>>Ned,
>>
>>If that last system was in FP/VFP, just how does one go about writing "reentrant code" in VFP/FP???
>>
>>Regards,
>>
>>Jim N
>
>The basic idea is simple. It is very similar to the way that you prevent two instances of the same application from running on the same machine. To prevent your application from running twice you provide some way for the application to know that another instance is already running. To create reentrant code in VFP you must provide some way for your code to know that it is being called recursively and act accordingly. The simplest approach is to block action if the routine is being called recursively.

I have understood the idea of reentrancy in VFP in a different manner, somewhere along the lines where MS says that "VFP wizards are not reentrant, but builders are". Don't beat me to confess where I read it; after all these megabytes of text, I know I read it but don't remember where.

As for myself, I know I'm always trying to make any data transfer or conversion repeatable (does this mean it's "reentrant" in some way?). Our local definition of an automatic data transfer says it's "a procedure which copies errors from one app's data into another's, but if we know our job, it also copies the corrections". In general, any procedure which does something on a set of data or on an object and its properties (which also fall into data category) is reentrant in this aspect if it is able to read its data and allow for additions or corrections from where it last left off, instead of doing things from scratch every time.

I feel that the word "reentrancy" is not proper for this kind of behavior - I wrote this in search for a better word.