Need to set the parent of this patch

Could you also update relnotes/next (documenting either the new flag or config option, whatever this patch ends up doing)?

At some point in this series (probably as a separate patch at the end), you should update relnotes/next to describe the API change and how to register stateful operations

In D6501#95188, @taapas1128 wrote:

I was talking about the way in which registrar takes in data. eg: see line 168 of hgweb_mod.py.

This example is not a goal in itself,

In D6503#94684, @taapas1128 wrote:

@martinvonz I haven't changed anything regarding rebase and update on interrupted states hence not added any tests.

Could someone look at the remaining patches in this series? They've been waiting for three weeks or so now.

In D6503#94684, @taapas1128 wrote:

@martinvonz I haven't changed anything regarding rebase and update on interrupted states hence not added any tests.

Do you need to update the metadata in Phabricator to tell it that D6515 is its new parent?

Results of tests are shown.

Please ignore my latest messages. I had not noticed yet that this was abandoned.

Thanks for splitting this out! Sorry about being so particular about splitting up the patch. I think it's an important part of writing patches that others will review. Hopefully it wasn't just frustrating and you can also see some benefit with it. The benefit is mostly to reviewers and future code archeologists and not to the author, so it can feel like wasting time. Remember that more people will read the patches than the people writing them (which is just you).