I continue to see a lot of false positives from this rule. It also doesn't handle reverted changes correctly; see e.g.  http://crbug.com/393243 .

Thanks for the report. I'm working on the fix to handle reverted changes correctly.

Thanks. I've also seen many bugs where e.g. some refactoring to make the requested fix possible has landed, but not the fix itself.

CL is pending review and i'll land it tomorrow. We will also go thro' the bugs which got updated with the reverted cl and remove the sheriffbot comment.

Thanks

Could you add a similar exception for bugs that have multiple human-generated comments after the last CL? That's a strong signal that they're still active.

I'm still seeing a lot of bugs where someone says, "I'll perform refactoring A first and then land fix B," and then only lands A. I'm concerned that many of those bugs are going to be closed soon.

The change to ignore issues when the last cl is reverted is added and deployed.
We went manually to the affected bugs and removed "Hotlist-OpenBugWithCL" label so it doesn't Archive automatically.

Assigning to Anthony for further inputs.

1. Reg "multiple human-generated comments" . Is "2" a good to consider or it should be more.
2. Do we also need to ignore any issues if it has "Blocked on:" label.

Thanks

1. I like the spirit of the heuristic, but I've seen cases where people come back and say this is fixed and can be closed.  The probably need human inspection anyways to make a decision, so the annotation is probably appropriate.  Perhaps having a larger grace period for secondary reeving might be appropriate.

2. Yes, that should be fine.

Here's what seems like another exception: if the bug is blocked on other not-closed bugs, it should probably be kept open, right? See e.g. issue 489441.

Added this to query to ignore blocking bugs (both blocked or has blocking bugs in it) the cl will be landed today.

Thanks

The cl to ignore blocking issues is landed and deployed.

We have manually cleaned up all the blocked and blocking issues affected by this rule.

Note: This rule is now suspended and more details can be found in issue 629092

Thanks