(In reply to comment #31)
> Can we get some tests too please? Pretty please, with sugar, and spice, and
> everything nice?
Litmus is the only way I can think of testing this, as this calls into the native toolkit and any tests I can think of would be no different to what we have for the toaster.
(In reply to comment #33)
> Instead of creating NotifyCallbackInfo, why don't you make the notify_
> callbacks take the nsAlertsIconListener as their closure-data?
I guess so, I was worried the nsAlertsIconListener would be destroyed before the callback happened but I suppose that's what XPCOM is for...
(In reply to comment #31)
> Can we get some tests too please? Pretty please, with sugar, and spice, and
> everything nice?
Litmus is the only way I can think of testing this, as this calls into the native toolkit and any tests I can think of would be no different to what we have for the toaster.
(In reply to comment #33) tener as their closure-data?
> Instead of creating NotifyCallbackInfo, why don't you make the notify_
> callbacks take the nsAlertsIconLis
I guess so, I was worried the nsAlertsIconLis tener would be destroyed before the callback happened but I suppose that's what XPCOM is for...