A long-rumored Google project that would combine Android and Chrome OS into a single platform has allegedly been scrapped. What does this mean for the future?
The future is impossible to see, as Yoda once observed. And when it comes to Andromeda, even the past is a bit fuzzy: This project, never publicly acknowledged by Google, had a grand aim in an era in which other client platform makers like Microsoft and Apple are adding complexity to their lineups. But the truth is, we never really knew what Andromeda was.
Most rumors pegged it as a single new OS that could replace both Android and Chrome OS, something that combined the best of both into a cohesive whole. When Google announced Android apps running on top of Chrome OS last year, most saw that as a half-step to the Andromeda future, a way to test the waters of Android/Chrome OS interoperability.
Sign up for our new free newsletter to get three time-saving tips each Friday — and get free copies of Paul Thurrott's Windows 11 and Windows 10 Field Guides (normally $9.99) as a special welcome gift!
"*" indicates required fields
A year later, we know that those efforts have gone poorly. And that Google has had a much harder time than it expected just getting Android apps running on Chrome OS. Perhaps Andromeda was an even more daunting problem.
As recently as last fall, I was openly wondering whether Google was on the cusp of actually announcing Andromeda and its plans for getting customers from two separate platforms onto this new thing. In my defense, we didn’t yet know that Android apps on Chrome OS were a complete disaster.
“The pieces are all in place,” I wrote at the time. “Structurally within Google. And architecturally within the products themselves. It’s what Microsoft finally did with Windows 10, whereas it had previously developed Windows for PCs and Windows phone separately.”
As it turns out, the pieces have apparently been scattered to the wind. This week, 9to5 Google’s Stephen Hall tweeted that multiple sources have now told him that Andromeda, whatever it is, has been killed.
So yeah, got a second source on this now: Andromeda was shelved. Some of the work being moved to other things, though. Trying to learn more.
In other tweets, Hall explains that Andromeda was real.
Andromeda was absolutely real. It was Android-based and sought to bring Android to different form factors. Google was preparing hardware … Hardware like “Bison” laptop, Huawei Nexus 7 tablet, others we never heard about. Assuming all shelved this point, but work won’t be wasted.
Hall also points to the obvious successor for Andromeda in a follow-up tweet.
Fuchsia, a separate project that you are all aware of by now, is not dead and effectively serves as Andromeda’s spiritual successor.
Ah yes, Fuchsia.
This one is interesting because Google has, in fact, admitted that Fuchsia is real. And we know a few things about this project: Like Android and Chrome OS, it’s a client operating system, and based on recent screenshots, it appears to be a successor to Android. Which, when you think about it, was the point of Andromeda as well.
If I had to guess—and I do—Google’s bad experiences with Android apps on Chrome OS have likely influenced its view of the future. And rather than meld these two things together, it will do what it should have always done: Base the future on Android, not Chrome OS. Because it’s much easier to add a full-featured web browser to a mobile OS than it is to run mobile apps on top of a web browser.
It seems rather obvious in retrospect. But then, this was my complaint when Google announced Chrome OS in the first place.
So we’ll see what happens. Just don’t hold your breath looking for answers. It’s now clear that Google’s future ambitions have been thwarted by reality, and it will be a while before it can move to the next big thing.
skane2600
<blockquote><a href="#122457"><em>In reply to rameshthanikodi:</em></a></blockquote><p>The term Real-time has been casually tossed around in recent years, but I seriously doubt that Fuchsia qualifies as an RTOS from an engineering perspective. The fundamental characteristic of an RTOS is deterministic timing behavior. That means that caching, garbage collection, anything that causes timing of the system to vary, is off the table.</p>
Bats
<p>Thwarted by reality? The reality is already Android IS ON Chromebooks. That's the reality. That's Google's starting point. So where exactly is the "thwart?"</p><p>After all, the fact that Google has been able to get Android on Chromebooks is a great achievement. We are talking about two different OS's in one device. That's like getting Apple's iLife suite working in Windows.</p><p>It's funny how Paul, is coming to these so-called "conclusions" based on his analysis of heresay. LOL. Is that actually professional?</p><p>After all, hasn't Paul learned from the iPad misstep? If the iPad hadn't changed the game, there would be NO Surface.</p><p>In all my years observing tech, one thing is for sure: All things are possible with code. </p><p>For Paul to come out and state the opposite, especially based on heresay…imo, just isn't stupid but also unprofessional.</p>
skane2600
<blockquote><a href="#122544"><em>In reply to Boris Zakharin:</em></a></blockquote><p>I've been down this road with him before and he'll never be convinced no matter what evidence you present. No doubt there are OS's out there created after NT that none of us here even know about. </p><p><br></p><p>He's right about one thing though: writing system software IS hard.</p>
skane2600
<blockquote><a href="#122585"><em>In reply to Jeff Jones:</em></a></blockquote><p>Yes, and of course, all OS's have some technical problems.</p>
skane2600
<p>Just as Microsoft doesn't need to succeed in mobile to survive, Google doesn't need to succeed on the "desktop". They should focus on their strengths rather than trying to create a single OS to "rule them all".</p>
skane2600
<blockquote><a href="#122538"><em>In reply to VancouverNinja:</em></a></blockquote><p>I don't see the connection. MS is essentially a non-player in the smartphone business and smartphones have the wrong ergonomics for productivity use.</p>
skane2600
<blockquote><a href="#122556"><em>In reply to quick_razor360:</em></a></blockquote><p>Ideally you'd want your customers to use your products anywhere those products would be appropriate and useful. But companies have to consider where they are most likely to be sucessful since they can't be in the "everything" business.</p>
skane2600
<blockquote><a href="#122601"><em>In reply to brduffy:</em></a></blockquote><p>The problem with Continuum is that it's not that useful to have a portable productivity machine that you can only be productive on when you're tethered to a desk. The solution for true portable productivity already exists – it's called a laptop.</p>
skane2600
<blockquote><a href="#122717"><em>In reply to Demileto:</em></a></blockquote><p>What screens? Anyway, If they ban laptops on all flights (unlikely) TSA would probably not know the difference between a portable keyboard and a laptop and might make you check it.</p>
skane2600
<blockquote><a href="#122829"><em>In reply to Demileto:</em></a></blockquote><p>It's not standard equipment on all planes and the current trend is to avoid them:</p><p>https://www.nytimes.com/2017/02/16/business/streaming-flights-movies.html</p><p><br></p><p>And what is this projection capability you spoke of?</p>
skane2600
<blockquote><a href="#122957"><em>In reply to Demileto:</em></a></blockquote><p>It's not clear if Trump even has the legal authority to ban laptops on domestic flights. Some carriers like SouthWest have never had screens, so there aren't likely to ever add them. But the point is that those disappearing screens were never computer monitors but integrated entertainment systems, so your argument falls apart even if laptops are banned.</p>
skane2600
<blockquote><a href="#123115"><em>In reply to Demileto:</em></a></blockquote><p><br></p><p>It's easy to say "you're just lacking vision" simply because somebody disagrees with you. Here are the relevant questions: Will Trump actually ban laptops on all international flights or will it just a remain an option? Would any laptop ban survive past the Trump administration (8 years max if not)? Are airlines going to keep or add screens to seats just because of a ban on international flights? If so, will they be willing to design or have designed new devices that allow a connection to computers rather than the entertainment units they used to use? If so, how would the computers connect to the screens and would they need an industry standard way to do it so all computer types could be used? What are the safety issues, if any involved? Would it require FAA authorization? Are there enough customers that buy devices with a Continuum-like technology to justify all of this?</p><p><br></p><p>Anyone can have a vision, but it's the analysis and grudge work that bring a vision to the real world.</p>
skane2600
<blockquote><a href="#122653"><em>In reply to VancouverNinja:</em></a></blockquote><p>Like a lot of MS customers, I could care less about Snapchat, Soundcloud or whatever replaces them next week. I probably haven't used iTunes in 5 years. Obviously some people care about these apps, but others won't consider the Windows Store to be "ready for prime-time" on the basis of these type of apps.</p>
skane2600
<blockquote><a href="#122501"><em>In reply to MikeGalos:</em></a></blockquote><p>There's is no general solution to "adapting to the interaction model at the moment". Yes, in many cases you can constrain your programs in such a way to maximize their ability to run on different platforms and form-factors, but in general you can't design your program without regard to multiple form-factors and have it adapt automatically no matter what OS you are designing for.</p>
skane2600
<blockquote><a href="#122528"><em>In reply to MikeGalos:</em></a></blockquote><p>So if one designs a UWP app placing individual objects across the screen filling the available design space on a desktop machine, those objects will appear on a Windows Phone in exactly the same relative positions as they did when it was designed and any text included will be fully readable?</p>
skane2600
<blockquote><a href="#122586"><em>In reply to Darmok N Jalad:</em></a></blockquote><p>I'm not sure if what you suggest Apple might do is equivalent to what Mike is saying about MS. If Apple were to simply drop Mac OS in favor of iOS that would obviously be easier than trying to create a hybrid. But if they did attempt a hybrid they would be faced with the same sort of problems MS faces. Of course MS wants the Desktop, Phone, XBOX, Hololens, embedded etc to use the "same" OS, if Apple just wants to create a hybrid with only 3 form-factors that would be a bit easier.</p>
skane2600
<blockquote><a href="#122933"><em>In reply to Jorge Garcia:</em></a></blockquote><p>It doesn't matter what it's called. If it's optimized for tablets it will be inferior for productivity, if it's optimized for productivity it will be inferior for a tablet. This is the issue regardless if the vendor is Apple, MS, or Google.</p>
skane2600
<blockquote><a href="#123240"><em>In reply to Jorge Garcia:</em></a></blockquote><p>Yes you can take iOS, add support for windowing and mice with those peripherals a standard in a device and viola! You've recreated a MacBook with iOS!</p><p><br></p><p>Or Apple could do like MS and have a device with an iOS tablet mode and a conventional MacOS mode. </p><p><br></p><p>Or they could just do the simplest thing: Let workstations be workstations and consumption devices be consumption devices.</p>
skane2600
<blockquote><a href="#122503"><em>In reply to Waethorn:</em></a></blockquote><p>It's nearly impossible to create an RTOS that runs on a processor that has an internal cache or uses pipelining. Each system call has to complete in precisely the same amount of time each time it is called regardless of the state of the processor at the moment of the call. In any case, an RTOS is important only in systems that have real-time requirements. If it's used for a general purpose computer, it's just over-engineering.</p>
skane2600
<blockquote><a href="#122557"><em>In reply to Jeff Jones:</em></a></blockquote><p>Using a separate core doesn't solve the pipelining and caching issues. If you really want to create an RTOS, you have to choose a real-time appropriate CPU for it to run on. Typically this ends up being a microcontroller.</p>
skane2600
<blockquote><a href="#122589"><em>In reply to Waethorn:</em></a></blockquote><p>While there may be orgs that claim their OS is real-time, if their system calls don't have deterministic execution time, they simply aren't. </p><p><br></p><p>Windows CE docs say:</p><p>"Guaranteed upper bound on delay in executing high-priority interrupt service routines (ISRs). The kernel has a few places where interrupts are turned off for a short, bounded time."</p><p><br></p><p>That's rather vague, but it's the only real-time accommodation I could find in the documentation (perhaps there is more elsewhere?). Nothing about deterministic system calls. </p><p><br></p>
skane2600
<blockquote><a href="#122590"><em>In reply to crfonseca:</em></a></blockquote><p>You're right, there's no reason. Everything that we currently do on a phone doesn't require a RTOS. </p>
Lateef Alabi-Oki
<p>I never saw the point of Andromeda when Android apps already work very well on Chrome OS today. </p><p><br></p><p>Google's next focus should be getting developers to make their apps compatible with Chrome OS so that they look decent on tablets and laptops. </p><p><br></p><p>I believe Chrome OS will eventually become the base OS for Google's hardware efforts. </p><p><br></p><p>They already use it in Google Home, On Hub, Google WiFi, Chromecast and of course laptops. </p><p><br></p><p>And now that Chrome OS can run Android apps natively, I see no reason why their phones won't run it in the future. </p>
skane2600
<blockquote><a href="#122940"><em>In reply to Roger Ramjet:</em></a></blockquote><p>"Very Early" Google had no software to sell or even give away. They introduced a novel search method at the right time and through ads made tons of money. Without a commercial Internet Google wouldn't exist. There was no option for ad-based revenue when MS started. Even today, it's not a model that many companies can emulate due to the power of established players. </p><p><br></p><p>The cost of a marginal copy of software has always been close to zero relative to development costs and yet companies have made billions of dollars selling software to customers. So obviously your "profit maximizing price is zero" theory doesn't hold up. Imagine the millions of dollars Autodesk would lose if autocad was free.</p>
skane2600
<blockquote><a href="#122928"><em>In reply to Jorge Garcia:</em></a></blockquote><p>"Had Google literally made an 80% clone of windows, that ran Android under the hood, .."</p><p><br></p><p>Or Google could just wave its magic wand …</p><p><br></p><p>Seriously, have you ever worked on a complex software product? </p>