"Jeff Eckermann" <jeff_eckermann@yahoo.com> wrote in message
news:20040805184412.78858.qmail@web20826.mail.yahoo.com...
> How much testing have you done? I have never had a
> problem with this, but then, I don't have to deal with
> any huge datasets.
>
I don't have any large datasets; largest table at present only has about
10,000 records.
Had created 2 views against the largest recordset for testing, one with a
WHERE clause, and 2 bound forms, each using one of the views as a
recordsource. The form using the view without the WHERE clause had the
conditions in the WHERE clause used as a filter.
I didn't put in timing code, but the filtered form seemed a little slower.
Dropped and rebuilt the test views, forms, etc. and now don't see much
difference.
Phillipe had noticed difference between Access 2000 and Access 2002/2003
mdb, but I had already been using 2002/2003, as will be distributing runtime
executable to my employees.
Can't explain the difference in performance from first testing, but will try
using filters in upcoming production version unless real world performance
suffers.
> >
> > I change the SQL of pass-through queries dynamically
> > at runtime to use as
> > record sources for reports. That wouldn't work for
> > forms, as not updatable.
> >
> > Is the best approach to use an updatable view as the
> > record source, then
> > change the view definition at runtime as with a
> > passthrough query?
>
> Why would you need to do this? If you want to show
> different data to different users, then you reference
> the user name (e.g. current_user or session_user) in
> your view definition.
>
Was just creating unique name for the temporary view, for which string
containing current_user would certainly be easier. All of my employees will
have similar table permissions at the present time, as small office and
similar tasks.
David P. Lurie
