{"thread":{"id":"49318","subject":"Allow git bisect in reverse mode?","startedAt":"2018-09-11T19:53:02Z","lastAt":"2018-09-11T19:55:52Z","messageCount":2,"participants":["Brad Davis","Ævar Arnfjörð Bjarmason"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"357901","messageId":"CAMK-fcye+nNyB9pFrp1nQwQC1+sQWMt_4yq2f7svJUiB=ZnnJA@mail.gmail.com","threadId":"49318","inReplyTo":null,"subject":"Allow git bisect in reverse mode?","fromName":"Brad Davis","fromEmail":"bdavis@saintandreas.org","sentAt":"2018-09-11T19:52:58Z","receivedAt":"2018-09-11T19:53:02Z","isPatch":false,"sender":{"key":"bdavis@saintandreas.org","avatar":null},"body":"I'm currently trying to determine when an issue was fixed in a long\nlist of commits.  I attempted to do this by running `git bisect` and\nmarking the commit where I knew it was broken as bad, and the tip as\ngood, but I got back an error message saying that good revs weren't\nancestors of bad ones.\n\nI'm currently working around this by mentally reversing the meanings\nof good and bad, but it feels unnatural and error prone to have to say\n`good` when I encounter the failure and `bad` when i don't.\n\nCouldn't git simply do the same thing and internally reverse the\nmeaning of good and bad if the bad revision is an ancestor of the\ngood?\n"},{"id":"357902","messageId":"CACBZZX7EijsVZcwb1BkZdYEzKJbyRWGJNBKXXNRt_i47R0o08w@mail.gmail.com","threadId":"49318","inReplyTo":"CAMK-fcye+nNyB9pFrp1nQwQC1+sQWMt_4yq2f7svJUiB=ZnnJA@mail.gmail.com","subject":"Re: Allow git bisect in reverse mode?","fromName":"Ævar Arnfjörð Bjarmason","fromEmail":"avarab@gmail.com","sentAt":"2018-09-11T19:55:39Z","receivedAt":"2018-09-11T19:55:52Z","isPatch":false,"sender":{"key":"avarab@gmail.com","avatar":"https://avatars.githubusercontent.com/u/45301?v=4"},"body":"On Tue, Sep 11, 2018 at 9:54 PM Brad Davis <bdavis@saintandreas.org> wrote:\n> I'm currently trying to determine when an issue was fixed in a long\n> list of commits.  I attempted to do this by running `git bisect` and\n> marking the commit where I knew it was broken as bad, and the tip as\n> good, but I got back an error message saying that good revs weren't\n> ancestors of bad ones.\n>\n> I'm currently working around this by mentally reversing the meanings\n> of good and bad, but it feels unnatural and error prone to have to say\n> `good` when I encounter the failure and `bad` when i don't.\n>\n> Couldn't git simply do the same thing and internally reverse the\n> meaning of good and bad if the bad revision is an ancestor of the\n> good?\n\nThis feature already exists. See --term-{old,good}=* in the\n\"git-bisect\" manpage.\n"}]}