I'm looking for a quick response as to whether coda might be suitable
for my needs.

I have a windows database system, a foxpro application running on
Win95/Win98 clients.  Currently the database is on a samba fileserver.

At the moment, the database is mostly used by client machines on the
lan.  However, there are a number of potential remote users, using
laptops.

On the plus side, the remote users are not entering data into the
database, only generating reports from it.  Also, the data is "rolled
over" once a week, and the remote users don't need access to the most
recently entered data, only the data entered before the last rollover.

On the minus side, I imagine the database program needs read/write
access to the files, but I don't really know what files it would
modify.

How I intend to do this: I will keep the lan users on the samba
system.  Once per week, after the database rollover, I will copy the
data into coda.  Then I will sync the laptops with the coda server.
When the laptops are in use in the field, they will be operating
either disconnected, or over a dial-up.

What I need to know is:
How does coda operate over a slow (dialup) link?  (In my situation,
I imagine very little data will have to be written back to the server,
and basically all the data will be cached on the client)

Would it be better to operate disconnected, or over the dialup?

When I do the weekly resync, can I tell coda to throw away any changes
made to files at the client end, and simply copy the server files to
the clients?  (Of course, it it was linux, I'd simply use rsync, but
it's windows)

Has anyone used coda for this type of application?

Is there a better way to do this?

Basically, I'd like to test coda for this, but I need to know if there
is something fundamental that would make coda unsuitable in this
scenario.

TIA

Alex