Gounder27, > I don't want to label all the 10000 files > as "patch1" because for patch1, we > only modified those 100 files.
Firstly everything everyone else has said I agree with - and importantly that CVS and ClearCase are not interchangable and each have strenghts and weaknesses. Chances are your employer spent up to $10K on the ClearCase license for you and as much again per person on training and training materials and consultants - spend the same time and effort with CVS and you'll get the same result - an SCM system that efficiently fulfills your business requirements. Also as previously pointed out: in CVS you should always tag all files. That all said I think it's an interesting question. Note: I use CVSNT mostly (free like CVS, runs on unix/linux like CVS, supports changesets, merge tracking, commit ids, audit etc) so I don't know which commands I may refer to that are syntactically incorrect for you. The most generic answer I can think of is to use merge: * cvs checkout -r may_release module * cd module * cvs update -j patch1 Patching the 100 files that were different for patch1 must have been a pain - in CVSNT I'd have just committed them with the patch number "cvs ci -B 1" and then I could use "cvs up -B 1" - but what I've written above should work now that you do have them tagged. A variation of the the 'cvs up -f' option (force match on HEAD if revision not found) is ultimately what you are looking for, ie: "cvs up -f may_release -r patch1" which is interesting - but not currently implemented. I've made a note of it for future reference though - thanks for the suggestion. You can track it's progress here: http://customer.march-hare.com/webtools/bugzilla/ttshow_bug.cgi?id=5296&tt=1 Regards, Arthur Barrett
