Hi, How fundamental to darcs phylosophy is the decision of considering a repo a branch?
The reason I ask is that I would like to tell people who work with me about darcs, but there's a major problem I can't solve. That is: suppose I have tree people working on a single project. Each one is working on a different line of development. So, every person will probably have three repositories: one at work, one at home and one at a remote server with backup system that also allows syncronization between all repositories. Since people sometimes have a few different ideas at the same time, it would not be unusual to have 5 or 6 repos per individual. There's a aditional problem: since home and office are usually under NAT, we can ssh to remote server, but we can't ssh from remote server to home or office. This means every time we have a conflict (say, between some code I wrote at home and other I wrote at office) we need to create a bundle of patches in order to apply them to the remote server. So, for a 3 people environment, we have around 12 repositories, with regular small annoyances when pushing conflicting patches. Is there something wrong with my comments? Do you think there's a better way to understand that issue that I wasn't able to see? My understanding is that this could be easy to solve if we could have a branch tree. In the above situation, we would only need 1 repository at the main server. And since darcs is smart it could even tell us when to branch, by considering branching the default way to go when we have conflicts, but keeping all consistent patches as a common tree. (Solving conflicts would be delayed to branch merging, which we would do at our own time.) Sorry for this long mail, but this is the single issue I have about using darcs as our source control tool for all our work. Thanks for your attention, Maurício _______________________________________________ darcs-users mailing list [email protected] http://lists.osuosl.org/mailman/listinfo/darcs-users
