Non-approving review tag e.g. "Consulted:"

DJ Delorie dj@redhat.com
Fri Apr 11 19:21:47 GMT 2025


Playing Devil's Advocate here for a bit...

I think this tag is a bit superfluous - every reply to a patch would
qualify for it.  What you're basically asking for is a list of
participants in the conversation around the patch?  We already record
that in patchwork with the Message-ID and in-reply-to tags.  Asking to
copy that info into git places a large onus on the patch writer - and I
admit I often forget to add this information to my own patches.  Plus, I
think that amount of "conversational" history doesn't belong in git
anyway.  That's what mail archives are for.

Perhaps all we really need is a tag that officially and more obviously
notes "opposition to consensus"?  Something like Failed-by: or
Rejected-by: ?  With some official rule that those must get converted to
Reviewed-by: (by the reviewer, of course) when the issues are resolved?

I will also note that, historically, we've removed most of the personal
contribution notes from the sources themselves, with the justification
that "we're all one group".  I think trying to capture all the personal
contributions to patch reviews goes against this previous decision.

Your proposal also doesn't have any protections against abuse; there's
nothing to stop a random person (or bot) from replying to every patch
with a Consulted: to get unwarranted credit for work not done.  Nor does
the Consulted: tag contribute to the consensus equation, since the rules
explicitly allows it to be ignored.

It sounds like what you really want is something like gitlab's
"unresolved threads" feature.  Perhaps you should argue for moving to
gitlab instead of trying to force a gitlab-like automation on our ad-hoc
system.



More information about the Libc-alpha mailing list