[00:04] SamB_MacG5: everything runs slowly with --coverage [00:53] what determines that a file is a branch === mark06 is now known as mark06-away === mark06-away is now known as mark06 === mark06 is now known as mark06-away === mark06-away is now known as mark06 [01:20] A file isn't a branch ;p [01:30] but like for it to be a branch ... the dir only needs a .bzr dir within it right? [01:34] No, a .bzr dir just means there's a bzrdir there. [01:34] A bzrdir _may_ contain a branch. It _may_ contain a repository. It _may_ contain a working tree. [01:34] It may contain any 2 of the 3, or any 3 of the 3. It's probably technically possible for it to contain 0 of the 3 too, though that would be stupid. [02:16] fullermd: of course a .bzr dir can contain none of them [02:16] * SamB_MacG5 runs "mkdir .bzr" [02:21] * SamB_MacG5 is rather puzzled by the way "bzr bisect run" does a "bzr bisect start" internally, but the tests try to set up a bisect range *before* that, which will be cleared out by that ... [02:23] You're supposed to do things _before_ I preemptively declare them stupid, y'know :p [02:26] well, I didn't actually do it, so there! [02:28] and why am I not reassured by the fact that none of the scenarios covered by the bisect tests have uncovered any real bugs in my attempt to add revspec ranges to "bzr bisect start" and "bzr bisect run"? [02:29] granted, maybe the bugs only exist in my head, but ... [02:29] it would be nice to find those too