I would think one way you might could go about this would be to use autocommit of the transactions (inserts/updates/deletes/replacements), and an interface that either was user-driven for pulling updates (hit a button to check for updates) or timed to look for updates (such as a meta-refresh if the interface is via a webpage to pull the updates).
Hope that helps....
I will be using Access for the database.Ouch. One word of caution: "simultaneous multi-user access" and MS-Access don't go well together. MS-Access if fine as long as you stick to single user. Try to get a more suitable database, for example PostGres, MS-SQL Server, or even MySQL. MS-Access is known to regularily cause data corruption, with simultaneous multi-user edits.
As for your approach... IMO you can have live updates on your data, as long as the user wasn't actually in the middle of editing it. At the least, you do need a central locking mechanism, making sure only one user can edit a particular field, or a particular record, at a time.
Less intrusive would indeed be a simple indicator, that might look a bit like a colored LED, indicating whether the data shown is up to date (green) or stale (red).
perlmonks.org content © perlmonks.org and Anonymous Monk, atcroft, bart, beable
prlmnks.org © 2006 edmund von der burg (eccles & toad)
v 0.03