Nulls MSAccess and SQL Server

I've recently upgraded a SQL Server 6.5 database to SQL Server 2000 together
with it's front end application MSAccess97 to MSAccess2000.
The MSAccess app, run on the client pc uses a Windows API to determine the
user
logon and machine name - this data is used for audit purposes and is append
ed
to all data records during inserts and updates on the SQL Server database.
However, since upgrade I'm finding that some of the audit fields in the SQL
Server
tables do not contain data although these fields are not nullable.
With the previous software if the user details were lost at the client data
changes
were blocked at the SQL Server end. It now seems the supposed nulls are
going across as empty strings therefore accepted by SQL Server.
I haven't changed anything, why would the handling of nulls have changed
either in
MSAccess2000, SQLServer2000 or in the interface between the two?Without seeing your code, it's hard to say what's going on. You can
put a Profiler trace on the app to debug and see exactly what Access
is sending over the wire. You can also create a constraint on the
table in SQLS to disallow the empty strings. You might want to
consider re-creating this functionality in a trigger instead of Access
VBA code.
--Mary
On Fri, 15 Oct 2004 07:55:13 -0700, "Peter"
<Peter@.discussions.microsoft.com> wrote:

>I've recently upgraded a SQL Server 6.5 database to SQL Server 2000 togethe
r
>with it's front end application MSAccess97 to MSAccess2000.
>The MSAccess app, run on the client pc uses a Windows API to determine the
>user
>logon and machine name - this data is used for audit purposes and is appen
ded
>to all data records during inserts and updates on the SQL Server database.
>However, since upgrade I'm finding that some of the audit fields in the SQL
>Server
>tables do not contain data although these fields are not nullable.
>With the previous software if the user details were lost at the client data
>changes
>were blocked at the SQL Server end. It now seems the supposed nulls are
>going across as empty strings therefore accepted by SQL Server.
>I haven't changed anything, why would the handling of nulls have changed
>either in
>MSAccess2000, SQLServer2000 or in the interface between the two?