Allowing reviewers to force NeedsInformation -> PendingReview sounds reasonable, as this is most often due to the developer intending to put the app back into the review queue, but failing to notice the subtly different side-effects of the "Resubmit for review" and "Send comment" buttons.
Another requested transition is NeedsInformation -> Rejected and NeedsInformation -> QAPending. These are trickier because they would usually require you to provide extra information (a reason for rejection, or a package name + supported distro arch series for approval). We could allow the former by just filling in a blank reason for the rejection, but I'd be more inclined to only allow NeedsInformation->PendingReview, and the reviewer can then take it from there, moving it forwards to Rejected or QAPending through the usual review process.
Discussed with ajmitch.
Allowing reviewers to force NeedsInformation -> PendingReview sounds reasonable, as this is most often due to the developer intending to put the app back into the review queue, but failing to notice the subtly different side-effects of the "Resubmit for review" and "Send comment" buttons.
Another requested transition is NeedsInformation -> Rejected and NeedsInformation -> QAPending. These are trickier because they would usually require you to provide extra information (a reason for rejection, or a package name + supported distro arch series for approval). We could allow the former by just filling in a blank reason for the rejection, but I'd be more inclined to only allow NeedsInformatio n->PendingRevie w, and the reviewer can then take it from there, moving it forwards to Rejected or QAPending through the usual review process.