Il 14/03/2011 01:30, Kristis Makris ha scritto: >> - files in 4 different branches were involed > > How is it possible to do a single commit against multiple branches in > CVS? Is that what you did?
We use CVS in a quite unusual way, although perfectly legal. We have a big product (a web application) which is made up of different modules. A version of that web application consists of a set of different versions of the modules. Each module has a dedicated path in the CVS repository and each module version is mapped to a branch. In a schematic way: Product version 1 may consist of: - module 1, version 1 - module 2, version 3 - module 3, version 1 Product version 2 may consist of: - module 1, version 2 - module 2, version 4 - module 3, version 1 Module 1 is in CVS under /modules/module1, Module 2 is under /modules/module2 and Module 3 is under /modules/module3. Then, for each version there's a branch, so that: - version 1 of module 1 is named MODULE1_v1 (only /modules/module1 and its subdirectories contain files in this branch) - version 2 of module 1 is named MODULE1_v2 (only /modules/module1 and its subdirectories contain files in this branch) - version 1 of module 2 is named MODULE2_v1 (only /modules/module2 and its subdirectories contain files in this branch) and so on. In this scenario, when we fix a bug or implement an enhancement for a particular product version, we may need to touch files in more than one module versions, so in more than one branch. From a user point of view, all the changes are to be seen together, because they all serve the fix of the same bug, so it makes sense to group them together in an "atomic" commit, with just a single comment log. > It seems that perhaps you committed multiple times, once per branch, > although it may have appread to all gone through at the same time > because the IDE was issuing these commits. If so, it may be easier to > identify the problem if you issue the commits on your own on the > command-line. Yes, it would help to diagnose the problem... I may try to see if I can investigate further. >> make sense to have this approach even to group together two different >> SVN changesets sharing the same comment log/bug#/committer, if the >> second one is issued soon after the first one (maybe the committer >> forgot to commit one file together with the previous changeset?). > > The revision numbers would be different, but given the same logs they > may be worth consolidating. I know I forgot to commit files before and > lazily pressed the up arrow on the command-line and used the exact same > log message. Exactly! >> What do you think about this? > > I think this idea would work. Unfortunately I don't know Perl at all, otherwise I would have tried to help you to achieve this result. Do you think this is something enough interesting to be worth an implementation effort by you? Thanks for your support. -- Mauro Molinari Software Designer & Developer E-mail: [email protected] _______________________________________________ scmbug-users mailing list [email protected] http://lists.mkgnu.net/cgi-bin/mailman/listinfo/scmbug-users
