Andrii, WDYT?

sgtm. I wanted to do that, but it was a departure from how Rietveld worked, and I didn't want to create extra differences. We can revisit that once we are in dogfood in chromium. That said, I prepared all the machinery required for this.

We should ask Andy, too.

+Andy

I’m all for less email.

Andrii, feel free to place this in whatever milestone you feel is appropriate.

Thanks gang.  Glad to know you're all on board.  I'll be patient.  :)

Btw, we might actually want to remove emailing others on CQ+1 or CQ+2 vote instead. CQ+1 in particular is close to being useless.

Well, they're not quite useless.   They tell you who CQ'd it, which the "CQ is trying da patch" emails don't, and do so much less verbosely than the "CQ is trying da patch" mails.

Gerrit does a good job of showing the CQ progress in its UI, so the CQ progress links aren't really needed as above-the-fold material anymore.  In the rare cases you want to look at it (the tree's closed or throttled) the link's always there in Messages.

I venture no one ever wants to see Bot data: { ... } beyond perhaps CQ maintainers.

Thanks for feedback.

As a maintainer I can assure you that I don't want to see bot_data either, and would much rather hide it somewhere :)

So, how about this:
CQ+2 -> send email to OWNER and maybe Reviewers (not sure)
CQ+1 -> don't send email
"trying da patch"  -> no email (Ravi -> please go ahead and land your CL :))

and for CQ status URL -> it actually deterministic and can be shown by commit queue plugin in Gerrit itself so that nobody needs to look at CQ message, other than myself debugging what went wrong :) Filed  issue 665031 

I wasn't aware different emails were sent to different people.  That sounds like a recipe for confusion.  I think I'd feel most comfortable if every human-initiated action was emailed to every human attached to the code review, whether author, reviewer, or cc'd.

I guess beyond that I only want mail for CQ failure (Tryjobs failed on the following builders...) or CQ success (Commit bot merged..), and there too I'd like every human to see that.

Different emails are already sent to diff people: posting your own comment != send email to oneself.
CQ **dry** run (which is what git cl try does by default) is not important to most reviewers, particularly if it fails on a new patchset uploaded as a result of prior review.

There is a trade off between confusion and spam, and I'd argue if CQ starts sending email on every CQ dry run failure to every reviewer, it'd add no value whatsover other than perhaps discouraging dry runs and saving capacity :D

That's a Gerrit option (Preferences > Email Notifications) which I've opted myself into the "Every Comment" setting precisely for this issue of confusion about who receives what emails.  I certainly don't mind that staying an option.

I also don't mind not seeing other people's CQ+1 or CQ+2 failures.

I do mind not seeing other people's actions.  If I write a CL and someone else +1's it, I want to know.  If I'm reviewing a CL and the author +2's it, I want to know.  If I'm CC'd to a CL and the author lands it without CQ, I want to know.  I want notification of every action a person takes, and the final merge if it lands.

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/infra/infra_internal.git/+/87bd58db6bcfadf23391cece830f38cbcaa96629

commit 87bd58db6bcfadf23391cece830f38cbcaa96629
Author: Ravi Mistry <rmistry@google.com>
Date: Mon Nov 14 16:24:47 2016

I tested the fix with https://skia-review.googlesource.com/c/4773/.
Verified that I got email for "Patch Set 1: Commit-Queue+2" but not for "CQ is trying da patch".

Marking as fixed.

Thank you!