Scott,
I understand and conceptually agree with all of the reasons for splitting a database; however, when done practically it can create issues when it's time to update the FE. To respond to your other points:
1. I obviously have to take into account data synchronization issues, but that's also a function of which data sources you use, e. g. internal tables or linked tables via ODBC. The point, though, is that it all involves no work or effort by the user and
doesn't involve additional components.
2. Obviously any changes require testing and can involve intensive work. The point again, though, is that the change is quick and immediate without the need to run additional software to implement the change on a user's computer.
3. There are plenty of things that can go wrong in a client-server environment involving PCs and laptops/tablets that get used wirelessly to access corporate networks. These are just a few of them:
a. There could be a bug in whatever software is used to download the FE changes.
b. The internal network may encounter a problem or go down.
c. The wireless network may encounter a problem.
d. If a VPN is used, there may be a problem.
e. Being the unpredictable gremlins that they are, users may do something unexpected.
Granted, all of these things can happen in a non-split database, but at least it's contained, consistent, and only involves access to one file. The complexity of a split database on multiple computing platforms creates a greater number of possible problems.
Also, corruption can be a problem even if the database is split.
The point of my question wasn't to get into a debate about the benefits of splitting a database. The benefits are well known, documented, and understood. I just wanted to find out if anyone had creative ways to distribute FE changes as easily as it is
to implement changes to a non-split database.
Eric