I’ve just wrapped up the latest in what has become a long series of periodic smartphone migrations, conveniently rationalized each time using words to the effect of “it’s for my job; I need to keep up-to-date on the latest-and-greatest technologies.” In this particular case, my motivations for upgrading my gear encompassed both technical and non-technical factors, but I didn’t upgrade as far as I could have, and my restraint was only partly motivated by fiscal concerns. As such, especially since cellular handsets are today’s highest-volume case studies of low-power design, I thought I’d devote this month’s writeup to this particular story.
For the past two years, and until the past few days, the Google Nexus One (developed by HTC) had been the handset of choice for my AT&T cellular account:
There was a lot to like about the Nexus One, thereby explaining why it was active in my cellphone stable for so long. It was thinner and lighter than the iPhone 3GS I’d been using previously, and toted a crisp (albeit washed-out in bright lighting environments) AMOLED display. It included a dual-microphone array along with an Audience audio processing IC for ambient noise suppression purposes. As one of the first handsets available with Android 2.1, it was an ideal platform for familiarizing me with Google’s O/S. And as with its Google-branded successors, the Nexus S and Galaxy Nexus, it’s far less function-restricted than a handset obtained from AT&T or another service provider:
- It enables ‘tethering’ over USB, Bluetooth and Wi-Fi, transforming the phone into a wireless WAN modem for a connected laptop computer or other device
- It’s carrier-unlocked (although my AT&T-tailored phone doesn’t, for example, support T-Mobile’s UMTS frequencies, and visa versa), as well as offering an unlockable bootloader, and
- It comes with a ‘stock’ Android build, absent the flashy (translation: memory- and performance-gobbling) branded GUI “skins” of its peers, along with bundling only a few third-party applications
So why did I abandon the Nexus One? For one thing, after two years’ worth of steady use, its Li-ion battery was no longer holding adequate charge, the replacement battery I bought for a few bucks on Ebay from a China-based retailer refused to reliably charge, and the retailer wanted me to ship it back to China before he’d replace it (yeah, right). But that was a minor nit; I could have always bought another battery somewhere else.
Far more frustrating was the Nexus One’s scant 512 MBytes of onboard flash memory storage, coupled with the fact that only 190 MBytes’ worth of it was user-accessible. To Google’s credit, the company had continued providing operating system (and corresponding app) updates for the handset all the way through Android v2.3.6…although the company has no intention of further upgrading it to Android 4.0 (aka “Ice Cream Sandwich”…Android 3.0 “Honeycomb” is a tablet-only version), thereby signaling its obsolescence.
But those O/S and app update patches incrementally consumed chunks of that 190 MByte allocation, forcing me to frequently de-install other programs in order to keep the handset above the 20 MByte-free threshold necessary to keep “push” email access, account sync and other important functions running. Android v2.2 “Froyo” had finally enabled the option to install (or move, if already installed) part-to-all of a program in an encrypted fashion to expansion storage such as the Nexus One’s micro SD card. But few developers embraced the feature, far fewer went back and retrofitted their existing apps with this capability, and those that did only harnessed expansion storage for a portion of their programs’ memory needs. So the 4 GByte microSD card in my Nexus One went almost completely to waste, at least from a program storage standpoint.
After experiencing the functional freedom of a Google-branded handset, I wasn’t motivated to consider a more locked-down Android alternative. But the few AT&T UTMS frequency-friendly Nexus S units I found on Ebay were all quite pricey (I’m loath to lock in to a multi-year contract, which is the only way to get a brand new unit via Best Buy). The Galaxy Nexus is even more expensive, and an official AT&T version is not yet available (imported GSM versions support AT&T’s UMTS bands, but don’t come with US warranties). And anyway, after a few years’ worth of Android experimentation, I didn’t feel particularly compelled to continue, at least for now.
I still had the old 16GByte iPhone 3GS, and I briefly went back to it. But, after six months’ worth of experience with a 640×960 pixel Retina display (I also own a 32 GByte iPhone 4 on Verizon, for work purposes), I cringed every time I peered at the iPhone 3GS’s comparatively fuzzy 320×480 pixel LCD. The iPhone 3GS’s back-facing camera, capable of capturing only 3 Mpixel stills and VGA-resolution video, also lacks LED flash illumination augmentation capabilities. There’s no front-facing camera for FaceTime or other videoconferencing purposes. And from a cellular data standpoint, the iPhone 3GS doesn’t support HSUPA, thereby capping its maximum upload speed at 384 kbps.
Conversely, I could have gone with a latest-generation iPhone 4S:
They’re currently available refurb’d in the 16 GByte variant for $500 plus tax and shipping, for example. As a techie, you might think I’d go for the latest-and-greatest, and I admittedly was tempted to do so, predominantly motivated by the handset’s 8 Mpixel still and 1080p video capture capabilities (Siri voice recognition…not so motivating). The system’s dual-core A5 (ARM Cortex-A9) CPU would stave off obsolescence longer than either the single-core ARM Cortex-A8 in the iPhone 3GS or iPhone 4, not to mention the comparative GPU capabilities of the three generations’ worth of hardware. And a 32 GByte storage-option selection, although pricier in the near term, would also ensure plenty of memory no matter how many (and how big) the apps I’d install and later upgrade, not to mention the audio tracks, photographs and video clips I might want to archive on the handset.
However, from existing experience with the iPhone 3GS, I felt confident that 16 GBytes of storage would be sufficient; I could always offload pictures and other data files via Dropbox if I ran short of space. Similarly, the iPhone 4′s (which looks just like the iPhone 4S, by the way, thereby negating the need for a third picture) CPU and GPU were seemingly more than sufficient from a performance standpoint, albeit not leading edge (but then again, without the battery life “hit”, either). The iPhone 4S supports up to 14.4 Mbps HSDPA, versus “only” 7.2 Mbps on the iPhone 4, but I doubt that AT&T will end up doing much network upgrade to this top-tier downstream speed grade, instead focusing on next-generation LTE. And I still carry a standalone camera with me (though I admittedly can’t tell you the last time that I turned it on), so I don’t really need an 8 Mpixel-capable camera phone (no matter how much I might want one).
Another iPhone 4-vs-4S differentiation, this one non-technical in nature, ended up (along with relative price) being a dominant factor in my upgrade decision. Through the iPhone 4 generation, it was possible to retroactively purchase AppleCare extended warranty coverage through the first year of the handset life. I did this when I bought my original iPhone 3 on Ebay (along with the iPhone 3G upgrade to it), as well as with the Ebay-sourced iPhone 4 on Verizon. With the iPhone 4S, however, Apple’s policy changed. AppleCare became $30-more-expensive AppleCare+, which is predominantly only available when you purchase the iPhone 4S, although the product page also indicates that you can also “buy it within 30 days of your iPhone purchase at Apple Retail (requires a Genius Bar appointment, inspection of the iPhone & proof of purchase).” In neither case, practically speaking, would I be able to obtain an official Apple extended warranty for a handset I might purchase used or refurbished, unless the original owner had bought the warranty with the phone (in which case the remaining coverage would be transferable to me).
I ended up buying a like-new 16 GByte iPhone 4 off Ebay for $320. The original owner had purchased it last October, but in his words, it was:
Barely used. Has been sitting in the box since last November. I had to switch to Verizon due to AT&T signal issues at my office.
Even better, he’d already purchased an AppleCare extended warranty for it, saving me an incremental $70, and the coverage doesn’t expire until October 2013. It arrived last Wednesday, the micro SIM from AT&T (replacing the SIM in the Nexus One) showed up on Friday, and everything looks and works great. And here’s the kicker; by selling both the Nexus One and iPhone 3GS predecessors on Ebay, I should be able to at least cover the price I paid for the iPhone 4 successor, if not turn a “profit.”
So, readers, sometimes it pays to not be an early adopter! And that guidance applies not only to you as a consumer of technology but also as a technology developer and implementer. Selecting mainstream system building blocks may drive your marketing counterparts nuts, because it means they won’t be able to promote the latest-and-greatest eye-catching product specifications; “quad core!”…”flux capacitors!”…”1.21 gigawatts of electricity!”…”42!”. But it also means assured supply, low cost, mature development tools, and already-squashed bugs, among other benefits. Which scenario would you prefer?
p.s…By the way, for any of you who might have chuckled at my migration from Android back to Apple iOS in the context of my earlier-described fondness for platform openness…that’s what jailbreaking is for To wit, the A4 SoC in the iPhone 4 can be more easily and completely jailbroken than can the A5 CPU in the iPhone 4S; another reason to go with prior-generation technology!