> Jeff plugs CVS:
>
> >These days I use CVS on unix machines a lot and am pretty
> happy with it.
> >(I understand CVS is bundled with the GnuCore stuff that
> Mark Klein has
> >ported to MPE, though I've never used CVS on MPE.  If I were
> trying to set
> >up a new source management system on MPE today, I'd set up
> CVS and see how
> >well I could make that work.
>
> It actually works pretty darn well! The only reasons not to use it are
> if you're concerned it preserving MPE file attributes and non-HFS
> semantics. CVS is a POSIX application, creates byte stream files and
> is case sensitive. But, for team or distributed development, it can't
> be beat.
>
> Regards,
>
>
> M.
>

Since our pc development team is using MS SourceSafe for its version
control, I have begun experimenting with it to store my cobol source.

So far It seems to work fine by having the SS database stored on an NT
server with the Shadow and Working directories on the HP3000 via SAMBA
shares. I created a couple of command files that convert between the
bytestream and cobol formats and, even though I cannot compile directly from
the Shadow directory, the process is not that bad.

This also gives me a little more control as my developers can submit their
changes to SS and I can compile and move them into production on the 3000 by
simply copying the contents of the Shadow folders, with my conversion
command files, and streaming the compile jobs. This will make the auditors
happy since the developers, except of course myself, will not be able to
touch production code.

So far it seems to work well, now if I could just get around the limitaions
of MPE security it would be even easier.