Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
On Mon, 24 Jul 2006 15:25:14 -0400, "Nick Duane" <nickdu at msn.com> wrote: > >Ok, that explains it I guess. However, from your last sentence you make it >seem like the feature would be difficult if not impossible to implement. >But it almost does it already. import will create the directory as long as >the directory contains at least one file (at least that appears to be the >behavior from my observations). So some directory information must be >making it to the server. Probably some confusion here. Cvs import will handle *files* like all other cvs commands and if these files are deposited in a directory structure then thes directories also get created. But directories are nothing else but a container structure exactly like the one from which you import. Cvs does not version directories and empty directories are not created unless they contain subdirectories that contain files. When you check out a module and there is a directory that exists on the server because at one time it contained an active file (and thus the RCS file that represents that now dead file also exists) but this file has since been cvs removed, then the checkout will *not* create the empty directory on your workstation. I have used cvs import *many* times in client/server configuration and it does not ever fail to work correctly. But by correctly I mean that it does not create a server side directory just because one exists on the client but is empty. /Bo Berglund