----Original Message---- >From: Larry Jones >Sent: 22 August 2005 19:10
> Dave Korn writes: >> >> We're just having a slight semantic disagreement here; > The key piece of the CVS design philosophy that you're missing is that > the vendor branch *is* the mainline for files that haven't been locally > modified. Heh, yeh, I finally grokked that now; Mark's explanation of the presence of the "branch" directive in the ,v files made it all clear. > Unfortunately, that does indeed mean, as you discovered the hard way, > that doing an import without doing the subsequent merge, you're > frequently left with a broken trunk. That apparently wasn't a concern > to the original developers since imports are usually rare and carefully > controlled, but it has become more and more of a problem as time goes > on, which is why the -X flag has been added to import in the new feature > release. Using it for all imports (typically by adding it to your > .cvsrc file or setting the ImportNewFilesToVendorBranchOnly option in > CVSROOT/config) keeps the trunk isolated from the vendor branch. Um, I'm not 100% sure I understand.... I thought that only affects _new_ files that have arrived with the update import; files that existed already, and that I haven't modified on the mainline, will still be dragged over to the branch, even with -X, won't they? > I suggest you re-read the section of the manual on tracking third-party > sources, particularly the part about updating with import: > > <http://ximbiot.com/cvs/manual/cvs-1.11.20/cvs_13.html#SEC103> I think you're particularly referring to the bit I quoted in my first mail, aren't you, about "For files that have not been modified locally, the newly created revision becomes the head revision"? Yeh, that was the bit I didn't read properly before I got myself into this mess. Ta for your help! cheers, DaveK -- Can't think of a witty .sigline today.... _______________________________________________ Info-cvs mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/info-cvs
