begin quoting MattyJ as of Wed, Dec 19, 2007 at 04:36:05PM -0800: > On Wed, 19 Dec 2007 15:39:28 -0800, SJS <[EMAIL PROTECTED]> wrote: > > >begin quoting MattyJ as of Wed, Dec 19, 2007 at 02:57:23PM -0800: > >>On Wed, 19 Dec 2007 14:40:48 -0800, David Brown <[EMAIL PROTECTED]> > >>wrote: > >> > >[snip] > >>I usually try to explain it to my developers by saying that > >>directories do not exist in Perforce. A path/branch exists by the > >>virtue of something being in it (a file.) So directories can't have > >>branches, but files can. > > > >What version-control system *does* version directories? > > > >Any of 'em? > > ClearCase is the only one I know of that does it. It's a pretty cool > feature if don't let it get out of hand, but really nice if you need to > re-arrange a directory.
The version I used last year didn't. At least, I never found a way, and none of the ClearCase advocates on the team bothered to enlighten me. And the file-rename violated the fundamental rule of version-control - it allows a bog-standard user to render it impossible to check out the project as of some previous date. Naturally, I discovered the latter problem the hard way. And then it came back and bit us a few times after that, what with point-and-click interfaces and accidental drags, drops, and renames. We also never got the remote-VOB thing to work, but had to deal with the standard checkout-edit-checkin workflow. Attempting to use the remote VOB 'capability' resulted in corrupted data, workstation crashes, and server crashes. I was underwhelmed. Of course, I had just come from using P4, which was amazingly robust, despite being run on an MSWindows server. -- ClearCase was the second-worst VCS I've worked with. MSVSS was the worst. Stewart Stremler -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
