I maintain and expand an application which is based (loosely now) on the CodeBook and Visual FoxExpress 5 frameworks.

We have shifted our development directory structure around considerably in order to work better with SVN version management. As a result I need to understand how the framework sets up values for PATH, CLASSLIB, and PROCEDURE.

This is what is confusing me. Even when the application is run in a network directory which bears no resemblence to the development tree structure, these settings still append the (nonexistent) development directories to the references to PRG and VCX files.

This seems crazy to me.

1) Why doesn't the BUILD process by itself insure that all components can be found at run-time?

2) If one must have voluminous lists of components in CLASSLIB and PROCEDURE, how does it help to refer to nonexistent directories?

Scratching my head. :^}

Peter