The Andromeda Strain

Veteran fans of thriller author Michael Crichton may recall that his career kicked into high gear with the 1969 release of a novel entitled “The Andromeda Strain.” The book described the impact of a deadly microbe strain delivered to earth from space via a military satellite.

Next week in San Francisco, Google is heavily rumored to announce the release of a new strain of operating system codenamed “Andromeda.” The new OS is expected to combine elements of Chrome with Android. Unlike current efforts to bring support for Android apps into Chrome, however, the new Andromeda OS is expected to bring some of the desktop-like capabilities of Chrome into Android to form a super OS that could work across smartphones, tablets, and notebook-style form factors.

Though details remain sketchy, the new OS is expected to offer true multi-modal windowing, as well as a file system and other typical accoutrements for a desktop-style operating system. In essence, this means that Google’s next OS—expected to be released late this year or sometime next year—will be able to compete directly with Microsoft Windows and MacOS X.

On many levels, the development of a single Google OS is an obvious one. In fact, I (and many others) thought it was something they needed to do a long time ago. Despite that, its impact is bound to be profound, and cause a fair amount of stress and, yes, strain, for users, device makers and developers alike.

For consumers and other end users, Andromeda will first appear as yet another OS option, because Google isn’t likely to immediately drop standalone versions of Android or Chrome OS after the announcement or release of Andromeda. Over time, as the transition to Andromeda is complete, those potential concerns will fade away, and consumers, in theory at least, should get a consistent experience across devices of all shapes and sizes. This would be a clear benefit for users, because they should have access to a single set of applications, consistent access to all their data, and all the other obvious benefits of combining two choices into one.

At the same time, however, the transition could end up taking several years, which is bound to cause confusion and concern for end users. Trying to choose which devices and operating systems to use, particularly as device lifetimes lengthen, could prove to be frustrating. Plus, if Google does move away from Chrome, as some have suggested, existing Chromebooks become relatively useless.

For device makers, Andromeda could represent an exciting new opportunity to sell new form factors, such as clamshell, convertible, or detachable notebooks running the new OS. They may also be able to create true “pocket computers” that come in a smartphone form factor, but offer support for desktop monitors and other peripherals, similar to Microsoft’s Continuum feature for Windows 10 Mobile.[pullquote]The launch of a new OS from a major industry player is always fraught with potential concerns, but the merger of two existing options (including the most widely used OS in the world) into a single new one heightens those concerns exponentially. [/pullquote]

Initially, however, Andromeda is going to be more of a challenge for device makers because of their need to deal with product categories like Chromebooks, that could potentially go away. Plus, like Microsoft, Google seems to be moving aggressively towards doing its own branded hardware products, and that takes away potential market opportunities for some of its partners. At the same time, the launch of a new OS with new capabilities and new hardware requirements seems like the perfect time for Google to make a serious play into their own branded devices.

For developers, Andromeda will undoubtedly prove to be a strain for a longer period of time because of their likely need to rewrite or at least rework their applications to take full advantage of the new features and capabilities that will inevitably come with a new OS. Plus, any confusion that consumers face about which version of the different Google OS’s to use will negatively impact future app sales and, potentially, development.

The launch of a new OS from a major industry player is always fraught with potential concerns, but the merger of two existing options (including the most widely used OS in the world) into a single new one heightens those concerns exponentially. As with Mr. Crichton’s book, the initial drama and tension are bound to be high, but eventually, I think we’ll see a positive ending.

Published by

Bob O'Donnell

Bob O’Donnell is the president and chief analyst of TECHnalysis Research, LLC a technology consulting and market research firm that provides strategic consulting and market research services to the technology industry and professional financial community. You can follow him on Twitter @bobodtech.

761 thoughts on “The Andromeda Strain”

  1. “if Google does move away from Chrome, as some have suggested, existing Chromebooks become relatively useless.”

    Chromebooks support OTA updating don’t they? What’s to stop Google from upgrading all of them to Andromeda?

    1. Supposedly Andromeda will have higher hardware requirements than Chrome, which could prevent that from happening.

  2. 1. Andromeda – combined OS (elements of Chrome and Android) —> Does this mean ’touch-screen’ similar to Microsofts approach?

    2. OS offers multi-modal, file system, and other desktop os features. Same approach as MS. Apple is bringing iOS features to MacOS. Will they end up in same place, or are there fundamental differences that will appear later?

    3. True ‘pocket computers’ in smartphone form factor… Will this satisfy the crowd of ‘I do serious computer work… I do word-processing, spreadsheets, compile code, presentation slides…’ ?

    4. If Apple is doing custom chips and integrating their iOS/MacOS and apps are talking to each other across devices. Is Apple then forcing Google and Microsoft to follow their lead, but with a different approach?

    1. re: 4-, MS has been converging Mobile and Desktop into a single ecosystem for years. Early on it was, ridiculously, by using their Desktop UI on phones (I still remember having to click the little red cross in a windows’ upper right corner to close apps on my HTC HD2). With the current Modern UI, I think they’ve got it mostly right, and, unsurprisingly, it is with a Mobile-ish UI on the desktop, not the other way around.
      Apple is still mostly only passing messages/data between different ecosystems, and porting a few apps. That works for a lot of use cases, but it’s not true convergence as in write once, run anywhere, and the Mobiles vs Desktop UIs are markedly different.

      Anyhow, my take is that MS is leading the “convergence” way, Apple is doing its own thing, and if anything Google is following MS. MS is still trying to stay alive in Mobile by leveraging their Desktop dominance, Google will try to invade Desktop from their Mobile base. The 2 should sharply meet in the middle ^^

      1. ” MS is still trying to stay alive in Mobile by leveraging their Desktop dominance, Google will try to invade Desktop from their Mobile base.”

        Do you see Google having an advantage? It seems that mobile computing has all the momentum. Also, I didn’t see this in the article, do you suspect Andromeda will more mobile OS like than desktop OS?

        1. Google has both advantages and issues. If work at the periphery of Google’s (RemixOS, Samsung customizations, plus another one from Rockchip whose name I forget) it’ll be 2 smilar-but-different UIs w/ app compatibility (broadly, 3 third UI scenario added to the Phone and Tablet ones), very similar to how Windows 8.1/Metro and its apps behave on Desktop vs Mobile.

          On the plus side for Google:
          – Legacy Desktop OSes are wildly over-serving most users. Most people are actually confused by overlapping windows.
          – Using the same UI+ecosystem on Mobile and Desktop makes a lot of sense. Half the learning, no messing with conversions and different apps…
          – It’s easier for Android to become nice to use on the Desktop (I’m doing it, it’s a hair from being Good, it’s just OK for now) than it is for MS to conjure up Metro apps; an Apple have ruled themselves out.

          On the minus side:
          – Updates !
          – Many people need or want legacy apps. Especially corps, but even individuals stick at home with what they use at work.

          Not a factor:
          – the toaster-fridge idiocy that even Apple is walking back on.

          1. “an Apple have ruled themselves out.”

            Not really. Nothing stopping from Apple coming out with a line of laptops (for argument sake Apple Book) based on their own custom SoC and running a flavor of iOS that’s optimized for KB / mouse use, just like watchOS & tvOS are a flavor of iOS optimized for their specific use case.

            And I actually think they will do that.

          2. It would be funny given their previous statements, and might be a bit harder than for Android, which has always supported kb+ms, hence the apps mostly do too (to different levels, for example kb shortcuts are possible but not mandatory)

            I’m fairly sure they could do the work and get most 3rd-party devs to do the work too; why they don’t is a small wonder… afraid of cannibalizing Macs ? Market too small ? Costs too similar to Intel+MacOS hence not worth the trouble ?

          3. Interesting thing is, it isn’t like Apple hasn’t been looking at doing something similar. Here’s an Apple patent from April 2013:


            If I didn’t know any better that looks almost like a Surface Book, which I think represents what the future of the laptop what ultimately gravitate towards (and by that I mean a device that can switch between touch UI & KB / mouse UI). I don’t see much room for pure tablets simply due to cannibalization by large-screen phones.

          4. I’m always wondering if those patents have any meaning at all business-wise, or if they’re just, given the very sorry state of the US patent system, a random defensive maneuver.

            In this case, the document seems indeed defensive: it describes a screen that plugs into a keyboard base, that might be touch-capable, that might get power and data wirelessly (which is a bit funny considering a) iPhones still don’t do that, and b) in this case there is a physical connection, going wireless is… puzzling). It basically describes anything from a tablet + keyboard to a Surface Book and more.

            Anyhow, the truth is in the pudding. I’ll believe Apple is converging when they announce/release something; for now they’re maybe backpedalling a bit from their toaster-fridge pathetic cop out. I think they should, I think the case for fragmentation is very weak, mostly legacy apps by now and not at all the UI/UX; and the upside quite big (lower costs from ARM vs Intel, no doubling up on skills and apps, no data/doc format issues…). Given they recently fragmented again with tvOS, I’m doubtful.

  3. I’m curious about Andromeda, and that’ll be sated in a week.

    0- How does it do the ChromeOS+Android melding. Apps-wise, there’s about 100% overlap because anything that’s on ChromeOS is also on Android, which has a zillion more apps. The advantages of ChromeOS are a) updates b) management c) xtop UI. Why not just add that to Android especially since it’s already 3/4 of the way there ? How does it work w/o compromising ease-of use in all 3 scenarios (phone, tablet, xtop) ?

    1- I’m assuming it’s a major new OS, with features that were up to now not available to either/both Android and ChromeOS. Is it also a new ecosystem in part or whole ? Specifically, Google has been adamant that OEM/custom ROMs not break compatibility w/ the appstore. (which is a bit funny: bootleg-ish Cyanogen for example were doing their own thing supposedly independently from Google, moved to break compatibility, were quickly brought back in line… benevolent looks the other way can quickly turn into death stares ^^). Anyhow, really curious if the old apps will work well on the new OS, and/or Andromeda apps on vanilla Android/ChromeOS. Historically, issues have been about working well with “only” keyboard and mouse, handling landscape hi-def low-rez screens, and handling windows’ weird formats and resizing. That can’t be back-ported, apps have to be updated. RemixOS sidesteps the issue by only allowing legacy windows formats ie small/regular/XL phone for non-resizing-aware apps.

    2- does it fix the update issue on the Android side (it’s more than solved on ChromeOS). A lot of progress could be made by isolating the telephony stuff in a black box, and forcing carriers and OEMs to do customizations via the Appstore, thus making the OS independent of both upstream and downstream crap.

  4. Historically, Google has not been very good at managing transitions. Most of the time, new stuff is developed independently of the old stuff, and there is a conspicuous lack of a consistent strategy. This has been the case recently for messaging/video conferencing, and has long been an issue with their Android/Chrome dual OS situation. We have also seen the same with the Picasa to Google Photos transition (no migration of tags etc.)

    Therefore, although it is fun to speculate on how Google has positioned Andromeda vs. Android and ChromeOS, I think it is equally likely that Google hasn’t really given it sufficient thought.

    I may be worrying too much, but failed OS transitions can be dangerous. OS/2 and the Copland project come to mind.

    1. It must be fun to work on apps at Google, starting with a clean slate every time, and moving to a newer shinier project 6 months in. No pesky backwards compatibility, userbase transitioning, roadmap to work on and stick to… As a user, I’m getting whiplash. I was vaguely thinking of pushing people out of Skype towards Hangouts which is marginally nicer; that’s on hold now. I’m actually thinking of turning Hangouts off, its outlook (ah !) is probably dim. Makes me kind of miss the old steady MS (the new one is as bad as Google, replacing good with bad then worse, see the Windows mail client)

      Google does seem serious and consistent, maybe even forward-thinking, on a few topics. Mostly, the PlayStore and app ecosystem, and the core OS.

      1. Regarding the core OS, not being able to make up their mind on which OS to use on the Pixel C until the last minute, does not exactly suggest consistency in their strategy.

        I maintain that Google is a Spagetti thrower kind of company, and should be analysed with this in mind.

        1. The “Pixel OS – random – last minute” thing is an urban myth though. Some commenter starting saying it, mainly because it’s an Android device with an historically ChromeOS name and the OS update procedure is ChromeOS’s superior one, not Android’s; and people just accepted it and are now parroting it.

          But there’s 0 corroboration from anywhere that Google had issues choosing an OS for the Pixel, or that it did so at the last minute.

          On the contrary
          1- it seems the new phones will be called Pixel too, which makes the whole thing rather coherent in fact, the tablet was just the first in the new series.
          2- the dual-OS ROM for quicker/safer updates has since been launched on phones too.
          So the 2 clues that led someone to say “that was meant to be ChromeOS !” have been proven to be bunk. The weird stuff on the PIxel C was just a harbinger of things to come.

      2. Our work recently switched to hangouts from Skype. Long story short, I feel really bad for the IT department. No real good independent app, random chat frame repositioning, no unread counters, no saving spot on message threads…its really a step back in usability. But since we’re all in the google ecosystem for email, it was just easier and more secure and more controllable from an IT perspective. Artists are pissed though…. Never thought I’d say I miss Skype…

  5. What new ability will Andromeda give me that current OSs don’t. That’s what I care about, and what I’m most curious about.
    If I don’t fully control it, that’s mainframe and I’m uninterested.

  6. > I think we’ll see a positive ending.

    Whether it will be a positive or negative ending is remained to be seen. It’s too early to jump into a conclusion at this time.

    However, it’s interesting to see Google is using similar approach as Microsoft which is one OS to rule them all. While on the other hand, Apple is using a different approach which is one specific OS for each group of devices (namely watchOS, tvOS, iOS and macOS) with seamless integration across the devices.

    In my opinion, Apple’s approach is better. A specific OS for a specific hardware would provide more specific optimizations and better user experiences on each devices. Some features (or apps) might be duplicated among devices but the user experience on each device doesn’t have to be the same. For example, the user experience of using messaging (iMessages) on Watch / TV / iPhone / iPad/ Macbook is unique on each devices.

    1. Once again, Bob is showing his true nature, anyone but Apple, basically.

      Far better to talk up something no-one has seen than to appreciate what Apple have done.

      The iPhone 7 launch podcast was an absolute embarrassment. He had nothing positive to say. Quite why he was involved I have no idea.

    2. I’m not sure we need distinct ecosystems for distinct form factors. That certainly was the case before, and we especially needed to put paid to MS’s attempts to use Windows’s Desktop UI in Mobile, but
      0- the Mobile UIs are here, they’re good on all 3 sides of the coin (yes, Windows is the edge ^^)
      1- Mobile ushered an area of UIs decoupled from the application logic simply because apps have to handle several form factors, from tiny phone to large tablet with several sizes in between. “Desktop” is just one more UI, not earthshaking. Touches map straightforwardly to a mouse, especially a touch-mouse which could be a requirement.
      2- there’s no conflict between the various means of interaction or presentation (touch vs mouse, landscape vs portrait, full-screen vs windowed or split, …)
      3- that’s true for the all-important OS shell too. On the Android side, there are specific Launchers for Tv-boxes (optimized for a remote), watches (optimized for tiny), laptops (optimized for… looking Windows-ish), game consoles (optimized for gamepad), old geezers (optimized for my dad’s failing ayes and fat fingers, I might switch to that soon ^^),…
      4- Apps can filter out use cases they really can’t handle (refuse to install on a TV box, on too small or too large a screen, on something w/o touch…)

      Sure, handling different layouts/situations is more work for developpers, and fragmenting the ecosystem is a way to make sure they do that work. It seems a bit extreme though; and limiting since many people are happy with imperfect as long as it is Good Enough.

      1. Well, it’s your opinion, I respect that. But, I don’t have to agree with you. And it seems Apple is also disagree with you. While you may argue to defend your opinion, Apple has been executing its opinion into real successful products. 🙂

        0- Of course mobile UIs are here. Hence the developers need to embrace a new kind of interaction with new suitable and appropriate UI system.

        1- Nope, touches don’t map straight forwardly to mouse, and desktop is not just another UI. You can’t simply put WIMP UI into mobile. UI is not just about how it looks, it’s about how it interacts with the users.

        2- It’s not about conflict, it’s about better user experience.

        3- All of that on a single OS? How bloated the OS would be?

        4- Or filter them at the OS level so the app developers knows since the beginning of the development to meet the standard user experience on each devices. Take a look at an Android tablet app that simply put the phone app to run on a large screen. Or have you tried to interact with an old Windows app (with WIMP UI) on a MS Surface? It’s terrible. You simply can’t do that on Apple platforms.

        One way or the other, app developers still need to provide the best user experience on each devices. It can be done using Apple’s approach or Google’s/Microsoft’s approach. Each approach has its own consequences, advantages, and disadvantages. In my opinion, Apple’s approach is better.

        1. 4- yes it’s a bit bad. But less bad than not having the app you want/need even with a clunky UI. And, on my Android+Windows tablets, I can hook up a keyboard and mouse and feel right at home using old Windows apps, which I do occasionally for emergencies, or when I don’t feel like lugging a laptop.

          3- not bloated at all, Android is modular so they put in whichever launcher is appropriate for the device/use case besides the ubiquitous default one. the Launcher is just another app in Androidland.

          2- yes, but since there’s no conflict, one user experience doesn’t impair the other one(s). There’s no compromise involved in making an app that runs well on both tablets and phones, you just do 2 UIs; there’s no more compromise involved in adding desktop to the mix.

          1- granted, multitouch maps fully only to a multitouch mouse, which is a possibility. Single touch + zoom suffices for a whole lot of apps though.

          0- Indeed. They don’t have to sacrifice old UIs on the altar of new UIs though. Apps have several UIs nowadays.

          1. 0- Yes, of course. Who said an app is only capable to have one UI? It’s out of my point. My point is an app on a specific device should be able to deliver the best UI and UX for the device. A developer who put an old app with WIMP UI on a touch device is lazy.

            1- As I said, better UX.

            2- Inappropriate UI impairs the UX.

            3- I need to confirm this on Android. But from what I’ve seen on iOS, Apple seems to deliver different OS image file for each iOS devices. When I updated my iPod touch, iPhone, and iPad to iOS 9, each reported different download size. I haven’t check this on my Android devices.

            4- It’s not just a “bit”. From user’s POV, it’s a huge different. Some users even rather to not using a bad app at all and switch to another (better) app. There’s always an app for anything on Apple platforms.

          2. 0- That’s the point. A developer who doesn’t make an scenario-appropriate UI is lazy. No need to rewrite a whole app.

            1- If I’m going to be doing serious work for any length of time, to me a better UX is w/ a mouse. And a keyboard and big screen of course. My hardware and apps do support that.

            2- Yes, but inappropriate UI-1doesn’t impair appropriate UI-2.

            3- Yes of course, the low-level OS is device-specific. The Launcher isn’t part of the low-level Android OS, except there’s a bundled default one ‘coz one is needed to launch the OS..

            4- Sometimes there isn’t, especially custom or specialty legacy apps. Most people with a white-collar job around me are stuck with Windows because of that. Not only that, but the apps are not Mobile-aware (Metro, web client…). They’re not at liberty to switch to something better though. The superior UX is being able to go on a WE with only a tablet in case there’s a work emergency or they’ve got some work to polish up, even if it’s with smelly old Windows/Desktop.

          3. 0- Again, you always look at the problem from developer POV. As a user, I don’t care what the developers need to do. All I want is an app with great UX. If an app can’t provide it, I’ll be happy to switch to another better app.

            1- A classic argument. A real work doesn’t have to always involve keyboard, mouse, and big screen. If your work need them, good for you. But it doesn’t always apply to ALL other people, as though if they don’t use them means they don’t have a real work.

            2- Again, you look at from developer POV, while I look at from user POV. The U in UI/UX is “user”, in case you missed it.

            3- I thought the public/official distributed Android image file is the same for all kind of devices (for the similar processor architecture, not by form factor). I haven’t heard an official specific Android image from Google is targetted for a specific device. Any reference?

            4- I can see the case with legacy apps. I believe statistically such case is not in majority. Perhaps there is a reason why those apps are no longer developed and maintained. It’s time to move on, maybe? 🙂

          4. 0- You’re the one who brought up lazy-dev. I was focused on UI/UX.

            1- Indeed. You’re the one arguing touch-only is always a better UX than touch OR kb+ms OR gamepad OR remote. My point is: not always.

            2- How is saying that one UI doesn’t impair another UI taking the devs’ POV ? From the dev’s POV, it’s more work. From the user’s POV, it’s invisible. Hence “doesn’t impair”.

            3- ??? There’s no public/official Android image file. There can’t possibly be a single image file because drivers, ARM vs Intel vs MIPS, ARM32 vs ARM64… The marquee event when a new version of Android is released is the publication of the *source code* on github, the binaries are only for specific Nexus devices, they might be considered cannon, but they won’t work on any other device and OEMs don’t work from those, they work from the source.

            4- It’s certainly past time to move on. Hey, have you heard Cobol and Windows XP are still in business ?

      2. “I’m not sure we need distinct ecosystems for distinct form factors.”

        It isn’t so much about having distinct ecosystems as it is about having distinct UI’s tailored to each device. The underlying architecture / ecosystem, for intents and purpose, can be the same.

  7. By the time Andromeda arrives in 2017, Microsoft will have had a whopping 5 year headstart with far wider hardware support, mature dev tools, & probably a 600M+ device userbase.

    Hey Google: good luck.

  8. In fact I think it’s Apple missing its best opportunity in years, to do the same with iOS. Apple iPhone users familiar and heavily invested in its app and service ecosystem would be enthralled by a clamshell laptop and/or hybrid tablet or even small desktop system running and synchronizing the same software on the same familiar platform as their phones, neither of which has seen anything new from Apple in years. (Of course there are heavy duty productivity uses for OS X.)

    Apple iOS productivity hardware with a real keyboard and large screen would be at least as compelling conceptually, and would hit the ground running in terms of user familiarity and tried and true platform and profitable developer community, as opposed to a beta-kinked rollout of a new (and certainly interesting) Google platform.

  9. I’m missing something with this post. Your rationalizations are well structured, but sadly it’s unreliable to count on whatever strangers may believe. Please elaborate, because I find you to be a worthy writer and I hope to learn more from you!