Обсуждение: Re: [ADMIN] PG 7.3.4 VS PG 8.0.3 Problem

Поиск
Список
Период
Сортировка

Re: [ADMIN] PG 7.3.4 VS PG 8.0.3 Problem

От
Chris Hoover
Дата:
Tom,

I'll send this over to the odbc list.  However, the odbc driver is
liked into our cobol environment.  I did not change the driver.  We
are running the 7.2.5 odbc driver.  This is why I am pointing to a
difference in the db responses.

The entire environment is the exactly same except for the database
versions (I am actually running the exact same code and odbc driver
against the 2 db's).  I can have the application connect via odbc to
my 7.3.4 instance and get the sqlcode of 0.  I can have the same
application with the same odbc connect via odbc to my 8.0.3 instance
and I get the sqlcode of 100.

This is what is pointing me to a problem with one of the db's.  It
appears that the sqlca code has had to of changed between the 7.3.4
and 8 code and is now behaving in a different (and unexpected) manner.

BTW, which version is correct?  Should a no rows return query give a 0 or a 100?

Chris

On 8/9/05, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Chris Hoover <revoohc@gmail.com> writes:
> > Anyway, I'm not sure how to give you a simple example.  We are running
> > a cobol application that is connecting to PostgreSQL via odbc.  The
> > application runs on one linux box and the database on a seperate one.
>
> Hmm.  I seriously doubt that the server's response to the query has
> changed any.  What seems more likely is that the ODBC driver has changed
> its behavior.  Can you try linking your app with the older ODBC driver
> and running it against the newer server?
>
> Also, pgsql-odbc is probably a better place to ask about it.
>
>                         regards, tom lane
>

Re: [ADMIN] PG 7.3.4 VS PG 8.0.3 Problem

От
Tom Lane
Дата:
Chris Hoover <revoohc@gmail.com> writes:
> This is what is pointing me to a problem with one of the db's.  It
> appears that the sqlca code has had to of changed between the 7.3.4
> and 8 code and is now behaving in a different (and unexpected) manner.

There is no "sqlca" in the backend protocol at all, so I really dunno
what is causing the change in behavior.

> BTW, which version is correct?  Should a no rows return query give a 0
> or a 100?

I suspect 100 is correct, but you'd have to go read the ODBC spec to
be sure.

            regards, tom lane