Implicitly unsubscribe bug contact when bug is Invalid

Sounds like a good idea, as long as the email goes to the right bug contacts as soon as the bug is reopened on a particular task.

This makes sense for the case of a bug with multiple bug tasks (although a lot of these cases could be better handled if we could transmogrify a product bug task into a distro bug task or vice versa).

But what about the case of a bug with a single bug task that has been marked rejected, but the reporter disputes this? No one will see any comments they add to the bug, which would be bad.

So maybe a good rule of thumb would be to only eliminate the initial bug contact for rejected bug tasks if there are other non-rejected bug tasks. How does that sound?

I think it would be confusing for people to sometimes get mail about a bug they'd rejected, and sometimes not. I prefer my original suggestion that a disputing reporter reopen a bug report if they want it reconsidered.

I have recently noticed this issue with the ubuntu-bugs mailing list. If a bug is incorrectly filed against Ubuntu and that task is rejected and a new one opened for Malone or Launchpad the ubuntu-bugs mailing list continues to get mail about that bug even though it no longer concerns Ubuntu. However, the ubuntu-bugs mailing list should continue to get mail if a bug with one task is rejected and commented on.

Perhaps only implicitly subscribed groups should be unsubscribed, and the rejector implicitly subscribed (but not the whole team), to ensure accurate feedback in the case of submitter responses to the rejection. My experience is that submitters will comment about the rejection, but not set the bug to a non-rejected state.

I guess "Rejected" now means "Invalid" or "Won't Fix".

Yes. I'd say Rejected now means Invalid in this context, since "Won't Fix" implies that the bug report is valid, so you'd want to receive further information about it.

Just as a motivation see https://bugs.launchpad.net/ubuntu/+source/blogtk/+bug/204895/comments/9

And here's another one: https://bugs.edge.launchpad.net/ubuntu/+source/blogtk/+bug/204895/comments/33

Even if there isn't an implicit unsubscription, having a way to unsubscribe manually would be a big help. In the case of bug #204895, xapian-developers is subscribed, and we can't see any way to unsubscribe ourselves, implicitly or otherwise.

I've reported a bug about being able to manually unsubscribe as bug #204980

On further thought, this is mostly a duplicate of bug 1342 (which I also reported). There's an important difference between "We're pretty sure this is not a bug in our project, but we're interested in any evidence to the contrary" and "This never had anything to do with our project, so we don't want mail about it". But that distinction would be most obviously expressed by the project being either present or not present in the bug report at all.

The only case in which this would be not covered by bug 1342 is where the bug report is filed against only one project, so the project cannot be removed without the report malfunctioning. But that can be addressed either by refiling the bug to the correct project (which is possible except for bug 80902), or by locking down the bug report (bug 73122).