Stop me if you've heard this one before: A new version of Android is right around the corner — and with it comes some lofty talk about how this'll finally be the year Android manufacturers start taking upgrades seriously.
That's the story with Google's upcoming Android P release, thanks to the software's integration of Project Treble — a new "modular base" for Android that makes it easier for manufacturers to process updates. In short, Treble keeps the guts of Android in their own standalone layer within your device's storage. The hardware-specific code needed to make the device run properly, meanwhile, lives in a totally separate lower layer. It's kind of like a fancy cake, only with less unsaturated fat than the kind you normally buy.
The idea, then, is that whenever a new version of Android comes out, the manufacturer can get it ready to roll with less work than what was previously required — because all that stuff in the cake's lower layer can now be left alone. Before, the manufacturer had to take that stuff and then mix it back in with the main operating system layer every time a new update came along. And that made for a lot of baking.
This is the first release cycle where a large number of phones are standing by and ready with Project Treble support built in — and already, the results are apparent: Google is gloating about the fact that Treble's presence allowed it to offer the new Android P beta on seven devices outside of its own Pixel phones — the Sony Xperia XZ2, Xiaomi Mi Mix 2S, Nokia 7 Plus, Oppo R15 Pro, Vivo X21, OnePlus 6, and Essential PH‑1. That's a noteworthy first for the platform.
We've worked closely with Qualcomm, MediaTek, and Samsung Electronics’ System LSI Business to co-develop their [software for controlling the chips used in phones], starting with Android P. Their [software packages] are now ready for Android P on a much-accelerated schedule, reducing the overall effort significantly. These silicon manufacturers are now able to provide a stable and high-quality release much earlier than before, allowing OEMs [a.k.a. original equipment manufacturers — a snazzy way of saying device-makers] to bring the latest innovations of Android to their customers across the globe.
Qualcomm got in on the action, too, putting out its own press release touting the fact that it was ready to support "fast commercial availability" of Android P for devices that use its processors:
By having early access to Android P, Qualcomm Technologies optimized its software on Snapdragon 845, 660, and 636 mobile platforms to ensure readiness for OEMs to upgrade to Android P at the time of launch. Qualcomm Technologies' leadership position and scale in mobile allows for Google to improve the speed of its OS upgrade cycle, and for OEMs to bring the latest software enhancements to consumers more rapidly than previous Android OS releases.
Well, whoop-de frickin' doo — right? I know. But hang on: There's a reason I'm quoting all this PR gobbledygook.
Yup, you guessed it: Time to put this all into perspective.
Android P and Project Treble's unspoken layer
Here's the thing: Project Treble is without a doubt a significant leap forward for Android. It makes a lot of sense for Google to split apart the layers of the code and reduce the amount of effort involved for manufacturers to process and deliver an update. That's meaningful progress — and quite the coup from an engineering standpoint.
But in talking about the layer-based changes Treble brings to Android, there's one more layer that rarely gets mentioned — something we'll call the frosting in our tasty little virtual cake. Frosting, you say? Why, yes: In addition to that lower layer of hardware-specific code and the upper-layer of Google's actual operating system code, there's an upper-upper-layer lots of manufacturers like to add into the batter.
It's the layer that lets Samsung have its custom TouchWiz interface, along with all the extra features it adds into its phones' software. Same for LG, Huawei, and any other manufacturers that put their own twists on Android. Their frosting doesn't just sit on top of the main Android code layer, either: It frequently mixes in with it, since the changes aren't only superficial decorations but often modifications and additions to the core Android product.
Project Treble, however, is all about that lower layer — the hardware-specific code. It has no effect on these higher-level changes so many manufacturers like to make to the operating system, and it does nothing to streamline that part of the process.
You can see signs of that fact right now, in fact: Maybe you've noticed that even with the broader-than-usual beta availability of Android P this year, the non-Pixel devices that support the software are currently running Google's version of Android — not their manufacturers' typical customized versions of it. Well, this is precisely why. So, yes: Those devices are able to support Android P in its purest form (with varying levels of success), but that isn't the same flavor of software that'll actually be rolled out to the devices once Android P's final release arrives.
And that's not all, either: There's one other big ol' lump of sugar we need to address.
The broader Android upgrade problem
The plain and simple truth, as I've pointed out before, is that while Project Treble makes the act of processing and delivering updates easier — with that above-mentioned asterisk — it still requires manufacturers to make upgrade delivery a priority if it's going to have any meaningful and wide-reaching impact. And if history's any indication, that type of customer-centric focus isn't necessarily something we can count on from most Android device-makers, no matter how hard Google may try to make it happen.
Back to that déjà vu we talked about at the start of this story: Google's taken steps to make the Android upgrade process easier for manufacturers before — and each time, we've heard the same narrative about how this is the year the device-makers are gonna get their acts together, gosh darn it! But then each time, without fail, we've ultimately been reminded that regardless of the circumstances surrounding them, most of the manufacturers just don't give a damn.
Setting aside the odd "Android Update Alliance" announced and then pretty much immediately abandoned in 2011, Google's nudges started with Android L in 2014, when the company offered its first early preview release so device-makers could have more time with the software ahead of its launch. With Android M the following year, Google served up that preview even earlier in the spring — more weeks to work with the code and prepare in advance should help, right? By the time Android N rolled around in 2016, Google had pushed back its time frame earlier yet and added an extra 36 days into its preview-to-final-release window.
And yet — well, you've seen the charts, right? By and large, Android upgrade performance has been getting progressively worse year after year, despite the ongoing efforts and advances. Google itself is the only real exception, with its consistently commendable support for its own Pixel (and previously Nexus) devices.
JR Raphael
(Click image to enlarge)
It's like I've said before: The underlying problem with Android upgrades isn't entirely technical. It's also the fact that the companies making and selling Android phones — with the exception of Google itself — have no real motivation to make timely, ongoing upgrades a priority. Idealistic as we may want this industry to be, the business realities driving manufacturers' behaviors are actually quite understandable.
So, yes: Project Treble absolutely does give manufacturers less work when it comes to updating their Android devices. But they'll still have some resource-requiring chores to manage — some phone-makers more than others, depending on how much meddling they choose to do. And, crucially, nothing about their underlying priorities or the business models that define them seems to have shifted.
Put another way, the years of subpar software support seem less about logistics and more about incentive. Project Treble could certainly have positive effects in some areas of the Android ecosystem, but it doesn't do anything to address that core issue.
And you know what? Maybe I'm reading too much into this, but I can't help but get the impression that Google is keenly aware of that. All the PR surrounding Treble's effect on Android P reads to me like a message directed at manufacturers more than anything — a deliberately public declaration that, hey, we've done our part. The chip-makers have done their part. If you fail to get software into your users' hands in a timely manner yet again, it's totally on you — and now, we're making that crystal clear to everyone.
Allow me to repeat a couple of highlights from those excerpts I quoted above, with my own emphasis added — first, from Google:
These silicon manufacturers are now able to provide a stable and high-quality release much earlier than before, allowing OEMs to bring the latest innovations of Android to their customers across the globe.
And then from Qualcomm, whose press release was clearly orchestrated with Google and meant to emphasize its partnership with the company:
Qualcomm Technologies' leadership position and scale in mobile allows for Google to improve the speed of its OS upgrade cycle, and for OEMs to bring the latest software enhancements to consumers more rapidly than previous Android OS releases.
A quote from a Qualcomm exec further drives the point home — again, emphasis here is mine:
"We are excited to work with Google to pre-integrate our software with Android P, making it production ready for OEMs," said Mike Genewich, director, product management, Qualcomm Technologies, Inc."Through our strengthened relationship, Google and Qualcomm Technologies are set to expand the market for Android and make it easier for OEMs to launch devices based on Snapdragon mobile platforms."
See it? "We've done this to make their jobs easier. What happens next is up to them. Capisce?"
And remember, too, that the manufacturers involved with the Android P beta thus far are Sony, Xiaomi, Nokia, Oppo, Vivo, OnePlus, and Essential. Notice some names missing from that list? This all seems squarely aimed in their direction.
So those three words I alluded to in the headline: Time will tell. We can talk endlessly about the awesomeness of Project Treble and how brilliantly it sets the stage for swifter OS updates, but what none of us can know is how much difference it'll ultimately make — because the real deciding factor is how much companies like Samsung, LG, and Motorola actually care. The exact circumstances may have differed in past years and the level of change may have been somewhat less grand, but it's hard not to feel like we've been down this road before.
There's every reason to be cautiously optimistic about what Project Treble could accomplish — but as someone fully informed about the context around it, there's also every reason to be skeptical.
Let's see what this summer will bring.
Sign up for JR's new weekly newsletter to get this column along with bonus tips, personal recommendations, and other exclusive extras delivered to your inbox.
[Android Intelligence videos at Computerworld]
