19 Comments

Summary:

An analyst predicts Apple will re-build iOS to work on faster processors this year. Based on the age of iOS and personnel changes made last year, this isn’t a bad bet.

SVP of iOS Software Scott Forstall at WWDC on Monday.

There are a lot of predictions being floated about what Apple’s going to do in 2013: an iWatch? An iTV? How about a plastic iPhone and an Apple TV SDK? A note from an analyst from Jefferies published Wednesday contains many of these predictions, but one idea about a move the company might make this year sounds likely: that Apple will make big changes to iOS that will allow the software to work better on mobile devices with much more powerful processors.

In a note to investors today, Peter Misek of Jefferies Securities notes (via AppleInsider):

“We think Apple plans to re-architect iOS to utilize more cores and better compete with Samsung. Also, we believe the way iOS interoperates with iCloud, gestures controls, and advertising will be substantially upgraded.”

I’m not sure about those specific things, but I do think big changes are coming to iOS for a couple reasons.

While iOS has seen six new releases since its debut in 2007, there have been few major changes. The arrival of the App Store in 2008, and push notifications in 2009 were the last big adjustments in how the software works. Consistency is good for users, and it’s been working for Apple. But iOS was developed at a time when mobile processors were slower and smaller — and iOS was no doubt built with those parameters in mind. At some point, as the analyst notes, iOS is going to need to make back-end changes to keep up with mobile processors as they become faster and more capable.

Another sign that change is coming? Since between the first iOS release to the debut of iOS 6, the same person was guarding and guiding the development. But Scott Forstall is gone now. CEO Tim Cook dismissed the former head of iOS Software last fall for reasons assumed to be related to the Apple Maps debacle.

However, I think this move will have ramifications for iOS in general, not just Maps. Both the software element and the engineering part of iOS have a new overseers: Jony Ive is now in charge of the Human Interface group, and Craig Federighi was promoted to lead iOS Engineering along with OSX Engineering.

New managers by nature often want to come in and force change (even if sometimes for its own sake). But in this case, as the software ages, as early iOS design decisions incorporating lots of skeuomorphic elements fall out of favor, as competitors catch up, and hardware continues to improve, it’s probably more tempting than ever.

You’re subscribed! If you like, you can update your settings

  1. You should know better than to quote either an analyst or AppleInsider.

  2. activeinspiration Wednesday, February 13, 2013

    Maybe since they’re not a hardware company they will open up there OS to a multitude of phones.

    1. “Maybe since they’re not a hardware company they will open up there OS to a multitude of phones.” that’s about the funniest thing I heard today, if it isn’t sarcasm

      1. I hope that was sarcasm, though I often hear advice offered to Apple from an outside perspective which fails to understand Apple. If they wanted to be Dell or HP they would have changed years ago.

  3. iOS doesn’t have a problem handling multiple cores or fast processors. It’s true that the Jelly Bean version of Android includes a bunch of upgrades to take advantage of newer processors, but that is more the result of Android being originally formulated for feature phones and therefore relatively inadequate in the smartphone world. Now they have caught up, but that doesn’t mean iOS needs to do the same thing.

    In fact, I still think Android lags when it comes to supporting truly high-power processing, and that is because it relies upon Java (OK – Dalvik). Although Obj-C is rather long in the tooth when compared with more modern programming languages, the fact that it is a very thin layer on top of C means that developers can have their code run natively (we have even written a bit of assembler code in a couple of projects when we needed real performance). In order to get this level of control and efficiency on Android you need to code against the NDK, which means trouble when trying to support all of the different Android hardware out there. I believe this is part of the reason why iOS has done so well with game developers – you can really get down to bare metal (not to mention reusing a lot of your existing C++ code base).

    1. Yep! iOS is OSX is NextStep is BSD… The micro kernel can do whatever you want. It’s like saying that Linux needs to adapt to scale. Of course it does, but the architecture isn’t the issue.

      We’re not outgrowing these OSes for a while.

  4. Grand Central Dispatch – already in iOS for concurrency. I’m not sure what really needs to be rewritten.

    1. This. GCD already has full core scaling at an OS level that is relatively simple to implement.

      If we were to see big changes in iCloud functionality such as adding a system wide storage area in addition to the per-app ones and interapplication communication.

  5. I think the most prominent changes to iOS will be design ones in the next year. Be prepared for flatter, less literal design language to match the minimalist flair of iHardware.

  6. So less rounded square app icons?

  7. The analyst has a point, but he kind of misses something important. Witness Apple has a multi-core problem with Grand Central Dispatch not working all that well as first evidenced in Mac OS X Lion and continuing somewhat with Mountain Lion, as well as iOS. Case in point, when’s the last time anyone looked at their iOS crash logs (after syncing an iPhone, iPad or iPod Touch with their Mac)? Take a close look at those logs (and also the logs immediately after a Mac OS X or OS X app has crashed), and don’t be surprised to find a segmentation fault (SIGSEGV etc.) The proliferation of apps crashing due to segfaults has been linked to old code inherited from the period prior to multi-core (for example some of the code integrated from NeXTSTEP into early days of OS X was from a single core world). What Apple needs to do is not focus too much on cosmetic UI changes but get back to focusing on a stable system (Mac OS X 10.6 Snow Leopard had gone a long way toward stability then Apple broke it with the “Back to the Mac” concept bringing iOS concepts into Mac OS X 10.7 Lion). I believe Apple wants to forget about Lion because it was very awkward. Time to clean up the old single core cobwebs Apple and make both OS X and iOS stable in the multi-core world of today.

    1. I agree about GCD. There is support for multiple cores and that code is meant to work identical on iOS and Mac hardware without any change and still get the benefits concurrency. Where it appears to have trouble is deep inside the beast which Apple is likely working hard on fixing. I doubt this article meant any of that when it mentions major changes to iOS for multiple cores and does not even address GCD. I have experienced strange problems on my Mac and sometimes with iOS where it seems like GCD code is having trouble. I did not think to look at the crash logs for evidence but that would make sense. I also see problems with the filesystem. That needs to be fixed. My new MBP with Retina had a hard failure within the first few weeks and I am now very paranoid about it. It seemed to start having issues with Lion and did not get any better with Mountain Lion.

  8. When you got nothing better to do nor have a clue about technology just write an article like this…

  9. Assuming the changes would be announced around WWDC, doesn’t it seem a little soon to expect drastic changes this year? Forstall wasn’t let go too long ago.

Comments have been disabled for this post