This should be ready to go now, although I'll wait until we're actually ready to act on it to promote and deploy it.

The biggest remaining issue on this workflow is that the pre-verification workflow is a confusing mess, per T7173.

In particular, T9263 may have motivated development of a local patch.

We haven't made any changes to this in the upstream recently that I can recall or find in git log, so if this behavior really has changed my best guess is that there's a local patch affecting it. The last upstream change which was expected to affect this was rPbcb957b (April 27).

Thanks for the report! I'll take a look at this.

V and Q with a single digit are ignored because of conflation with "Version" and "Quarter" in common usage, respectively.

This might be a more general issue:

I suppose we can version detect and just drop this flag. It only matters if the user also has local pre-push hooks which prevent the push, which are presumably/hopefully very rare.

I'm worried it may intimidate users who are worried they can not live up to the requisite level of professionalism.

It looks like some other activity also occurred in the same transaction set. Does the email reveal what it is? That is, do the first two lines look like this?

Older versions of an image are available in the table below. For example, this shows iterations to a high quality professional picture of a cat:

Can you show me the mail headers for the mail you received but did not expect to receive? In particular, the X-Phabricator-Mail-Tags header?

Corp + Org will also move here after CORGI so this will generally make a good deal of sense as things consolidate.

Ignoring the actual invite workflow for now, the major technical issue with supporting this is handling the eventual merging of accounts.

T9302 has a followup bug on this, there may be an issue with the interaction between this query and Owners packages.

You can just type 500 email addresses into this box (and, at least locally, it works fine, although admittedly I haven't done it in production personally):

I can't find an upstream version of this external newer than the one we're using, so I think we can close this. Last update I see on Sourceforge is from 2010.

I marked D14023 as resolving this since I think I took care of most of the really silly/old stuff and this now works in a fairly reasonable way, at least for the most part.

That fixes the link thing, if anyone runs into the other thing we can fix it once we have a repor.