Re: [ServiceWorker] setTimeout/setInterval are unreliable in ServiceWorkers (#838)

@ojanvafai What is happening at the moment? I assumed it would be extremely inefficient to terminate a SW as soon as the event queue is empty, are you say that this *is* in fact what is happening?

Why would a user-agent, or operating system, terminate a SW on a device is resource rich and far from exhausted?

BTW I fully accept you original points about setTimeout/setInterval as long as they are still there with worker threads.

NB: While the GeoLocation team is implementing serviceWorkerRegistration.travelManager one could always implement serviceWorkerRegistration.calendarManager. A calendarManager event would be sufficient to re-instantiate a terminated Service Worker. (calendarManager being managed in the browser as will be travelManager)

---
Reply to this email directly or view it on GitHub:
https://github.com/slightlyoff/ServiceWorker/issues/838#issuecomment-187484086