Matt

After reading that thread, I agree that it's related, but I think that even if we choose not to simplify ancestor merges in general without a `--simplify-merges` flag, semantically `--skip-empty`...

Until that command has been implemented, I've had success with `git fetch && jj rebase -s 'all:roots(mutable())' -d 'trunk()' --skip-empty`

> Oh, but `jj upload foo --bar` will have to be rewritten to `[["fix", "foo"], ["git", "push", "--change", "foo", "--bar"]]`? Then, we'll need an explicit `*args` syntax to specify which...

Speaking for gerrit users, with my new behavior added in `jj gerrit upload`, it directly maps the `jj` change ID to a gerrit change. This means that which of the...

@Wyverald might know more. You might want to try USE_BAZEL_VERSION=6.5.0 (with bazelisk). Bazel 7.0 flipped bzlmod on by default, so it could be related to that.

For anyone who doesn't want to read the slack thread, the core issue is that go mod tidy won't populate go.mod with direct dependencies unless they haven't upgraded to go...

FYI, a decent workaround is to use the `--output_base` flag. It's not ideal, but if you're switching between configs, you can also switch between output bases and have a whole...

That's correct, yeah

> Background snapshotting processes like VisualJJ or `watchman.register-snapshot-trigger` inevitably have this problem. To mitigate this class of problems, I think we'll need a wider lock to serialize local operations. This...

> Would it be possible for watchman to simply queue up changes and _wait_ until a previously triggered snapshot operation has completed before requesting the next one? AFAIU, this would...