>David;
>
>One way to do it is to install the necessary files on one computer and map from the other computer. The runtime files are only required on the computer with the exe.
>


No, thje runtime files are needed by -each- machine that executes the .EXE; while you can try to run without properly installing and registering the runtime, the results are spotty at best, and anything that relies on the runtime's self-registration registry entries to locate common components will fail miserably.

>There is no question in my mind that it is best to have a second computer to test on. This is a clean machine (only has windows operation system on it) and you then map from the second machine. Testing will prevent problems at the user end due to programming anomalies. There is no cure for users that are not computer literate. Well, the desire to acquire knowledge will do it but that takes effort. To be honest I have four computers on my home network. I load on a clean machine, and map from a clean machine and then “attack” the program by working on the same record.
>
>One thing about config.fpw files I do not use drive designations. I might end up installing on a different drive. Buffer = 5 is almost always safe.
>
>Barbara is right that you can run more than one version of the program on one computer and test locking. When it comes to the installation though I prefer to use more than one (clean) machine. My friends and family ask why I have so many computers. I answer I want happy clients and have not had any calls from clients who were having problems with the application. I do get calls for additiona to programs I write. If you piss off the client you are going to be in trouble. It is impossible to do PR after an installation has problems.
>
>Tom