Typical Gerrit usage outside of Chromium has two-levels of Code-Review labels, where +1 indicates "Looks good to me, but someone else must approve" and +2 indicates "Looks good to me, approved".
The Chromium Gerrit setup was confusing until I found the FAQ and the note that "Only committers can set the “Code-Review +1” label." (https://polygerrit.appspot.com/). Specifically, it is confusing that committers set Code-Review +1 (rather than +2) and non-committers don't have any flags they can set, just comments.
I'd propose that Chromium adopt the move typical usage, where all reviewers have the ability to set "Code-Review +1" and only committers are able to set "Code-Review +2", which gates submission. This also maps to the old Rietveld behavior better, since there's no replacement for a non-committer "lgtm" right now.
Typical Gerrit usage outside of Chromium has two levels of Code-Review labels, where +1 indicates "Looks good to me, but someone else must approve" and +2 indicates "Looks good to me, approved".
The Chromium Gerrit setup was confusing until I found the FAQ and the note that "Only committers can set the “Code-Review +1” label." (https://polygerrit.appspot.com/). Specifically, it is confusing that committers set Code-Review +1 (rather than +2) and non-committers don't have any flags they can set, just comments.
I'd propose that Chromium adopt the move typical usage, where all reviewers have the ability to set "Code-Review +1" and only committers are able to set "Code-Review +2", which gates submission. This also maps to the old Rietveld behavior better, since there's no replacement for a non-committer "lgtm" right now.
I personally totally agree. Unfortunately, enough other key chromium stakeholders disagree that we're not going to be using Code-Review+2 in chromium and related projects in the immediate future.
There's a chance we'll be forced to change by some upcoming Gerrit changes, and there's a chance we'll naturally migrate after people get used to the new tool, but we're not going to include switching from a 3-state system to a 5-state system along with the rest of this migration.
If this ever changes, it will be due to other factors; keeping this bug around to track a single request is not reflective of the probability of this being addressed.
Comment 1 by sanfin@chromium.org
, May 25 2017