Quite aside from lack of features,
its performance is poor compared to 8.1.
        regards, tom lane



pgsql-sql by date:

Previous
From: Jeff Boes
Date:
Subject: psql client: technique for applying default values to :variables?
Next
From: Dirk Jagdmann
Date:
Subject: foreign keys with on delete cascade and triggers