Users are having issues setting the approval flag for patches I submit. Is this intended, or is there a permission that isn't set for my account?
@Richard Can you add example issues in the "Blocks" field in this issue so that we can investigate please. In general, the baseline/default permission model in Bugzilla is: Issues and Objects (attachments, etc) can be modified by: A) The user that created that object (issue, attachment, etc) Examples: - Creator (Reporter) of issue can close the issue - Creator of attachment can set attachment field (flags, etc) values B) Users with the explicit or implicit (group) "editbugs" permission (like users matching *@FreeBSD.org) Example: - Committers (and those with editbugs permissions) can set any issue/attachment field values. C) For flags, users that match the user (email) value of the flag when it has been set (?, meaning 'requested') Example: - If a flag has no current "?" (requested) value, only editbugs users can set it (to +, -, cancel (null)) - If a flag has current value ? <email>, user matching that email can set/update its value (eg: maintainers) What this effectively means is that a flag must be set to ? <email> FIRST, before a user matching that email can set/change the flags value. In practice, the Bugzilla autoassigner script currently does this (sets ? <email>) on the maintainer-FEEDBACK flag (on issues) if the maintainer of new issue can be determined (using the summary contents: "category/portname"). This doesn't occur for the maintainer-APPROVAL flag (on attachments), so it needs to / should be set by the attachment creator (or committers/triagers) before a user (matching the <email>) without editbugs permissions can set/change its value
@Richard Apologies, I missed the existing See Also link (to Bug 220286)
(In reply to Kubilay Kocak from comment #1) Thank you for your detailed answer. It was perfect and I completely understand the issue now, I will be sure to start setting the request field as I have been failing to do. (In reply to Kubilay Kocak from comment #2) No problem, I know I did lack details when I initially opened the bug, and I should have gave a little more information. Going to close this as it is not a bug, simply lack of knowledge of usage on my part. Thanks again.
That quirk of usage creates a problem: if someone other than a maintainer submits a patch and doesn't flag it maintainer-approval?, the maintainer is then unable to flag it maintainer-approval+. Expecting all bug filers to know this behaviour and remember to set maintainer-approval? on a submitted patch has, in my experience, been unreasonable.