Re: [INTERFACES] ODBC - MSysConf - Mailing list pgsql-interfaces

From Karsten Kaus
Subject Re: [INTERFACES] ODBC - MSysConf
Date
Msg-id 36C611BD.2A69@kdschmid.de
Whole thread Raw
In response to ODBC - MSysConf  (Karsten Kaus <kk@kdschmid.de>)
List pgsql-interfaces
David Hartwig wrote:

I still do not know, whats going on, but I get it running somehow:
I must not need the controlpanel to set up the DSN. If I stay within
Access when
linking the tables and configure the DSN from the panel Access opens up,
then
everything works (as far as I checked). I even get write-access - wow.
Access seems to make a copy of the driver-settings. I set up a DSN with
read-only access (wich shows up in the contolpanel after linking the
tables) and
try to change to read/write in the controlpanel - it does not work.
I have to link all the 70 tables new with a freshly created DSN, within
Access.
Hope I get this distributed to all the stations without too much
handwork.
For this first try, I'm only creating a faxlog for 6 of our users. But
in 2 weeks it's
going to be harder.

( By the way: Most of the records I will use are bigger than 2k. Can
someone give me a
keyword how to handle that?)

>
> What version of the PostgreSQL driver are you running?
I thought it was the latest: PostgreSQL 6.40.0004
>
> What version of Access?
Access 97 - Access 2.0 does not recognize the PostgreSQL Driver.
>
> What version of the Driver Manager?   The driver install will upgrade to the DM to 3.0 but it is optional.
I should be already above that.
>
> I probably should have asked this before we got this far.
>
> Karsten Kaus wrote:
>
...
>
> >
> > MSACCESS        fff27265:fff26281       EXIT  SQLExecDirect  with return code
> > 0 (SQL_SUCCESS)
> >                 HSTMT               0x00b31870
> >                 UCHAR *             0x040ce9a0 [      -3] "SELECT Config, nValue FROM
> > MSysConf"
> >                 SDWORD                    -3
> >
>
> Here is where it gets strange.   The above statement should have a return code of -1 if the table does not exist.
Andthe error message should say so.   (I just this verified this on my 
> system.)   The next few statements tries to cope with the result set and it is not happy with what it is getting.

Thanks a lot so far, but still I'm not too happy with that.

KK

pgsql-interfaces by date:

Previous
From: Peter T Mount
Date:
Subject: Re: [INTERFACES] libpq and multi-threading on Linux?
Next
From: Aaron Dunlop
Date:
Subject: Perl5 interface under win32