Support skipping a version in bisect #950
Assignees
Labels
Comments
|
@codebytere I already submitted PR for the assigned issue. Can you assign me this? |
|
@Hanaffi I don't see a PR from you in the queue. 🤔 |
|
@erickzhao Its there: #1092 :) |
|
@codebytere @erickzhao Can you give some hints on this one? IDK where to start |
|
Hello. As git bisect uses the same searching procedure for both bad and skipping situations, is it a right approach to just add a skip button in the renderer? |
|
Hey @chan95821, I think the git code actually uses a pseudo-random number generator to determine the next commit to test: https://github.com/git/git/blob/a36e024e989f4d35f35987a60e3af8022cac3420/bisect.c#L620-L651 |
chan95821
added a commit
to chan95821/fiddle
that referenced
this issue
Mar 10, 2025
Bias was set to better choose an appropriate version to test based on the skipping algorithm in git bisection. Because RunnableVersion does not indicate whether a specific revision has been skipped, it will choose a random version regardless of its skip status.
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
Similarly to
git bisect skip. Sometimes, a given version is not a good one to test, e.g. because it crashes on launch (looking at you, v8.0.0-nightly.20191021). I'd like to be able to test a nearby version instead, since I don't know whether to mark this particular version as good or bad.The text was updated successfully, but these errors were encountered: