compiling cruft?

[Dr. Hawkins]

I'm starting to notice a pattern here.

It seems like the first time I merge and compile a version of my code, the
resultant stack compiles quickly, but on subsequent compiles, it slows down
and takes many minutes longer to compile.


To prepare a compile, I have to reload in version 7 (supposedly as of
8.1.4-rc1 I won't), and run a routine that saves the file with a new name
including "to_compile", and loads all of the other stacks as substacks, and
then exits.  I then compile with a current version.

I just don't see where the cruft could accumulate, though--there is a "save
as" with the new name before the others are loaded, and thus well before
the compile.


-- 
Dr. Richard E. Hawkins, Esq.
(702) 508-8462