suggest you start by splitting your database. tables go into a backend file
that sits on your server. all other objects go into a frontend file, and the
backend tables are linked to the frontend. a copy of the frontend is placed
on the PC of each person who uses the database. see
http://www.granite.ab.ca/access/performancefaq.htm for more tips on
optimizing performance in a shared database.
hth
Thanks for that suggestion. Actually, I tried that first and the problem is
that I have lots of code that opens tables, sets indexes on them, and makes
changes to them. All this code (I discovered) would have to be changed
considerably to work in a linked-table environment, which doesn't allow
references like
set weborders=psbcust.openrecordset("weborders",DB_OPEN_TABLE)
nor
weborders.index="orderid"
SO I was trying to do this the less-efficient way.
Is it possible?
thanks,
dave

Signature
Dave
PHONE SOURCE
www.thephonesource.com
> suggest you start by splitting your database. tables go into a backend file
> that sits on your server. all other objects go into a frontend file, and the
[quoted text clipped - 38 lines]
> > dave cook
> > phone source
tina - 17 Jan 2006 16:48 GMT
sorry, dave, i have no other suggestions for you.
> Thanks for that suggestion. Actually, I tried that first and the problem is
> that I have lots of code that opens tables, sets indexes on them, and makes
[quoted text clipped - 57 lines]
> > > dave cook
> > > phone source
Tim Ferguson - 17 Jan 2006 17:36 GMT
> the problem is
> that I have lots of code that opens tables, sets indexes on them, and
> makes changes to them.
This is a very old-fashioned way of working with databases, and generally
comes from not making the break from file managers like dBase.
You need now to move up into the late sixties, and embrace the world of
RDBMS -- although Access is not strictly a RDBMS, it's close enough to
warrant a similar approach. Indexing is a once-only design time process;
making changes to data should all be done via queries. As long as you
have made a good job of the schema design (that's the "R" part), the db
engine itself will use whatever indexes are appropriate (the "MS" part).
Chuck out all those Open Table methods!
Hope that helps
Tim F
David - 18 Jan 2006 00:11 GMT
I guess this newsgroup is only for people who do things the ideal right way
and not for those who are willing to accept a good-enough real-world solution.
Maybe there isn't a short-term (possibly kludgy) way to make it run
multi-user without rethinking all my ways of doing things. But neither
answer I received on this forum actually said that. I felt lectured rather
than helped- the message I received was "you're wasting your time - do it
the right way or not at all"
If one of your answers had explicitly said "Sorry, it's not possible unless
you do the re-design necessary to conform to a more modern approach," then I
would accept that advice. But my sense is that I'm not being told what is
necessary, but rather what is fashionable.
- dave
But I won't bother this newsgroup with my amateeurish questions anymore.

Signature
Dave
PHONE SOURCE
www.thephonesource.com
> > the problem is
> > that I have lots of code that opens tables, sets indexes on them, and
[quoted text clipped - 15 lines]
>
> Tim F