Case in point:
MS Access was designed to have multiple database managers manipulating
the files themselves directly and uses another file for locking
information. However-- as anyone who has ever worked with the process
will tell you: Don't do it. Data corruption (often unrecoverable) will
result.
The lessons we have learned from MS Access are:
1) Don't have 2 unrelated backends trying to access the same data and
2) Don't do it across a network.
IMO, this shows a fundamental design flaw in MS Access at least given
how it is marketed.
Don't try to do the same with PostgreSQL.
Best Wishes,
Chris Travers