The Pixel C was probably never supposed to run Android

Was Android actually Google's third operating system choice for its new Pixel C tablet?

The Pixel C finally went on sale this week—but our full review notes that the convertible tablet feels like hardware in search of the software to make it a compelling product. Perhaps that's because, internally, Google engineers seem to have been searching for a compelling Pixel C software package for the last year and a half.

The contradiction between hardware and software is visible all over the tablet, so two examples will suffice. The hardware's keyboard and big screen would point to it being a productivity device, but the software's lack of a split-screen mode and apps optimized for the screen's size hamstring the Pixel C. The hardware seems geared for voice command functionality, given its array of four top-mounted microphones, but the software doesn't support Google's always-on voice commands.

It's also odd that the device hails from the Pixel team. Google has two big hardware brands: "Nexus," which covers flagship Android devices, and "Pixel," which denotes flagship Chrome OS devices. With the Pixel C, though, the Pixel team broke ranks and produced an Android tablet.

In our view, the Pixel C's irregularities all have a single explanation: the Pixel C was originally a Chrome OS device.

Back in July 2014, a new "Ryu" board (a "board" is just a reference to "motherboard"—a Chrome OS device under development) popped up in the Chrome OS open source repository. Further trips through the Chrome OS source code revealed that "Ryu" had a light bar, USB Type-C connectors, an Nvidia Tegra SoC, and wireless charging. That sounds an awful lot like the Pixel C (especially the wireless charging, which is used to charge the keyboard via the tablet's battery when closed).

Open up the Pixel C's software and take a look at Android's build.prop file—which lists all sorts of base information about the device—and you'll see "ro.product.name=ryu" listed in the properties. Based on this commit, it’s safe to say that at one point Google was definitely developing Chrome OS for its new Android tablet.

It appears that the Pixel C was planned as launch hardware for a new, all-touch version of Chrome OS which at some point got canceled—necessitating a switch to Android. The story is a lot more complicated than that, though. What follows is the best timeline we could piece together showing the Pixel C's troubled development history.

Plan A: Project Athena—Chrome OS' (canceled) touchscreen interface

We've long heard rumors of a "Chrome OS Tablet" being worked on at Google headquarters. Some of this touch work was clearly visible to the public. Since at least 2014, Chrome OS has had a semi-hidden virtual keyboard, which could be enabled via a developer flag. While most of Chrome OS' touch work could be attributed to the "laptop with a touchscreen" form factor of the Chromebook Pixel and a few other Chromebooks, a virtual keyboard would be mostly useless in a laptop form factor. This was destined for a tablet.

The Chrome OS team was actually working on a new interface called "Project Athena," which appeared to add a lot of functionality focused on touchscreen usage. In July 2014, right around the same time "Ryu" (the Pixel C) was introduced into the Chrome OS repository, the first references to Project Athena started to pop up.

The biggest addition we knew about was an experimental window switcher, which would have taken Chrome OS' window management UI from a taskbar-based interface to a cascading, scrolling thumbnail UI that looks a lot like what Android has today. Chrome OS' taskbar interface is great for a mouse, but the tiny buttons would make it a poor experience for touch users. No one from Google explicitly said this new thumbnail UI was for a Chrome OS tablet, but the change would provide much larger targets for touch use, and it would be less efficient for a mouse user because of the greater distance the mouse would have to travel. In our hypothetical Chrome OS Tablet world, we'd imagine that the cascading thumbnail window management would be for touch users, while the taskbar would stick around for mouse users.

Athena also added a few swipe gestures to Chrome OS. Dragging up from the bottom edge would enter "overview mode"—presumably the above thumbnail-based window switcher. Dragging in from the left edge would switch to the previous task (just as in Windows' touch UI), and dragging in and holding would enter a split-screen mode. Chrome OS is a windowed OS, so split-screen mode would seem redundant unless you consider that the usual mouse-based window resizing doesn't work well with touch. A "drag in" gesture like this would be great for our theoretical Chrome OS tablet.

Project Athena never shipped, though. In December 2014, about five months after the introduction of Ryu and Project Athena into the Chrome OS code base, the project was canceled.

Plan B: "Frankenboard"—The Chrome OS/Android Hybrid

With no touch interface to run on the 10-inch touchscreen and no mouse to drive the mouse-centric Chrome OS, the Pixel C was in a pickle. In March 2015, Android code started appearing in the Chrome OS Ryu device tree.

Chrome OS really can't run without a mouse, so apparently the team decided that making Ryu boot Chrome OS and Android would fix this. We know that Android wasn't a replacement at this point, because some of the work involved getting the Chrome OS boot loader to talk to Android.

Digitimes actually nailed this news a month before the commits became public. In February 2015, the site said Google was going to start pushing "2-in-1 Chromebooks" that would boot Android and Chrome OS. The device was going to be built by Quanta Computer, the same company rumored to manufacture the Chromebook Pixel for Google, and the report said it would be "Google branded," AKA a Pixel.

Just like with Project Athena, this idea lasted about five months—in July 2015, the dual-boot project seems to have been scrapped. A comment on a Ryu commit mentioned the cancellation while giving us more evidence into the device's hybrid nature, saying "Abandoned. frankenboard is dead." "Frankenboard" of course would be an excellent nickname for a device that was setup to boot Android and Chrome OS.

Plan C: Christmas is coming! Just flash Android and ship it

With Chrome OS now apparently out of the picture for the Pixel C, it looks like the decision was made to put Android on it and ship it out the door. In September 2015 at the Nexus launch event, Google announced the Android 6.0-powered Pixel C. This was just two months after the apparent cancellation of the dual-boot plans, leaving very little time to get regular Android up and running on the Pixel C.

At the launch event, the Pixel C immediately seemed like a rush job. The product's announcement was tacked on to the end of the high-profile Nexus 5X and 6P launch, where it was almost immediately forgotten about. While Google had about 30 to 40 Nexus phones set up for the attending journalists to try, there were a whopping two Pixel Cs set up at the event. The Nexus phones went up for pre-order the day after the event and shipped about 20 days later. For Pixel C customers, Google followed up the announcement with two entire months of radio silence, leaving many to wonder if the device was ever coming out.

Google's silence was only broken with a message along the lines of "Surprise! The Pixel C is for sale right now!" on December 8. The Pixel C's launch timing meant it missed the lucrative Black Friday/Cyber Monday shopping period, a sure sign of a troubled launch. In fact, we're still not sure what the retail box for the Pixel C looks like—our review unit was luxuriously shipped in a plastic bag inside of a brown cardboard box.

A quick look at the Pixel Team's recent Reddit AMA shows that the team is scrambling to implement basic Android features like always-on voice recognition, promising that "Android N" will bring features that the device desperately needs (like a split screen mode).

The path to a Chrome OS and Android merge runs over the Pixel C

Connect all the dots from the public Chrome OS commits, and our timeline looks like this:

• June/July 2014: "Ryu" (the Pixel C hardware) and "Project Athena" (touch-centric Chrome OS) are introduced into the Chrome OS repository.
• December 2014: Project Athena is canceled.
• March 2015: Commits for making Ryu dual-boot Chrome OS and Android are posted.
• July 2015: "Frankenboard" (the dual-boot Ryu project) is canceled.
• September 2015: Google announces the Android-powered Pixel C (code-named "Ryu").

So why did the first hiccup happen in the first place? Why was work on a Chrome OS tablet abandoned? Our guess is that it has something to do with Chrome OS and Android merging.

Introducing a new Chrome OS form factor, only to have it superseded by a crazy hybrid OS a year or two later, probably wouldn't be received well by customers. Chrome OS customers, a large portion of which are schools and businesses, would especially value platform stability. Convincing those customers to move over to the new hybrid OS will be hard enough, and creating and then killing a new form factor in the span of two years wouldn't inspire a lot of confidence in Google. It's about not further complicating what will already be a rocky transition.

This narrative explains why the Pixel C seems so lackluster today; it wasn't supposed to be this way! It was supposed to be a flagship Chrome OS tablet worthy of the "Pixel" name. It seems those plans were canceled, though, and what we're left with is the Android-powered Pixel C—just another Android tablet.