Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Android does this as well, so I guess like the guy you are responding to, i don't see a real difference.


To my understanding, Android offers something similar and also allows apps to wake up arbitrarily in the background. Reasonable people can disagree on whether the latter ought to be an option or not (or, perhaps even more reasonably, can be happy that both types of devices exist so people can choose which they'd prefer). But the complaint regarding Android was that it let poorly-behaved apps drain significant amounts of battery in the background and it's not true that Apple has now added that possibility.


So you know all the technical implementation details of both?

Or are you comparing random API doc material and actual technical details?

Apple says a lot of things, not all of them turn out to be technically accurate.

For example, do you know it actually will not allow poorly behaved apps to drain battery, or is this just an assumption based on what apple says will happen?

What always happens in these discussions is people say "apple's docs say x, so it must be like x". Quite often, when people actually go and look at how it operates, it isn't like x at all.


The discussion was around architecture. Of course it may not behave as intended when actually implemented because of bugs. Or because of Apple outright lying in their communications. Sure, those are possibilities, but they're irrelevant to architectural criticism.


But you have no real architecture details, only a small number of bullet points on how it's supposed to behave and a simple but not horribly descriptive API. That isn't an architecture, that's just marketing materials. They are completely irrelevant to anything.

Given only that, you cannot possibly make informed commentary on how it will behave in practice, you are just parroting a story.

In particular, you said "it let poorly-behaved apps drain significant amounts of battery in the background and it's not true that Apple has now added that possibility."

You cannot possibly assert this with any real details to back it up, because you do not know how this architecture operates past "apple says they won't be woken up enough or run long enough to drain battery". If you have the actual details necessary to back this statement up, please add them.

Unless you are talking at such an abstract level where everything that matters is an implementation detail, in which case it's very easy to design perfect architectures that have no problems!

We don't believe people when they make crazy claims about crypto, without seeing the actual details and implementations. We should not trust Apple or Google's marketing points about their architectures when trying to make "architectural criticism".

In the end, if you really believe "architectural criticism" is possible without actual detailed design info, carry on.

But to me, that's a worthless discussion based on what are essentially talking points.

In any case, all that matters in the end is performance in the field, so this entire discussion is mostly technical masturbation until real users have phones in hands.


I'm not sure why you think I have no real architectural details. What do you know about what I know?

Regardless, I think your pedantry is based on a statement I intended to be read in a less formal fashion than you're choosing to read it. When I said that it was not a possibility for poorly-behaved apps to drain significant amounts of battery under iOS 7, I didn't mean that it was absolutely impossible under any and all circumstances including system bugs, unforeseen architectural weaknesses, or gamma ray induced bitflips. People say "Linux uses memory protection to keep processes from corrupting the memory of other processes" and we all understand that they don't mean that it's literally impossible for memory problems to ever happen.

Likewise, my point was that it is generally true that iOS 7 doesn't allow apps to wake up as often as they want. That is a design goal behind its architecture. Doubtlessly, neither its architecture nor its implementation are perfect — I'd be surprised if it were utterly impossible for an app to end up running whenever it wants in the background. But it'd be boorish to belabor that point.

The original post that has led to this discussion was saying that Android has been criticized for being designed to allow apps to run in the background arbitrarily and now iOS has been redesigned to allow that too. That's not true. And the conclusion of hypocriticalness that was drawn from this faulty premise was untrue.


"I'm not sure why you think I have no real architectural details. What do you know about what I know? "

They do not appear as supporting details of your argument, so ...

Again, you are trying to make it seem like pedantry and that i am addressing only extreme cases or bugs , and my point is you have offered no details to support any part of your argument.

"Likewise, my point was that it is generally true that iOS 7 doesn't allow apps to wake up as often as they want. "

I quoted your statement about battery life, and said you have offered no details to show this to be the case. Rather than offer details to refute that, you have now instead said "I said something different".

I'd appreciate it if you would stick on point and address my contention that you have not offered details about this statement:"it let poorly-behaved apps drain significant amounts of battery in the background and it's not true that Apple has now added that possibility."

Please offer details to back this up. IE what architectural details you know that you believe make it the case that apple has not added the possibility of apps using large amounts of battery in the background.

If the details are "apps can't arbitrarily wake themselves up", then your statement about the possibility of background apps draining battery life is easily shown to be wrong, and i'll be happy to refute it for you. If it is something else, i'd love to hear it, so i know exactly what argument i am addressing.

Let's stay with this one part of the argument, please.


Don't worry about the trolls, your points made sense to me.

The crazies have come out in this thread.


  > But you have no real architecture details, only a small
  > number of bullet points on how it's supposed to behave
  > and a simple but not horribly descriptive API.
And WWDC presentation.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: