Steve McCrystal wrote in a message to David Bowerman:

 DB> The code to unload the functions after use is not included since I
 DB> never bother to unload them. Too many pieces of my REXX code make
 DB> use of them to bother with continual loading/unloading --
 DB> especially when you unload the functions from one thread while
 DB> another thread is making use of them.

 SM> I think I'll just load everything I might need and let it alone. 
 SM> Will it work to just call a routine from STARTUP that does so?

I used to do that but now just load from the first REXX code that gets run.  I 
can't remember why I stopped though it may have been tied to some problems
that I was having with the peer networking which had me dropping bits and
pieces until I found the problem.  More precisely, I never did find the
problem but after it disappeared, I was leary of changing the STARTUP.CMD file 
again.

My REXX coding starts from a boilerplate file that defines common constants,
loads the various helper files and so forth.  Of course, it can look a bit
silly when "Hello World!" takes 100 lines of boilerplate but easier in the
long run and keeps me from committing the easier dumb errors.

 SM> If the folks at SciTech could get *their* REXX code written
 SM> properly, I could just leave OREXX active and forget about it all!
 SM> :^)

Always the problem with other's code. ;-)

 SM> Thanks again for the help.

You're welcome.

Regards,
       David

--- timEd/2 1.10.y2k+