It's incredible the answers non-technical people come up with to explain technical things. The reason RAGE takes more CPU/GPU power than Quake III is because it does more. Not because John Carmack wants your gaming PC to feel slow. The same is true of Apple when they ship a new app that does more.
Samsung is just a generic hardware vendor for Android PCs, like Dell. They don't create operating systems or apps.
It probably also has a psychological component. Just seeing a new generation gaming PC makes mine feel slow.
RAGE is doing substantially more of everything than Quake III, so I think the difference between Quake III from 1999 and RAGE from 2011 is a poor parallel to draw with iOS6 to iOS7.
A better comparison is that Windows 8 runs faster than Windows 7, which in turn ran faster than Vista. Linux kernels continues to improve in performance too.
I think people substantially underestimate how hard the iOS GUI pushes the performance wall, from one release to the next.
There was a talk at WWDC this year (session 419, if you want to watch it) that goes into some detail about how the blur effects on e.g. the lock screen, notification center, control center, toolbars, etc. work.
You can watch the talk for the details, but as a practical matter several of these effects involve seven (!) full GPU passes, just for these effects alone, let alone whatever UI is under or over them. And not even the very latest-gen devices can do it consistently full-screen at 60fps. iOS displays motion underneath blur effects in very small areas (e.g. a toolbar) on new devices but if you think about it, you never see motion underneath, say, the lock screen. The reason is that not even the latest iPad Air has the GPU to apply some fullscreen blurs in enough time to meet the draw deadline.
This actually surprises a lot of people--I can't count the number of designs I've gotten where some designer has made a large blurred UI element and expects something to move underneath it. Which, depending on the size of the element and the target hardware you can sometimes get away with, but you are in for quite the lengthy performance test, and a complicated spectrum of fallbacks for all the devices that just aren't fast enough.
And that's just one blur effect. You take a look at all the other "new" iOS things that are less visible, like the endless stream of new background technologies (background fetch, silent push, NSURLSession, multi-app audio routing, various iCloud syncing stuff) the new bluetooth stuff (iBeacon, HomeKit, Multipeer), turning WebKit into an LLVM compiler, and all the other things that are going on in iOS it really is "doing more of everything". I remember a time when iOS was basically just a few OS processes plus 1 active app, and as we speak ps on my phone lists 87 processes.
Yes. The UI equivalent of automotive tail fins is expensive. That's the sole point of it.
At some point in the "near" future it'll be energetically a net positive to snap a pix on the cam, analyze the eye pupils and accelerometers and recent eyeball movements, and if you're not looking at that part of the screen, its not going to waste energy displaying it. Could even see this happening with backlighting. Watching other people use their phones is likely to get very distracting for 3rd parties.
You missed an easy Windows comparison by focusing on Windows 7 and 8: Vista brought big graphics changes but ran poorly on a lot of existing hardware, similarly to iOS 7 on the iPhone 4.
> The same is true of Apple when they ship a new app that does more.
There are often better explanations. GMail is fast. All the time. No matter how much mail you have. That's because it runs from a local database and all server interaction is abstracted behind sync and local accesses to data. Google Play is slow because it is a REST app with a native app veneer. It's harder to make an app like Google Play use sync'ed databases. So sometimes it does puzzling things like taking tens of seconds to go to the next track even when you are shuffling only your downloaded music.
Fast apps can "do more" and remain fast. Slow apps are slow for architectural and algorithmic reasons rather than because they "do more."
> GMail is fast. All the time. No matter how much mail you have. That's because it runs from a local database and all server interaction is abstracted behind sync and local accesses to data.
I don't know about Android, but on iOS, the Gmail app is much slower to launch and navigate than the stock mail app. I think that, on launch, the Gmail app makes some requests before showing you any of even your locally cached data, which always struck me as poor design.
That's also how the GMail website still acts, which confuses me as well. Why can Facebook load a complex web app instantly while GMail still shows a loading bar on launch?
actually, when you load facebook, it shows an interstitial which looks like a page of posts, but is really just a bunch of blocks carefully arranged, and it has nothing to do with the final load page.
Huh. I checked back a few times. It looks like facebook loads an interstitial with boxes, that actually corresponds to the layout of the page, but not the content. This includes sentence layout (the boxes cover the area that gets replaced with text, similar for images). Now I'm curious what's going on.
When a new iphone comes out, owners of old iphones could feel that their phones are inadequate, and are searching for "is the iphone [n-1] slow?".
Also, with each iphone generally comes a new ios version, which is more or less written for the specs of the newest phone. That being said, this post is being typed from an iphone 4s I've owned for two years. I've never had any complaints about performance. Never noticed a slowdown.
Re: samsung: while they don't make the operating system, they do bundle apps and sometimes even their own filesystem (Galaxy S1). This particular filesystem made android slow.
You can argue that this means Samsung does indeed make their own operating system.
The term OS, as it is used by media and most people today, is the software that is installed on the device before you install any additional apps to it. E.g the "App Store" is considered a part of iOS, the Internet Explorer as part of Windows, and the Homescreen and Dock of Max OSX is considered a part of OS X.
Since a Galaxy and a HTC One ship with different Applications, a different user interface and other key software components (you mentioned the filesystem on the Galaxy S), Android is less an operating system but but the cross platform foundation that most operating systems are based on today.
> The same is true of Apple when they ship a new app that does more.
Presumably youre implying here that people who complain about their iphone being slow are the people who upgraded their OS while keeping their old hardware?
The spike for the iPhone 3G suggests that the effect is as much or more in people's heads [1] than due to new software. The 3G had all the same parts that affect how fast a phone works: the CPU, GPU, RAM etc, were all the same models as in the first iPhone. Even if iOS 2 were designed more for the 3G than the original iPhone, it couldn't have required significantly more resources because no more became available!
The spikes for iOS 4 and iOS 7, which are relatively larger than for the other releases, may be justified. iOS 4 brought limited multitasking and more significant OS-level changes than iOS 3, and iOS 7 brought the visual refresh which is more CPU- and GPU-intensive.
[1] Or due to the upgrade process itself, rather than changes in newer OSes, slowing down phones
>Even if iOS 2 were designed more for the 3G than the original iPhone, it couldn't have required significantly more resources because no more became available!
That wouldn't matter if Apple were simply adding code that tested the hardware and manually added slowdowns on the older model. Your observation could be seen as lending credence to the conspiracy theory.
Isn't that the problem with conspiracy theories? There is no disconfirming evidence ever. All evidence is either neutral or somehow confirms the conspiracy theory.
New releases of iOS usually go hand-in-hand with a new device. They design the OS to be optimal for the latest and greatest hardware, not so much the previous generations. As much as I like a good tinfoil hat theory, I'd say this is just a case of Apple adding new features (eye candy, mostly) and not worrying so much about the performance on older devices.
It's also to do with Spotlight re-indexing the contents of the phone. Every email, every contact, every web page still in history, etc.
There's some "optimisation" happening too, as bits are shuffled around so the most commonly used parts are fastest to access.
This applies for OSX as well as iOS. After upgrading, just wait an hour or two and everything will settle down. The impatient will simply have to cope with slower speed since the device is still sorting itself out.
Agreed - after a restore usually an iPhone will get unusually warm, I assume because it's doing this indexing and such. I don't recall if it happens after an in-place upgrade, though.
I wonder if Apple will ever add some sort of indication when the indexing, etc. is happening, so users don't become frustrated and/or wonder why their phone is hot all of a sudden.
I wonder if this is also due to new major iOS versions being released at the same time as new iPhones, and these updates add more intensive features (e.g. apps open in the background) and are designed more with the most modern devices in mind.
I know that my 2011 iPad 2 started feeling much slower after installing iOS 7. (Though luckily I finally tried the "Reset all settings" trick and it made a big difference, which points to a bug.)
> I wonder if this is also due to new major iOS versions being released at the same time as new iPhones, and these updates add more intensive features (e.g. apps open in the background) and are designed more with the most modern devices in mind.
So you wonder if it's the same theory posited in the article itself..?
The theory in the article is that Apple is intentionally making the older devices slower.
The theory you are responding to is that Apple is intentionally adding new features to iOS, which incidentally makes older devices slower because they're trying to do more work.
Is it a "conspiracy to make your device slower" when Apple adds new features to later versions of iOS? They're not intentionally adding in "slow down a bit here and encourage the user to upgrade to a new phone" code. So there's no conspiracy.
One of the interesting contradictions in software is that, logically, things are actually supposed to get faster as new versions come out, since there should have been both more time and more data to use for optimization. However, in practice, new versions of software are usually slower than their predecessors, presumably because of the addition of new features. It'd be interesting to see a product developed with a "no slowdowns allowed" policy.
Pretty sure Snow Leopard was a release aimed at doing just that, rewriting a lot of the core system libraries for x86_64 and also clearing house on a lot of old cruft. No new real features were delivered with the release.
Then Lion came along and slowed things down to a crawl. Definitely the slowest OSX release I've had on my old macbook (circa '08). I eventually downgraded back to snow leopard until mountain lion came out, which was noticeably faster than Lion, but still not quite as smooth as snow leopard. But now Mavericks is back to being pretty good.
All-in-all, still not a very straightforward 'improvement' path performance-wise -- at least not from the end consumer's point of view.
WebKit has long been known for having a good set of automated performance tests and has a "zero-tolerance policy for performance regressions": http://www.webkit.org/projects/performance/
The websites themselves are what get slower and use more memory over time, not the browser. Try using Tumblr on your phone someday and check the data usage after - it doesn't compress those multi-MB gifs for mobile at all.
Some large websites (I've heard this about Google Search, specifically) have ironclad latency limits. If a feature can't be added without increasing latency above XXms, then it's not shipped.
The article specifically notes that the spikes happened when the new iPhones are released, not just announced. So, new Iphone envy is less of a reasonable explanation..than the release date usually coinciding with new versions of iOS being deployed.
I have a much simpler explanation for this data. More people search about all aspects of their iPhone ahead of Apple releases -- whether it is slow, where to buy a new one, how to sell their old one, etc. This hypothesis is supported by the same data source (Google Trends data).
This has more to do with Apple's marketing strategy (hush-hush until a big media launch day) vs Samsung's (slow drip releases across various phone service providers). And also Apple's brand equity -- that it has more loyal followers eagerly anticipating new releases, vs Android users who tend to be commodity/pragmatic buyers who basically only research phones when their contracts are expiring.
This might be one of the best examples of the quote 'statistics don't lie, but liars use statistics." First, paint a picture of some insane scenario where Apple would be purposely nerfing phone speeds (pay special attention to how much of a weasel word 'speed' is in this context), then, (gasp!) it turns out that a completely unscientific data aggregation found the author's scenario to be true!
The big problem, though, is that the average consumer doesn't decouple processing speed and network speed.
Release-day consumers congesting the network with updates, app/music (re-)downloads, testing of new functionality, etc. seems to be the most likely reason; why would the number of people searching for 'iphone slow' sharply drop after a day or two? Are we to believe that 95% of iPhone users upgrade on the first day?
It bothers me that the term "Big-Data" is thrown around casually for basically using Google Trends canned.
The hokey conspiracy theories are needless when there are multiple sites that specifically benchmark the iPhone when new releases of iOS come out - 10 seconds with Google [1].
So the author ( & professor! no less) should have gone with his first instinct that Apple is the largest company in the world and would be foolish to deliberately wreck an older product...this is an embarrassment of an article needlessly tagged with "Big-Data" and expounding a bunch of non-sensical conspiracy theories.
It is very unfortunate that the author did not get benchmark data, because the benchmark you have linked to supports this alternative explanation, which is in the later part of the article:
>This data has an even more benign explanation. Every major iPhone release coincides with a major new operating system release. Though Apple would not comment on the matter, one could speculate — and many have — that a new operating system, optimized for new phones, would slow down older phones.
The benchmark data shows an older phone performing much slower with an updated OS.
Whether this supports the conspiracy theory implied in the first part of the article depends on Apple's intent, as the article says.
Hmm, a google trends search for "iPhone covers" shows similar spikes. Clearly people are embarassed about their slow phones and want to hide them. And not that searches for iPhone-related terms spike when, you know, people have just bought new phones.
"Apple does not force iOS updates on its users. However, it does notify you as soon as an update is available"
Alas, that's something of a lie! OK, something of a falsehood for polite society. In fact, Apple downloads its iOS installer update to your device--and that can take up over 3gb of space! You have no choice in that.
And, if you are not careful, you can inadvertently click a button in iTunes that updates the OS.
No, the reality is that Apple is forcing through its updates; making it all too easy for people to reflexively click yes on the update; and eating up precious space on 16gb devices (well, 14 gb).
Does it actually matter what the motive is? If Apple is making older devices slower as a side effect of something else, (and not informing the users who choose to update) is that really any better? And how can you ever prove they aren't aware of the effect and choose to do it anyways to sell more new devices?
Seems like a lot of this speculation could be replaced by a few benchmarks to see whether the phone actually got slower... but I guess that would take more time than looking at a Google trends graph.
Very interesting to see some data on this. I also have this feeling that at every iOS release the older devices get terribly slower. And Apple seems to make no efforts to change this, on the contrary, iOS7+ is filled with transparencies, bouncing effects and other gimmicks that add no utility but demand hardware, and now swift generates code so slow that new devices will have to be bought to be able to use the apps made with it. Just perfect for Apple.
>on the contrary, iOS7+ is filled with transparencies, bouncing effects and other gimmicks that add no utility but demand hardware
No, Apple doesn't want their devices to feel slow. For example, my iPad 3 doesn't gaussian blur UI backgrounds. If Apple was nefarious as you're suggesting, they would force my iPad to render the expensive blurs.
I definitely believe this is part of the business strategy for OS vendors. Microsoft actually successfully did this with new versions of Windows too, up until Vista (their incentives were less direct since they didn't produce the hardware directly, but they still got a royalty for every new PC purchased with a Windows license, which was all of them). Every new version of Windows would require a comparatively beefy machine to successfully run. Vista wasn't all that different in its requirements differential, but the market had started demanding devices that traded down computing power for more convenient form factor (the first generation of netbooks were big when Vista hit the streets) and expecting these devices to run Vista just as fast as they had run XP. Microsoft's failure to anticipate this shift in consumer behavior is mostly responsible for Vista's bad reputation.
Apple can continue to behave this way specifically because only Apple makes hardware that runs Apple software. If they opened up to third parties and someone created a computationally weaker but aesthetically preferable device that ran their software they'd be in trouble.
I like how you read an article whose point is that there's no sense in making up conspiracy theories, and infer from it support for a conspiracy theory.
Swift should theoretically deliver faster apps with less message passing overall and other big improvements.
At the end of the day the latest phones need to keep up with competitors and added features and resimplified experiences that are more costly to implement cannot be avoided. The only way to really avoid it is disallowing older phones to upgrade - which is far less desirable IMO.
I think this is one of the primary reasons I like Linux. Current versions of Linux run acceptably well on old hardware whereas current versions of Windows just chug along on old hardware.
Computerphile did a piece on this (intalling Linux on a very old PC)
XP to Ubuntu with an 8yr old Hacktop - Computerphile
> People suddenly feel that their phone is slowing down. It doesn’t show that our iPhones actually became slower.
This could be tested. Developers could collect data on the actual speed of their apps, and if there is anything shady, we will have hard evidence on our side.
I am sure there are enough un-updated phones in our drawers to test out what the speed used to be back then.
I'll propose a different conspiracy theory: perhaps it's the other way around - new versions of iOS are specifically timed to coincide with the release of new hardware in other to make people feel like their phone is now slower and get them to upgrade...
I don't think it is the old phone which is slow and which causes people to search. It's the new phone which is slow. People get disappointed, and search. This explains why spike does not appear during announcements.
The thing that is the most striking about the graph of those searches is that it is time locked to releases and NOT to announcements. Seems strange. If people are upgrading the OS then there is a very simple explanation.
Samsung is just a generic hardware vendor for Android PCs, like Dell. They don't create operating systems or apps.
It probably also has a psychological component. Just seeing a new generation gaming PC makes mine feel slow.