I'm working on CVS conversions and discovered that my branch conversion algorithm is flawed. It means that the postgres repository I created is probable wrong for all branches. In working this out I wonder how fossil handles this situation: 1) I create a branch b1 2) add fie hello.c to b1. 3) switch to trunk. 4) add an other file hello.c to the trunk. Are the 2 files hello.c related? I would expect not, because they don't have a common ancester. When I merge the branch with the trunk
fossil says there is a conflict I assume since #trunk:hello.c is not related to #b1:hello.c. What does one do then? If I go ahead and commit then hello.c is still what it was on the trunk b1:hello.c is ignored. I think fossil should not allow the commit on this merge Is the lesson that the merge is not possible? (Which i think is right) And the solution is then to rename one of the files so that the conflict goes away? Side Note In CVS files are allways on the trunk even if created on a branch. When created on a branch the trunk version has state dead. and the branch version is branched from this not existing 1.1 version It also showed that I have been doing the wrong thing when adding files to CVS on a branch. I would add them to the trunk and then do cvs tag -b b1 hello.c While the correct way seems to be cvs co -r b1 cvs add hello.c cvs commit -- Rene
_______________________________________________ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
