Bugzilla

I'm working on [expanding one of the notifications tests](https://phabricator.services.mozilla.com/D261217) to handle new messages arriving in EWS. If you are fixing this bug and want to compare behaviour with IMAP, run the test, it logs all of the folder property changes that happen. 

Basically, if `BiffState` changes to 0, or `NewMailReceived` changes to anything (this isn't a real property, but it's the "change" we're listening for), the notification code does its thing. This is poorly documented and untested, so I'm still having to work out the true intent of some things. E.g. I don't think there are supposed to be notifications if it's _the user_ that checked for mail (`server.performingBiff == false`), but there are if an automatic change happened (`server.performingBiff == true`). But the current behaviour isn't at all consistent with that theory.

I'm working on [expanding one of the notifications tests](https://phabricator.services.mozilla.com/D261217) to handle new messages arriving in EWS. If you are fixing this bug and want to compare behaviour with IMAP, run the test, it logs all of the folder property changes that happen. 

Basically, if `BiffState` changes to 0, or `NewMailReceived` changes to anything (this isn't a real property, but it's the "change" we're listening for), the notification code does its thing. This is poorly documented and untested, so I'm still having to work out the true intent of some things. E.g. I don't think there are supposed to be notifications if it's _the user_ that checked for mail (`server.performingBiff == false`), only if an automatic change happened (`server.performingBiff == true`). But the current behaviour isn't at all consistent with that theory.

At this point in time I've modified the test to check that what I think is supposed to happen, happens. Obviously it's not all passing.