In D8973#136734, @martinvonz wrote:

In D8973#136435, @sebhtml wrote:

In D8973#136267, @martinvonz wrote:

For the record, I'm -1 on this, for the following reasons:

Comparing with hg push/pull doesn't seem right. Those command *have to* exchange ancestors (e.g. hg push -r abc123 doesn't exchange just abc123). Let's look for other examples. Oh, hg export has a -B flag (I had no idea). Interestingly, it turns out to use semantics similar to what my guess about your hg log -B was :) It apparently relies no scmutil.bookmarkrevs(). Looking at other callers of that function, I found that hg email and hg strip also have -B options with the same semantics. I'd be much more comfortable taking a patch that used the same semantics for hg log -B. It's a little unfortunate that that won't match the bookmark() revset, but that's still much better than adding a third way of matching revisions (I count hg push/pull -B as matching the revset semantics of giving you only the commits pointed to).

In D8973#136174, @pulkit wrote:

Folded next couple of patches in this one and queued. Thank you!

In D8973#136267, @martinvonz wrote:

For the record, I'm -1 on this, for the following reasons:

I also ran all the tests.

It seems to work as expected:

I added a test here https://phab.mercurial-scm.org/D9060

In D8973#134422, @pulkit wrote:

Thanks for the patch. Can you add some test case for this?