See: https://www.pivotaltracker.com/story/show/103103536 When a new version is submitted, it should automatically disable all existing *pending* versions. This prevents an add-on to clog the review queue with multiple versions when really, only the last one matters. (Descoped: ~~We must also make sure that an automatically disabled version can go back to pending if necessary, if more recent pending version was removed~~ -- Having versions going back to pending is just too much trouble, at least for this story.)
Andrew, Currently we allow reviewers the option to reject/publish Obsolete versions which seems weird. Should I file a bug for that?
(In reply to krupa raj[:krupa] from comment #2) > Andrew, > > Currently we allow reviewers the option to reject/publish Obsolete versions > which seems weird. Should I file a bug for that? Rejecting an Obsolete version seems a little weird I admit. Assuming it actually works (e.g. publishing an Obsolete version does all the necessary to make it Published) its pretty harmless I guess. And it works around the descoping in #c0 if we need to approve such a version.
The new added version of an add-on has the status "pending" and the old version of the same add-on is made "obsolete". The "obsolete" version can only be downloaded. Verified on MP-dev FF44(Win 7). Postfix screenshot: http://screencast.com/t/bnbiqkcT47 If there is something else to be checked, please reopen.