Response to the proposed directory tree:
It's fine as you have proposed it. I have a slight
preference for "sim" over "mc". I'd like to see
a documentation directory at least for the simulation
directory, and I think it wouldn't hurt for the
data either. (E.g., high-gain run, low-gain run,
how many good events, what was trigger, etc)
The /calcom/data directory will presumably be
empty for a few months... I also had created
a directory called "resources" and put things
like the general cernlib documentation in them;
this could be the super-directory in which
the "paw" directory is contained. I do not have
strong feelings about any of these points,
they just seem more logical to me.
- Will