Possibility to create groups of users with more permissions

[llrs]

During last meeting Martin Maechler asked if there was the possibility to have a group of people to confirm bugs and/or to disable notifications for status changes. Reporting this here before I forget for the next meeting.

I haven't found any way to limit the notifications sent but I found that it is possible to create groups with restricted permissions on Bugzilla.

If I understood correctly the documentation such group could be set up with the following options:

Product R:
r-confirm: SHOWN/SHOWN, CANCONFIRM, EDITBUGS

Here confirm would be the user group of people that would be allowed to confirm bugs. Who belong to that group can only be set by the administrator of the Bugzilla. The other flags on the same line are different permissions these users would have:

• SHOWN/SHOWN: just bugs shown (ie not security bugs). Not much documented
• CANCONFIRM: Allows them to confirm bugs.
• EDITBUGS: would allow to change the Os and other data related to the bug. I'm not sure if this also grants being able to close issues, if this is too much, it is not really needed to confirm bugs but maybe it is interesting to confirm versions, or limit the OS a report affects.

[hturner]

This was discussed in the September 2021 meeting (minutes) where Martin Maechler noted that even marking as CONFIRMED creates noise (in terms of email notifications). So the idea of giving triagers more permissions on Bugzilla seems unlikely to be progressed.

[llrs]

I created the issue precisely after that meeting. I understood from the meeting that even if the group creates emails/noise it could also help raising the signal/noise ratio.

I think he mentioned this possibility some meetings later in a passing comment (so, it might not be in the minutes) and I didn't feel he was against it but wondering if it was possible and helpful.

[hturner]

Ah okay, I got the order wrong (minutes aren't perfect!). I'm happy to reopen this to allow further discussion.

[llrs]

I could have understood it wrong, but I thought that a clear confirmation might be helpful.

[gmbecker]

I had the same impression as @llrs that @mmaechler (and @bastistician ) were not opposed to this but probably best to just see what they say directly about it.

Feedback from @s-u and other R-core members would probably also be good.