Please use Discourse to discuss Phabricator. (And please don't bump unrelated tasks from five and a half years ago.)

Is this really a harbormaster problem though? I would have located it within Differential.

The broader structural issue here is:

A second issue is that some of the "publish" behavior depends on the commit/revision relationship. This can still be sequenced correctly entirely in the publish daemon, but "close related revisions" and "publish the commit" are not entirely independent.

This appears stable.

See PHI1807. At time of writing, arc land can delete the local master if you land it onto itself. This isn't a big deal (it gives you the command to get it back), but not intended and undesirable. Although it isn't recommended, arc land is supposed to support working in master and landing master into itself.

A sometimes-suitable replacement would be some bin/request mechanism which simulates a web request in process. There are some limits to how authentic this can be, but for some substantial subset of hangs it may still be useful.

PHI1776 is another report of this behavior.

A related issue is that it is difficult to identify the set of "published" hashes in the general case. We would like arc branches to show all unpublished history, but stop at published history.

The major issue I'm trying to deal with now is that if you have a tree like this:

P2143 has some graph reachability code that I don't actually need yet, but may by the time I get here.

The current ls-remote script as written does not actually work; I think it only returns branches in the remote that:

The actual extension is a 100-line Python script...
The real Pandora's box this opens...

The real Pandora's box this opens is that now that arc is exposed to whatever issues arise with the raw Python/Mercurial API, there's a much weaker argument against not writing more extensions.

I put the extension in phabricator/ earlier by accident, but that was just me goofing my local paths. The extension is purely client side: hg can do ls-remote just fine (it's just "do hg pull, but stop half way through"), there's just no UI command for it.

Would the use of an extension make additional setup necessary for observed repositories vs. hosted repositories, or are observed repositories out of scope for these set of changes?

I am considering reviving the command server since I don't remember it having any critical failures, but offhand:

I'm going to try writing a Mercurial extension instead.

I believe this is roughly the form we end up with, which is completely ridiculous and extremely difficult to test or have any confidence in:

As a tool developer, this kind of thing is infuriating: