WWDC 2014

I'm just going to open with this:  Every single Android proponent who claims that the iOS 8 announcement is a reaction to Android, and an unimpressive copy of everything that Android's been doing, is wrong.

It's not copying Android.  Not even a little.  Yes, there are features that were clearly in Android before they were in iOS.

Palm would like to have a word with all of you.

Improved inter-application data sharing?  PalmOS.  Multitasking via cards?  WebOS.  Customizable input methods?  PalmOS.  Pop-up notifications?  WebOS.  Interaction with notifications without leaving your app?  PalmOS.  iOS/OS X Handoff?  VERY late WebOS.  You can claim that Apple didn't come up with it, but you can't claim that Android did it first.

Now that that's all out of the way, there's a simple fact that Apple's integrated these pieces much better as technology caught up to support it.  Sure, they waited.  TouchID is, simply, the most reliable fingerprint scanner I've ever used.  iOS devices have insanely long battery life, even with a 4G/LTE radio running.  Siri isn't remarkable, but it's functional, and improves incrementally.  This has been Apple's game since 1984, so complaining about it now means you've ignored three decades of development.

So, what does that mean?  Well, not much.  iOS 8 has some neat improvements.  So does OS X 10.10.  The two combined are where it gets interesting, because Handoff lets you rapidly adapt your workflow to suit your needs.  Mail Drop cuts the bullshit of sending large attachments.  Swift sweeps away a decade-plus of Objective C cruft.  AirDrop will work the way we've all wanted it to work.  iCloud file storage is finally uncrippled.  Per-app battery usage is a power user tweak that's been long overdue.  HealthKit gets something right that S Health didn't:  Apple's connecting it to one of the most prestigious medical centers in the country, in addition to the biggest Electronic Medical Record vendor in the USA.  This isn't just a fancy user interface, this is connectivity.

OS X 10.10 unifies the user experience, which I think is going to upset some people, but it's also going to ease adoption of OS X for iOS users, and vice versa.  Things will be familiar, and that's not a bad thing, though I suspect it will bring detractors just like it did when iOS 7 came out, or when Microsoft introduced Vista, or Windows 8.  That's a minor detail, compared to the continuity features, and that's where things get interesting:  No one else could have pulled it off.  Google could have made it function, but how many people are swapping between Android phones and Chromebooks?  Microsoft could as well, but likewise:  How may users swap between a Windows 8 system and a Windows Phone?  Apple's just in the right place at the right time, owing to a massive iOS user base and the use of iOS as a gateway drug into OS X.

This is some Star Trek-level stuff coming around, and while I expect Microsoft and Google to answer in kind, I'm okay with that:  It just means that everything's going to get better, and it probably won't matter much who "wins," since the real winners are us users.

The Problem with Wearable Computing

2013 wrapped up with an interesting explosion in wearable gear:  The Smartwatch came back with a vengeance, giving us the Pebble and the Galaxy Gear, Fitbit’s new Force wristband gave a similar option to fitness/Quantified Self adherents, and Google Glass came along and already brought up new concerns about privacy and information overload.

The issue is that most of these products are deeply flawed, and simply aren’t going to work for the average user in their current form.  While I’m an avid Pebble owner, and have happily seen the device evolve from an occasionally-useful toy to a functional companion device, it’s still a wristwatch, and wristwatches seem to be vanishing as people rely on their smartphones to tell them the time.  I can’t blame them, of course.  Cell phones have long been more accurate timepieces, relying on synchronized time signals from the towers, and ultimately, GPS radios. Adding killer features to the wristwatch ultimately appeals exclusively to those willing to wear a wristwatch in the first place.

The fitness wristbands are a different animal, of course.  It’s nice that the Fitbit Force doubles as a timepiece, but the main advantage to it is that it carries the full Fitbit sensor suite in addition to all of the display capabilities of their belt-mounted sensor.  It’s also more discreet than a smartwatch, and serves a purpose beyond the typical watch.  Sadly, it’s got another problem:  Motorola, Apple, and Samsung have all started to build low-power motion-tracking capabilities directly into their phones.

Google Glass, on the other hand, provides a potentially always-on, always-connected heads-up display for your life.  It can record audio and video for you.  With sufficient image recognition and location awareness, it could very well be the first commercial product to allow for augmented reality, a pipe dream of many wearable computing advocates.  Unfortunately, it’s got two problems that are very hard to get around.  The first is a matter of fashion, but the second is a matter of capability.

Fashion’s an important part of product design.  People will accept a less-capable product if it doesn’t look out of place.  Google Glass and the Pebble both need some work on this front, though many of the fitness bracelets are pretty good about this.  For all of its technical flaws, the Jawbone Up was a sufficiently attractive product to turn heads.  The Pebble is a bit clunky, and I’ve taken to replacing the wrist strap to get around this problem.  It certainly helps, though the actual device could stand to be a little thinner and the sides could use a slight beveling.  Google Glass is a very different animal in this regard.  It’s very difficult to hide that display element, along with the camera, earpiece, and battery at present, and many people find the device clamped onto someone’s head fairly off-putting.

This isn’t a unique-to-Glass problem, either.  Medical devices strive to blend into the patient’s body as well, from prosthetic limbs to cochlear implants, as the concept of “normality” tends to be a major issue.  What’s truly funny about this is that it even goes back to the Pebble:  I’ve had to learn to ignore it, because if you check your wrist every time that you get an email, text message, or other important notification from someone, people start to think that you’d rather be somewhere else, and are actually checking the time!

So, how do you blend these functions together?  Unfortunately, this is going to lead us right back to head-mounted units:  You need to keep the people around you comfortable with your “enhanced” awareness if you want it to take off, and that requires a situation where the awareness is understood, or where your hardware is fundamentally invisible.

Let’s start with the former, an area where Glass and its ilk can excel right now.  Non-social tasks, such as surgery, sports, and vehicle operation can be aided with Glass without the handful of people around you really noticing a difference.  There’s little difference between using Glass as a navigation aid and relying on an on-dash GPS, so long as that’s all it does.  A surgeon certainly doesn’t work alone, but they are often more focused on the patient than their coworkers, so a headset won’t make much difference.  A skier may have friends on the slopes, but they’re often wearing a helmet and goggles, so adding Glass, or the Recon Snow, really wouldn’t register.  It probably even helps that some winter sports enthusiasts like to record their runs using a helmet-mounted camera.  In short, if you’re already focused on a task, a device that melds into that focus is a good fit.

Now, for invisibility, here’s the real problem:  Watches are more invisible, but you have to look at them.  Glass is less invisible, but that’s mostly a function of the technology.  We need smaller camera modules, smaller batteries, and slimmer batteries, all without compromises.  Once you get that into something that looks like a pair of prescription glasses, you’re most of the way there.  I’m sure we’re not too far on this:  Only 25 years ago or so, I met a teacher who was hard of hearing.  She had hearing aids built into her glasses, and I didn’t notice for quite some time.  While the hearing aids are definitely designed to restore an ability we all take for granted, the concept remains the same for adding a new sensory option.

I think that there’s a lot of room for improvement across the board, and I’m really interested to see what we get in 2014, though I’m certainly skeptical that Google Glass will shrink down by enough in the coming year, and I certainly don’t expect Apple to jump into this market just yet.  There are simply too many compromises in the core requirements of invisibility and battery life for the typical Apple product.  Will it start to spill into niche markets, however?  Absolutely, and I’m very interested to see where that leads.

Listening In On Encryption

ExtremeTech covers this in detail, and there's certainly some reasons to be concerned, but I don't see it as something to lose sleep over.

In summary, by listening to the noise generated by the voltage regulator in a modern CPU, you can make a pretty good guess as to what bits are being processed.  Get it right, and you have the decryption key, which unravels the entirety of the GPG algorithm.

I can think of a few ways around this, some more elegant than others.

• Generate white noise that might confuse the eavesdropper.
• Disable the CPU scaling capabilities, thus removing much of the noise from the voltage regulator.
• Junk math.  Run some parallel decryption-like operations on a random data source.

I'm a bigger fan of the third method:  You don't compromise overall system behavior, and you don't need to rely on speakers being active.

The real reason I'm not worried in this particular case is that the encryption that was cracked is GPG. Once someone comes up with a fix and submits a patch, it will be officially rolled into an upcoming release, and in the meantime, the patch will be out there for those who want it.

Overall, really cool stuff, and just another example of why security is a process, and not a product:  It's never enough to own a lock, you also have to use it properly and maintain it over time.

Phonebloks/Project Ara

Phonebloks was already an interesting concept, and now Motorola's getting behind it.  I think it's a solid idea for the reason that I think the new Mac Pro is a solid idea: Not all components are rendered obsolete at once.

That said, it's not without problems. 

1. Size.  Durable, modular connectors take up valuable space.  If you're comfortable with a larger phone, this isn't a problem.
2. Integration.  The big push in mobile has been to get into System-On-A-Chip (SoC) designs, merging the CPU, RAM, and all wireless capabilities into a single chip. 
3. Efficiency.  Again, the SoC solution fits here:  You use less power when you're not "leaving" one chip to get to another function. 
4. Performance.  See above. 

Mobile's a little different from desktop in this regard, and I see a huge potential for this in low-income and underdeveloped populations.  That's a huge market that's massively underserved.  I don't see it selling well to a luxury market, however, and I absolutely expect this to be picked up by the tinkerer/hacker communities.

Just for some basic off-the-cuff thoughts, here's some lifecycles that I've seen for mobile components, in terms of how often they could be replaced:

1. Displays - Two years.  Apple went through three iPhone generations on the same display, then two more on the Retina Display, and now they're on year two for the 16:9 Retina Display.
2. CPU - One to three years.  Upgrading is nice, but you can realistically stretch a CPU until it's been outclassed twice before really worrying about replacements. 
3. RAM - Two to three years.  This depends on the OS, but there's been a push to make things more efficient, not less.  I suspect that this will add some life to the memory module. 
4. Graphics - Variable.  Do you need to play games on your phone?  Upgrade yearly.  No?  Maybe every three-four years. 
5. Wireless - Variable.  Bluetooth can probably get by with anything from two years to indefinitely.  Wifi, maybe two-three, again, depending on your own needs.  I've got some devices at home still using 802.11g. 
6. Cellular - Ideally, every two years.  Power efficiencies, speed boosts, jumps from 3G to 4G and beyond, there's definitely a few good reasons to swap out occasionally. 
7. Cameras - Whenever you'd like.  If it's even halfway decent today and has a flash, you have to be pretty into snapshot photography to consider this a major factor. 

Honestly, I'm more concerned about the system chassis than I am about the modules:  As things get faster, will you be able to shuffle data to and from modules as quickly as you need?  That goes to the bus/chassis design, and that's something that I would certainly leave to Motorola to solve.