You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Okay, I can accept that; it isn't documented AFAIK, but it makes sense that we wouldn't want to run property observers during deinit, lest there be side effects. But now hold my beer and watch this:
The observer is called! Okay, that is incoherent. If it's wrong to call property observers during deinit, it is really wrong to call property observers after deinit.
The text was updated successfully, but these errors were encountered:
@mattneub Sure – though in the compiler it's the same logic that determines whether or not to perform a direct access, so ideally a fix for one would also be a fix for the other (that is, depending on compatibility implications – I believe the defer-within-init to trigger observers pattern is used by quite a few people).
Right. defer probably shouldn't have this behavior, but of course any method you call during deinit will result in the observers running, so it's not like it's not allowed at all.
Additional Detail from JIRA
md5: 659d8175f6537ca6db993b3627e234f6
is duplicated by:
relates to:
Issue Description:
The property observer here never fires:
Okay, I can accept that; it isn't documented AFAIK, but it makes sense that we wouldn't want to run property observers during deinit, lest there be side effects. But now hold my beer and watch this:
The observer is called! Okay, that is incoherent. If it's wrong to call property observers during deinit, it is really wrong to call property observers after deinit.
The text was updated successfully, but these errors were encountered: