Solipsism Gradient

Rainer Brockerhoff’s blog

Browsing Posts in Apple

WWDC 2020 opens next June 22nd and all indications are that the highest-impact announcement will be the Mac’s migration from Intel to the ARM architecture.

While CPU architecture migrations are infrequent — they happen every decade or so — Apple has a good track record of pulling them off successfully.

The first major migration was the move from Motorola 68K to PowerPC chips around 1994, followed by moving from the Classic Mac System 9 to Mac OS X around 2000. Relevant here was that for some time Mac OS X ran older applications in the “Classic Environment”: a compatibility sandbox that emulated the APIs of System 9 and the instruction set of the 68K.

This worked reasonably well as PowerPC CPUs were several times faster than the old 68K ones. It also introduced the concept of “fat binaries“; the same application file contained code for both old and new environments.

A better historical precedent is the move from PowerPC to Intel processors in 2006. This was more traumatic for developers, as PowerPCs were “big-endian” and Intel CPUs were “little-endian”. This meant that, except for strings, values stored in memory, files or transmitted over networks had a different byte sequence ordering. To have the same program source code work on both systems you could no longer assume it would just work, but had to bracket your instructions with macros or function calls that would do nothing on one platform but swap bytes around on the other.

If you’re not an oldtimer like myself you probably never had to think about this — every Mac, iPhone, iPad, Apple Watch or Apple TV use little-endian values, and I even had to dig into documentation to make sure of it. ARM CPUs can be run in big-endian mode by setting a special bit at boot time but this is not the default, and no Apple device uses that mode.

Now, this meant that in 2006 developers could not just migrate their apps to Intel by recompiling; we had to look through every line to either check that it was endian-neutral, and if it wasn’t, those special macros had to be used. For people who had very CPU-specifically optimized code — perhaps even in (shudder) assembly language — separate code sections were necessary.

Having done all this, you recompiled your app twice; once for PowerPC, once for Intel; and the magic of fat binaries allowed you to ship it all in one app. Later on, some apps even needed 3 or 4 different code sections, depending not only on endianness but also on whether they would run on a 32- or 64-bit CPU!

Another — today mostly forgotten — aspect was that Apple prepared for the Intel migration by gradually modernizing and building their developer toolchain in-house. LLVM, Clang, LLDB etc. allowed Apple to ensure that, for whatever CPU they wanted to support, compilers were ready beforehand and could be optimized continuously later on, without depending on outsiders.

Still, in 2006 Apple had to ship special hardware, “Developer Transition Kits”, to select developers for testing. For software that couldn’t be converted to the new architecture, Apple introduced a limited compatibility box: Rosetta. If I recall correctly, it did on-the-fly translation of PowerPC code into Intel code, which was then cached. Because of its limitations it didn’t work for many larger applications and was soon phased out.

Moving in parallel to the PowerPC to Intel migration was a slower-motion shift in operating system APIs. Most notably, this involves Carbon and Cocoa.

Carbon was a C-based API introduced in 2000 to ease migration from Classic System 9 to Mac OS X. Cocoa, introduced around the same time, was an Objective-C based API for modern object-oriented programming, itself an evolution of NeXT’s OpenStep system. Underneath both APIs, in the now well-known layer model, was Core Foundation, which could be used from both types of apps; and some apps (like my own) could mix calls to both APIs with some care.

Not too long after the Intel migration, Apple announced that 64-bit was the future, and that Carbon would not be migrated to that environment. This process was stretched over several years and involved redefining what APIs were really considered “Carbon”; some, like the File Manager, were “de-carbonized” and lived on until macOS 10.5 (Catalina) came out.

Cocoa, on the other hand, continues to be used everywhere in macOS. The Finder, the Dock, Xcode, and Safari are all Cocoa apps. Even when Swift came out a few years ago most of it was built on top of Cocoa and Objective-C objects; the notable exception is the Swift toolchain itself.

So, after all this, here we’re looking at Yet Another Hardware Migration for Macs. Let’s look at the implications.

Economically, it makes sense for Apple, as many others have already commented. They’ll no longer be bound to a foreign evolution roadmap on which they have little influence. They have extensive experience in producing high-performance, low-power CPUs for their mobile devices, and the latest versions already outperform Intel in some situations.

Technically, it’s a huge win. Switching to ARM64 — and not just the standard ARMv8.x architecture licensed from ARM, but with their own, extensive modifications — will allow them to have unified GPUs, Neural Engines, memory controllers and so forth on all their line, with more uniform device drivers and low-level programming.

For 99% of developers, I think nothing will change. The new chips are little-endian also, so a simple recompile will have Xcode produce a fat binary for the new Macs which should run outright. Of course, if you have assembly language sections in your program and/or write kernel extensions/device drivers, time to learn a new architecture…

Snags will come for people who dislike, or can’t use, Xcode. Some have to use Intel’s compilers, for instance; I know too little about such cases to have an informed opinion, sorry.

Some pundits seem to expect a sudden concurrent change in macOS; something like Objective-C and/or Cocoa being obsoleted in favor of Swift and SwiftUI. Or even the Mac going away entirely, some sort of huge desktop iPad taking its place. In my view this won’t happen. For one, what would developers or even most Apple engineers use for development?

A big question is: will Apple be able to provide an Intel compatibility box on the ARM Macs? Certainly Boot Camp will not be available. Running a virtualizer like VMware Fusion or Parallels seems almost as difficult, unless the new CPUs have some sort of hardware assist to decode x86-64 instructions. This may not be as outlandish as it sounds; current Intel/AMD processors already break x86 CISC instructions into RISC micro-operations which are then cached and executed by the “inner” CPU. This is a gross oversimplification but in theory nothing — except silicon space — bars Apple from breaking x86 instructions into ARM instructions.

A Rosetta-like box seems more feasible for running individual Intel applications, but who needs that? Game users? Performance will be limited. Most virtualizer app users want the complete OS running and with native speed. Linux/BSD might be available soon; perhaps Windows for ARM.

But what about Catalyst, some of you may ask? Here I can only shrug. In its present form it certainly is not an important future technology for macOS. While simple apps can be done with it — perhaps purely for the benefit of developers unfamiliar with AppKit — can you envision a Catalyst Finder? SwiftUI is still very new and primitive, and will continue to be layered on top of AppKit/UIKit for some time. They may merge in the future, or be renamed gradually like Carbon was, but that’s a long time out.

Finally, hardware. I don’t think the existing A13 SoCs would be applicable to any Mac, though. Some version of the Mac mini would be the obvious candidate to be the first to get the all-new CPU. It would then percolate up through the laptop line and the iMac. In these cases, reduced power usage would be a bonus — even for the iMac, it would mean a smaller power supply, less heat and a thinner enclosure.

The Mac Pro should be the last Intel redoubt. Multiple CPUs, OEM graphic cards, generic PCIe cards — Apple will have to address a huge range of problems there and this will take years.

Enough handwaving for now; the usual disclaimers apply and I’m really looking forward to the keynotes next week.

Update:
— corrected date for the 68K->PowerPC migration. Thanks to Chris Adamson for catching the error.
— fixed some awkward language about virtualization. Thanks to Maurício Sadicoff.

LEGACY NOTICE

No comments

Today I have removed my remaining apps from the Mac App Store and posted official “Legacy” status notices on the product pages. My software development activities are (now, officially and indefinitely) on hiatus.

I know this notice has been long overdue, as I’ve been unable to update anything for more than 2 years. My apologies; I’ll try to post more about this here in the following weeks.

Boom: the Return

No comments

A few years ago I wrote a series of posts about Apple’s then-new Lightning connector for iOS devices:

No doubt you’re noticing a trend there… 🙂

Anyway, the recently-released iPad Pro seems to have the much-awaited USB3 capability on its Lightning connector. It does ship with a Lightning-to-USB2 cable, though, and USB3 capability isn’t mentioned in the tech specs.

The main objection to this actually happening is that Lightning, with its 8 pins, doesn’t have enough pins to support the standard USB 3 specification. This is, again, the old assumption that Lightning cables are “just… wires leading from one end to the other”.

To restate what I posted previously, if you actually look at the USB3 pinout, there are the two differential pairs which Lightning already has, and one additional pair for USB2 compatibility. So a legacy wire-to-wire USB3 cable would need 9 pins — but, remember, Lightning connectors don’t work that way!

In other words, if you plug in an old Lightning-to-USB2 cable into an iOS device, the cable itself already has to convert the two differential pairs to USB2’s single pair. So, no need to have the extra legacy pair on the Lightning connector itself — a future Lightning-to-USB3 cable will generate that as well, and use the two high-speed pairs when plugged into a USB3 peripheral. The current pinout is, therefore, quite sufficient.

A Tale of Two Certs

6 comments

I’m keeping this post updated as details develop…

About ten days ago, something strange happened on my Mac: I was debugging the next version of my RB App Checker Lite app and suddenly I saw the dreaded dialog box:Damaged

Completely abnormal, especially as I was debugging using the Developer ID version (not the Mac App Store version!) from inside Xcode. When I opened Terminal, the same dialog; when I opened Safari, same thing! No new process was allowed to run. Of course I had to reboot to be able to do anything, everything worked fine afterwards, and I couldn’t reproduce the problem, so…

OK, a couple of days ago I concluded all was ready and I uploaded my app for review. A few hours after I announced so on Twitter, the reports began appear: the sky is falling! Major Mac App Store meltdown, everybody was getting the “damaged” dialog, Apple’s certificates were the culprit. I started testing my local apps from the MAS and, sure enough, the MAS leaf cert had expired; no problems, some of them asked anew for the AppleID password, some didn’t. RB App Checker Lite showed the expiration but no other problems, but I pulled it from review just in case.

Two days of confusion and frantic coding later, I had submitted (and pulled!) 4 more builds until I was reasonably sure that everything was working correctly. Thanks to several fellow developers on Twitter, the upcoming version seems to show everything correctly; it turned out that my receipt checks were somewhat obsolete. I usually publish the direct download version only after the MAS version has passed review, but decided to release version 1.1.4, build 351 immediately: you can get it here. It has a long list of improvements and fixes.

Meanwhile, the consensus is that rebooting and re-entering the AppleID and passwords (or even deleting and reinstalling) the affected apps solves 99% of the problems.

There are actually several different unfortunate problems here. First, the “damaged” dialog seems to be caused by some sort of cache or memory corruption in the system processes that coordinate to implement GateKeeper and the app store updates; some reports say killing the “storeagentd” process solves this problem without rebooting. (My system doesn’t seem to run this, FWIW.) What not everyone knows is that this dialog appears before the app it allowed to run; that is, it’s not affected by any checking done inside the app itself!

Second, asking for a new AppleID password. This is caused by the app itself checking the store receipt; something strongly recommended by Apple, since otherwise, it’s easy to copy a downloaded app to another computer and having it run there; I remember some early games not doing this and being widely pirated.

When an app is downloaded from the MAS, a proper receipt for that AppleID and that computer is already inside. A missing or corrupted receipt is the only normal circumstance in which the “damaged” dialog should appear. But if you copy the app to another computer, this will be noticed by the app itself.

Once a MAS app starts up, the first thing it should do is to check the receipt. It’s a complex process and not everybody implements it the same way. At first, checking the receipt’s cert chain would cause the receipt to be rejected in the case of expiration; the app exits with a special numeric code (exit 173) and this code signals the system to put up the dialog asking to confirm the purchaser’s AppleID and password. This, in turn, will cause a new receipt to be downloaded, and the app can now run with no problems. Update: reports indicate that, in at least some cases, the system doesn’t respond properly to exit 173.

A few years ago receipts began to include a new field containing the receipt’s creation date, and developers now had to check the certs against that date (and not against the current date), therefore obviating the need to reenter the password. Unfortunately this was not widely divulged, and Apple’s own sample code hasn’t yet been updated accordingly; I confess to not seeing this myself!

As is usual in disasters, several things have to go wrong at the same time: some bug corrupts a critical system cache, certificates expire normally, some apps incorrectly test for expiration, receipts are corrupted or the AppleID validation servers become slow or unreachable (because of the huge number of simultaneous requests), and… boom.

Many articles, unfortunately, published factual errors or wrong assumptions.Let’s try to counter a few:

  • Apple “allowed” their Mac App Store certificate to expire. Wrong on several levels. First, there’s not one but 5 (!) certificates involved in any app from the store: Apple’s root certificate: and 4 others: two intermediate and two leaf certificates.
    The way these certs work is by so-called certificate chains; every cert vouches for the lower-level ones. At the top is Apple’s Root certificate, which is one of a hundred or so in the System Keychain. There are two different certificate chains in every MAS app; the first is used in the code signature:and the second is used to sign the store receipt:Note the expired certificate there? This is a leaf certificate. These, usually, have a short life — one or two years — and the intermediate certificates usually last a little longer.
    So, when a cert expires, is that a serious problem? No – unless it is the root cert, which is why they all expire somewhen in the 2030s — hopefully, by that time, they’ll have figured out something better, Apple will have updated the cert via Software Update, or the horse will have learned to sing.
    The root cert can be updated via Software Update because it’s stored in System Keychain — but it’s impractical to push cert updates to each and every signed app, bundle or library; there are many thousands of them! So an expired cert in the code signature doesn’t affect the app at all. What’s important is that the certs were valid when the app was signed. When and if you get a new version of the app, all certs will probably be new ones. So there’s no “allowing” a leaf cert to expire — they do so naturally.
  • Apple “pushed” a new certificate that expires in 2035. This is probably just looking in the wrong place — not knowing which certificate had expired, someone glanced at the root certificate and noticed the “new” 2035 date. Nothing new to see, of course; that cert was created in 2006! Even more confusingly, someone else deduced from that that Apple let their original root cert expire; also wrong.
  • The system hasn’t been updated to check SHA2 (256) certificates. Wrong; it’s true that older systems used a version of the OpenSSL library that understood only SHA1 (128) certs, but that actually means 10.5 or so. Newer systems understand SHA2, and in any event, since the MAS went up, Apple has always recommended developers to not use the system’s OpenSSL library (I think it’s not even included anymore), so only very old apps would be affected by that.
    Update: Glenn Fleishman has informed me about the SSL situation: there’s the new 1.0.x library branch and the older 0.9.x branch. Both apparently got SHA2 support in 2010, when 1.0.0 and 0.9.8o came out, but some developers seem to have kept older versions, no doubt for valid reasons; space precludes, etc.
  • Apple is blaming developers. Apparently this can be traced to a single report of misinformation from an anonymous Apple Support person. As I write this, Apple hasn’t yet said anything; I doubt they’ll say anything over the weekend.
  • This is a serious security/cryptography failure.  Nope. This confusion arises from the fact that digital certificates (and libraries like OpenSSL)  are used for both secure, encrypted communications and for app/receipt signing. In the latter case, an expired cert doesn’t expose any information or makes the system or apps easier to hack.
  • Developers are better off not doing any, or little, receipt checking. Not really. True, apps which don’t do full receipt checking might have not been affected in this single instance, but under usual circumstances they’re more vulnerable to hacking or piracy.
  • Apple’s store/system infrastructure is brittle and can’t be trusted. True, it’s a very complex system that depends on many twisted little interlocking parts to work properly. And, as we’ve seen, this particular instance of failure is as self-amplifying as electrical grid failures _ once it starts, the demands on the working parts grow so huge that those fail, too. In Apple’s defense, it’s very hard to test for or simulate. Let’s hope that all involved have learned something from this incident; I certainly have learned a lot.

Update: forgot to comment on this particular post:

But when I tried to convince my Mac to run this app as an unsigned app, I encountered what is extremely likely to be the store DRM: I initially got the “your app was bought on another machine” message, so I tried deleting the receipt, but then I got the dreaded “app damaged” message, at which point I removed the signature.

…the only way I can see is to create a new root CA which I install on the machine as a trusted root, and redo the signing chain, and even that might not work if the DRM is somehow tied to the signature chain.

While I can understand the frustration implicit in not being able to run purchased apps “forever”, I think this is a fundamentally wrong approach. Let’s educate developers to check receipts properly, as I mentioned above. Figuring out a way to run store apps (or even developer-ID purchased apps) without “DRM” means that anyone else can use the same method to install pirated copies; we wouldn’t be able to trust users anymore.

Much worse, re-signing someone else’s app and expecting it to run is an even greater violation of trust. The days when you could hack someone’s app with ResEdit and having fun making it look different, or do unexpected things, are long gone. I implement very strict checks that my complete app bundle has not been altered in any way and that it’s running with my original signature, otherwise any user could freely alter files, hack the code, change graphic resources or even — and such cases have happened! — repost the app somewhere else as being their own. No, flawed as the current approach may be in implementation, I see no better alternative.

Update: reports are in from some helpful fellow developers, confirming my suspicions of cache corruption — RB App Checker Lite says the app bundle and receipt contents are OK, yet the apps will not run. I use the same APIs (hopefully) that the system processes use — but those APIs can take a long time to run, so the results are cached somewhere.

Update: Apple sent email to all developers:

In anticipation of the expiration of the old Mac App Store certificate, we issued a new certificate in September.

As I said — no “let[ting] certificates expire”. They all do.

We are addressing this caching issue in an upcoming OS X update.

Confirms this is a caching issue, as I suspected.

…some apps are running receipt validation code using very old versions of OpenSSL that don’t support SHA-2. We addressed this by replacing the new SHA-2 certificate with a new SHA-1 certificate last Thursday night.

I’m a little surprised that the number of apps using“very old versions” justifies going back to SHA1; but, OK.

Please ensure your code adheres to the Receipt Validation Programming Guide and check that all receipt validation issues are resolved.

Good, but:

  • the link goes to the page detailing the online receipt validation. Very few apps use that IMHO — you have to be online every time the app runs, you have to have a reasonably fast connection, and app launch will be significantly slower. Linking to this page (Validating Receipts Locally) would’ve been better;
  • it would have been more helpful to call out specifically the certificate expiration check and update the sample code to properly use the receipt creation date.

Update: I’ve now submitted rdar:///23611335 — a bug report to call attention to this documentation problem.

Update: Fixed the “Validating Receipts Locally” link, which was also pointing to the wrong page. Sorry. Also, here’s one way to do the correct date checking (copied from Matt Steven’s code):

X509_STORE *store;
// set up the store
X509_VERIFY_PARAM *param = X509_VERIFY_PARAM_new();
X509_VERIFY_PARAM_set_time(param, time_from_receipt); // option 1: verify using a specific time
X509_STORE_set1_param(store, param);
X509_VERIFY_PARAM_free(param);
// call PKCS7_verify() using configured store

As you may have seen elsewhere, Apple has just approved publication of RB App Checker Lite 1.1.3 (build 320) on the Mac App Store, and I’ve simultaneously published the Developer ID version (build 321) and updated the product page.

While the main focus of this version is to do some additional checks, fix bugs and display some information that users have requested, the XcodeGhost story broke just after I uploaded the first binary for review, and I had to remove it in order to see what RB App Checker Lite could do to detect this breach of security.

Assuming you’ve read the excellent summaries linked to in the previous paragraph, you may be interested in what I found inside the “ghosted” version of Xcode. Briefly, the hack alters the script Xcode uses to link an app’s binary; it also inserts look-alike versions of the CoreServices framework into the iOS, iPhone Simulator and OS X SDKs inside Xcode.

All this of course breaks Xcode’s code signature and, under normal circumstances, running such a hacked version would — after the customary delay for checking, 2 to 5 minutes — be detected by GateKeeper and it would advise the developer that “‘Xcode.app’ will damage your computer. You should move it to the Trash.” And the previous version of RB App Checker Lite would advise that “…requirements and resources didn’t pass static validation” and point at the changed file.

You’d think that that would take care of the matter, but it turns out that the affected developers turned GateKeeper off entirely, no doubt to get rid of the several minutes delay. After that, versions of their apps uploaded to the App Store would have been linked with a static library containing categories on Cocoa classes such as UIApplication, UIWindow and so forth; this static library having been hidden inside the added frameworks.

Needless to say, the new version of RB App Checker Lite also detects the added frameworks and warns: “3 frameworks are suspect: they use system names but are NOT signed by Apple!”.

This is both good and bad news. The good news is that this specific version of XcodeGhost — or any similar hack that hides code inside bogus frameworks looking like Apple’s frameworks — can be detected. The bad news is that this specific tactic depends on passing a casual visual inspection of the SDKs inside Xcode; in other words, the names and file paths used look reasonable and mostly duplicate Apple’s names and conventions.

This works because Xcode is a huge application; it contains nearly 5 thousand auxiliary executables. The latest Xcode beta has several SDKs for each of the 7 platforms it supports, and each SDK has an included instance of all system frameworks, both public and private, for that particular combination. Unfortunately, not all these frameworks are currently signed by Apple — only 2/3 of them are, and not in a consistent manner. (In all fairness, the percentage has been creeping up a little with each release.)

Therefore, unless you check the entire app contents with GateKeeper, RB App Checker Lite (or even the codesign command-line utility), it will be humanly impossible to pick out visually — by inspection in the Finder — if anything has been changed inside Xcode. So keep GateKeeper turned on! One suggestion Apple should implement is running GateKeeper tests for Apple-signed software even if GateKeeper has been deliberately disabled.

So, what to do about “infected” apps? Unfortunately the news is not good there. (By the way, I’m surprised that no infected apps were — as yet — found on the Mac App Store.) As I said, infected apps contain linked-in categories on Cocoa classes, using plausible English method names. Writing such categories is perfectly legal and even plausible — I’ve done so myself. Having code inside these categories do things that are allowed by the app’s entitlements, such as sending/receiving data over the net, is also perfectly legal and plausible. There seem to be some utilities out already that purport detecting such code, but I suppose they’d turn up a lot of false positives unless they check for these specific combinations of symbols — not very future-proof.

By the same token, Apple can’t really do these tests comprehensively when an app is uploaded to the store. They can and do check for private or “suspect” APIs being called, but as far as I can see the present XcodeGhost doesn’t use anything like that.

Coming back to RB App Checker Lite: it currently does NOT look inside executable code at all. Should it do so? I’m reluctant to implement that; it’s not clear what exactly to look for, regarding hacks like XcodeGhost, and it would mean that checking Xcode and similar huge apps would take tens of minutes or even more. I’m open to suggestions, however… comment here or email me!

Watch Update

No comments

Tomorrow (we suppose) the Apple Watch will be out. For months, there’s been lots of interesting documentation on Apple’s site — but it’s all about WatchKit, the framework used on the iPhone side to run “Watch” apps. Almost nothing about the Watch itself. I think most of my previous speculations were confirmed: specifically, the part about the Watch mostly being a remote display for the iPhone:

Perhaps… just a sequence of drawing orders? The important part is that there’ll be a single process on the Watch for doing the UI, and all the application-specific parts can be offloaded to the iPhone.

So, for now, the application logic will all be on the iPhone side — where the actual WatchKit part runs — and “assets”, meaning storyboards, xib files, and PNGs with pre-rendered icons, buttons and so forth, are downloaded to the Watch and displayed as needed. My back-of-the-napkins calculations about battery life (around 15 hours) still seem valid: Tim Cook said that you’d have to charge the Watch every night. I also said:

Watch OS … will not be a stripped-down iOS; maybe even not a Darwin derivative. It will be a highly optimized embedded system that runs as few processes as possible. It will be very robust because it will be able to do only a fixed set of functions.

Of course, this clashes with everybody else’s assumption that of course the Watch will be running iOS. Apple continues to be very careful about this: the OS that actually runs on the Watch is named nowhere that I could find. Likewise no hardware specs beyond the two screen’s pixel sizes were revealed. Details about the OS may not be revealed until next year, when developer apps supposedly may run on the device itself. It might make make sense for Apple to repurpose, say, the OS running on the smaller no-app iPods.

Beyond speculations about functionality, rumors have concentrated on price and updatability. I’m not competent to speculate about prices, but John Gruber’s final thoughts on the issue seem very reasonable.

Opinions are split on updatability, since few of Apple’s products can be upgraded, and none can have their hardware updated to a next generation. Then again, here’s a completely new type of product, smaller and (in some versions) more expensive than any other; it’s also, perhaps, the most personal Apple product ever. If you get an expensive Watch, say, as a graduation present — with an engraving, perhaps — you’ll be very reluctant to dispose of it and get a new one in a few years, even if the new version does much more.

At absolute minimum, the battery will have to be replaceable, and in my opinion, the entire Watch module (probably including the battery, probably excluding the display) will be upgradeable for a fee once a better version comes out — maybe not forever, but for at least 2 or 3 generations. We’ll see.

Apple’s (pre-)announcement of the Apple Watch left the tech world in the usual disarray. Is it an expensive knock-off of Android watches (people tell me there is such a thing!)? Is it an attack on the high-end Swiss watch market? Is it an attack on the low-end Japanese watch market? Is it an even more transparent lock-in attempt on soi-disant “Apple fanbois”? I’d answer “no” to all those questions, but right now I’m more interested in the hardware and software technology of the watch.

Notice that the above link doesn’t mention iOS anywhere, but this other link has the magic word: WatchKit. Quote: “WatchKit Apps. Soon your favorite apps will feature controls and interactions unique to Apple Watch, enabling you to enjoy them in dynamic new ways.

Speculations about WatchKit since then usually have mentioned one or two assumptions:

  1. WatchKit will be written in/accessible only from Swift;
  2. WatchKit apps will run under iOS on the Apple Watch.

The first is, of course, wishful thinking from developers investing in the new Swift language. The second is, in my opinion, completely unwarranted and I’ll try to explain why.

This post is the most plausible so far: “WatchKit apps will ship as embedded binaries in iPhone apps, using the same basic principals [sic] as iOS 8 extensions. There will be some mechanism for the watch paired to an iPhone to detect and automatically install these ‘apps’ based on what is available on the paired iPhone. Delete the container app from the iPhone, it disappears from the watch. Xcode will have a template to add a WatchKit app to an iPhone app project.

Let’s back off WatchKit for a second and look at what we’ve seen of the hardware. The entire main board is shrunk down to a single unit: the S1. If you stop the middle introduction film at 4:46, you’ll see that it’s really a collection of chips and SMT components on an encapsulated multilayer board — not really a “single chip” as the narration says, but many large CPU “chips” nowadays are like that, too. Other than the S1, there’s of course the “Taptic Engine” assembly which does the wrist tapping, the crown sensor assembly, antennas and display, and the most important part: the battery.

Battery life is the make-or-break feature of the Apple Watch. iFixit’s disassembly of the Moto 360 watch shows why: there’s a square peg battery inside a round casing, rated at 320 mAh. Even though Motorola apparently build their own batteries, they don’t have enough volume to do a round one. Apple doesn’t have a volume problem and their casing is square, so they’re free to use all remaining volume for a longer-lasting battery.

The 320 mAh rating and the typical battery life of 12 hours of the Moto 360 means that the watch consumes, on the average, just under 27 mA. But they run Android on the watch, using an off-the-shelf TI ARM processor with attached RAM, flash memory, and so forth, so that figure is not surprising. In other words, it’s a stripped-down cellphone/MP3 player.

Suppose that Apple did its usual optimization of battery size, usage, etc., in a stripped-down iPod nano. It’s half the size of the nano, which has a 30-hour life, so we can assume half the battery, meaning 15 hours. OK, that would be marginally acceptable, perhaps.

But remember, the Apple Watch needs an iPhone nearby. In fact, many of the published functions, such as Siri, cellphone call response, GPS and so forth certainly use the iPhone’s hardware and software for that. Remember that one of the culprits of excessive battery usage is generic apps and processes running on the device. Remember that Apple, since the first iPod in 2001, has been very aggressive in optimizing their embedded systems. Remember that the first iPods and iPhones didn’t have any generic apps running on them, either. Remember that Apple already has technologies like Clang, OpenCL and Metal…

All that said, why run iOS and generic applications on the Watch at all? So here’s what I think likely about the real implementation.

  • Watch OS (or whatever it’s called — did they explicitly call it anything?) will not be a stripped-down iOS; maybe even not a Darwin derivative. It will be a highly optimized embedded system that has a few apps running in as few processes as possible. It will be very robust because it will be able to do only a fixed set of functions.
  • In other words, it will run only those things that may run while the paired iPhone is not available; we don’t know yet, but that might be just the timekeeping and pulse measuring apps. If the iPhone is there, the Watch will also work as a specialized I/O and display device for the apps installed there.
  • WatchKit will run on the paired iPhone inside a special server process; a matching iOS app will show installed Watch apps — probably those apps will be from the normal App Store, since they usually will have an iOS counterpart.
  • So, an installed Watch app will have at least some sort of preference app or pane on the iPhone; no use typing in passwords and such on the Watch, right? The part written in/for WatchKit will contain a server plugin that does the heavy lifting, data collating and communicating with the outside world, but it will also contain the application logic itself, commanding the Watch to do or display certain things.
  • I don’t mean to imply that the Watch will run a full WebKit client and the iPhone a web server, that might be overkill. Perhaps a useful subset of that, perhaps some variation of Display Postscript, some interpreted command language, or just a sequence of drawing orders? The important part is that there’ll be a single process on the Watch for doing the UI, and all the application-specific parts can be offloaded to the iPhone.

One consequence is that you can forget the idea of “jailbreaking” the Watch to connect to a non-iPhone, of course. Another one is that battery life might be at least a day, maybe even two or more. Nothing on Apple’s site so far contradicts any of my reasoning.

So, will WatchKit be accessible from Swift apps? Certainly. Will it itself be written in Swift? I doubt it for now. Maybe in iOS 9 some of the frameworks in iOS (and OS X) will have been rewritten, assuming that by then the Swift optimizer will be good enough. But that won’t be the case in a few months.

Possible but unlikely: WatchKit may have an API to download actual application code to the S1, which may (or may not) have an ARM-like architecture. Only in such a case — and since there will be no Cocoa/iOS frameworks on the Watch — I would expect the downloaded code to be in Swift (without optionals!), for extra safety; can’t have the Watch crashing and rebooting, right?

Update: Marcel Weiher kindly reminded me of CarPlay, which apparently works like that; nobody would say that cars are running iOS. On the other hand, in that case, the device is connected over USB (that is, reasonable bandwidth) and the car doesn’t have any battery life problems.

Comments welcome.

The direct download version of RB App Checker Lite 1.1 (281) is out, and RB App Quarantine also has been updated to build# 281 (its version is still 1.1 since nothing has changed for the user). Check out the release notes!

Most important: this version of RB App Checker Lite considers the new signing rules in TN2206 for OS X 10.9.5 and 10.10. In particular, for applications, code signatures are now checked recursively, both version 1 and version 2 resource rules are shown if present, and the spctl utility is called to check the Gatekeeper assessment.

However, this signals a new development: direct download and Mac App Store versions of RB Utilities will now, unfortunately, have different functionality. In particular, the MAS version of RB App Checker Lite (which is currently in review) will not be allowed to call  spctl, as this utility requires special entitlements to work; and the MAS version of RB App Quarantine (also in review), for the same reason, will not clear the quarantine flag. So for now you may want to download directly from the product pages.

Apple’s rules for calling attention to such differences are a little tricky to navigate but hopefully the MAS versions will be approved and in a future release I’ll develop a way to bring this functionality back through an optional download. Research is underway!

Photos licensed by Creative Commons license. Unless otherwise noted, content © 2002-2020 by Rainer Brockerhoff. Iravan child theme by Rainer Brockerhoff, based on Arjuna-X, a WordPress Theme by SRS Solutions. jQuery UI based on Aristo.