Re: [w3c/ServiceWorker] Expose GeoLocation to workers (#745)

> They're designed to wake up for an event, briefly handle it, then go away. This should be especially brief if the user doesn't have a tab open to the site.

No they are _not_. This has been explained to you many times before by wonderfully talented people like [Ben](https://github.com/w3c/ServiceWorker/issues/840). Simply repeating a lie/conjecture over and over does not make it any the more true.

I hope the irony of you literally just typing this: -

> You can't simply state these things, you need to give reasoning. By showing your working, it may become obvious that your understanding of these features is incorrect.

over [here](https://github.com/WICG/BackgroundSync/issues/141#issuecomment-306433580) was not wasted on anyone.

The fact that the spec says a UA _**may**_ terminate a SW at anytime by no means implies inevitability or compulsion to do so and, if the two quoted implementations are anything to go by, it simply doesn't happen. I'm not sure if you're FUDding intentionally or you simply don't understand the performance benefits of Service Worker recycling.

Even with the stipulated restrictions on global state, Service Worker re-use is extremely desirable for the performance of a subsequent Push Notification or Travel Events as well as the reduction in the device resources needed purely to spin-up/run-down a service worker. Why would one _not_ leverage this low-hanging performance fruit? In fact, in the absence of memory and/or CPU resource load I can't see why Service Workers don't enjoy extended lifespans. 

Anyway, "To reiterate my previous comments: " See how the ServiceWorker paradigm fits shimlessly with user movement: -

Travelling->Travelling->Travelling->------------At Rest---------> Travelling scenarios. 
< ------- Service Worker 1 ------------>...................................................<-- SW 2 --->

Then, of course, is the beauty of not having to foreground and/or instantiate a client/complete UA if all the SW wants to do is tell fleet management (or facebook friends) that you've moved! This is a marriage made in heaven.

Yep, Service Workers and Background Geolocation truly are a Dream Team!

WRT other FUD on wake_lock sledge hammers: -

> This shouldn't be any worse for battery,

> It's possible that memory usage will be greater keeping a page open vs a service worker, but I'm not sure by how much. 

Really? Not even anecdotal? Can't "simply state these things" . . . :-(

I'll let that sort of informed opinion and scientific analysis go through to the keeper if you don't mind. There's too much to be achieved and too much work to be done.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/w3c/ServiceWorker/issues/745#issuecomment-306492755