I've had some pretty hairy merges in the past, I often needed to do some research in between merge sessions. It annoyed me that I had to put the whole repository in "merge mode" while I was doing this. And the only way to tell if the repository is in "merge mode" is to look at the rev-id and see if it said "NNNNNN+MMMMMM" or something like that.
When I was working in that mode (non trivial merge) what I wanted was the ability to craft a specific "merge-rev" using individual steps, and then at the end specify the two branches I wanted to merge, and the revision that I wanted to be the result of the merge. I guess you don't have to use a revision for the merge-result, you could just use the current working state. I ended up wanting a fair number of convenience scripts that worked like teamware resolve's command line interface. Things like 'diff parent child', 'diff ancestor parent', etc, etc. When I want to run through lots of conflicts that will be fairly easy, then I want a GUI so I can step through them efficiently. When I need to research a hairy merge, I want a GUI focused on one specific file without the rest of the cruft so I can research that one file. But that might happen on a different day as a different task. SIMPLE conflict sets should continue to use the current set of steps. That works fine. But it would be nice if there was a less convoluted way to do it in multiple steps. I ended up creating a snap shot of the sources (a repo with a throw-away history) to store my progressively implemented conflict resolutions. --chris This message posted from opensolaris.org _______________________________________________ tools-discuss mailing list tools-discuss@opensolaris.org
