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.
> OK, so if I do specify a vendor branch when I do the import then on > the first such import the source files get revision 1.1.1.1 and are on > the branch (and of course also on trunk). Why "of course"? There's quite a bit of magic going on at that point... > But now, when the next load of files are imported using the same > vendor branch, what happens then with files which have been altered? Depends on your definition of "altered" here. If you mean altered between the two imports then yes, the new revision will be 1.1.1.2 and there will be no revision on the trunk. Yet, the Vendor branch is special in that you do not have to explicitly update to it in order to get its revisions. Simply checking out your (multi-)imported module without specifying any tags, revisions or dates will give you the most recent revisions from the vendor branch. In this case the 1.1.1.2 revision will actually be the HEAD! That is because Import changes the "default branch" of files, i.e. the branch that will be retrieved when you do not explicitly request a specific branch. It is also the branch to whose tip revision the HEAD symbol refers. Once you modify and commit a file however, the new revision ends up on the trunk (typically as 1.2) and the default branch for that file is changed back to the trunk, hence all future commits will go to the trunk as well. You could however still do update imports to the vendor branch and then merge those changes with your local changes. This latter part of course only really makes sense if you do use Import for managing third-party sources. In your case, I would just use Import (with vendor branch) to get all those historic revisions into your repository and then continue work as usual. After the initial imports files will all have revisions like 1.1.1.1 or 1.1.1.8 but as soon as you start committing manually (in contrast to using import) they will continue life as 1.2, 1.3, 1.4, etc., just as if there had never been a branch involved in the whole process. > Will they not wind up as 1.1.1.2 and no revision on TRUNK? > If this is so then we have effectively bypassed TRUNK for everything > except the very first import.... Actually no. Even the very first import goes to the vendor branch instead of the trunk. However, due to the inner workings of Import this does not imply that the 1.1 revision is "dead" as it would be if you manually added a file on a branch initially. I guess the biggest problem here is understanding the "default branch" concept and that semantically it is yet subtly different from both the trunk and the vendor branch. I'd say, why don't you just try it all on a test repository and see for yourself? ;) Hope this helps. -- Oliver ---- ------------------ JID: ogiesen at jabber.org ICQ: 18777742 (http://wwp.icq.com/18777742)