Linux App Support is Coming to Chrome OS

Posted on May 3, 2018 by Paul Thurrott in Chrome OS, Chromebook, Dev, Linux with 123 Comments

Linux App Support is Coming to Chrome OS

Image source: Android Police

It appears that Linux app support is indeed coming to Chrome OS. The obvious question, of course, is why?

After all, Linux isn’t exactly well known for its vibrant apps ecosystem. And Chrome OS already has access to the world’s biggest apps ecosystem, Android. So what’s up?

Simple. This is a developer play.

The one thing that’s still missing from Chrome OS even with Android app supported added is developer tools. And that is the one thing that Linux can offer from an apps perspective.

This is born out by the description of this feature which, as many Android/Chrome-related blogs are starting to discover, is now available in Beta form in the Chrome OS Dev channel. “Run Linux tools, editors, and IDEs on your Chromebook,” the description notes.

For Google and its employee and third-party developers, this is the final piece of the puzzle. It is what will make the Pixelbook and other high-end Chromebooks viable as developer boxes.

Interesting stuff.

 

Join the discussion!

BECOME A THURROTT MEMBER:

Don't have a login but want to join the conversation? Become a Thurrott Premium or Basic User to participate

Register
Comments (131)

131 responses to “Linux App Support is Coming to Chrome OS”

  1. NextWithoutFor

    Bingo! And with more devs leaving their MacBooks behind for Linux PCs, this is a very interesting ... development.

    • ck

      In reply to NextWithoutFor:

      Source? At least according to the Stack Overflow surveys for the last 6 years, the use of both Mac and Linux dev environments is growing... at the expense of Windows.

      • ebraiter

        In reply to ck:

        Yup. In the past year according to NetMarketShare in the past year, Linux went from 2.24% to 2.29%. ChromeOS from 0.29% to 0.31% mcOS from 8.23% to 8.69%.

        At this rate, maybe one of them will take over Windows in say 120 years.

        • curtisspendlove

          In reply to ebraiter:

          “At this rate, maybe one of them will take over Windows in say 120 years.”


          While MS is certainly not in trouble in any way, Windows is not as infallible as many seem to think.


          I dont see most “normal” people using Windows (or for that matter, any traditional “desktop OS environment”.


          Windows, macOS, Linux Desktop...all will be “legacy” systems.

          • skane2600

            In reply to curtisspendlove:

            I don't think anyone is claiming that Windows is "infallible", but to suggest that "normal" people don't use Windows is embarrassingly ridiculous. There are 400 million PCs running Windows 10 alone and given enterprises reluctance to upgrade suggests that a significant percentage of those 400 million are owned by everyday people.


            Of course there are many people who use Macs and to a degree Linux desktops.


            Edit: OK you did say "most" which mitigates my criticism somewhat.

            • curtisspendlove

              In reply to skane2600:

              “...but to suggest that "normal" people don't use Windows is embarrassingly ridiculous.”


              That is what I get for posting late at night.


              I meant to say, within the next decade, most people won’t be using Windows.


              I continue to see this attitude of “I don’t need a traditional PC anymore” spread.


              And for the most part it works. Not for me, not for you, but we aren’t “normal”.


              I also hear from a lot of people who keep the PC around for extreme edge cases they can’t yet do on a phone or tablet. Or something very specific that is just currently easier.


              These problems won’t exist forever. For instance, I was pleasantly surprised I was able to set my father’s printer up solely via its touchscreen and his android tablet.


              No PC necessary.

              • hrlngrv

                In reply to curtisspendlove:

                . . . No PC necessary.

                Especially if phones could be docked as needed to large external monitors, keyboards and mice. Not to run software locally on phones but for such systems to serve as terminals connected to servers running the desktop-like software users may need on occasion.

                As I see it, it's running complex applications locally which won't be viable by the end of this next decade. It'll still be available for those with poor network connections or rigid regulatory environments, but most people and nearly all noncommercial use will shift to servers except for lightweight apps.

                • skane2600

                  In reply to hrlngrv:

                  Except the superfluous device in this scenario is the phone. Just incorporate a processor in the monitor (becoming a terminal) and you're done. But the real problem is that you're not going to find these "terminals" everywhere so they aren't going to solve the computing "on the go" problem the way laptops do and if they are in the office, they just become another tethered solution -all the bulk of a desktop PC with more restricted functionality.

                • hrlngrv

                  In reply to skane2600:

                  You certainly won't find monitors with processors many places. However, a phone with a dock which has multiple usb ports could connect monitor and input hardware to a processor with networking.

                  Laptops may be a better compromise for many between mobility and capability, but some people only want to carry a phone.

                  Anyway, my focus is on complex applications, which is seldom computing on the go. Docked phones, tablets, laptops, and desktops could all have (most already do have) clients for connecting to remote application servers. That is, for many, PCs are no longer necessary since it's no longer necessary to run complex software locally. It's the need to run such software locally which is dying, not the complex software.

                • skane2600

                  In reply to hrlngrv:

                  Your scenario involves a change to the way things are today. Having terminals available is no more of a change that what you propose. But that aspect doesn't matter because whether it's a terminal with a keyboard and mouse or with a monitor with a mouse and keyboard with the phone controlling it, there's not going to be a lot of these "stations" around in public places because there's no reason for businesses to provide them.


                  While it might be true that some people want to carry only a phone they are still going to have to connect to tethered equipment to efficiently run complex applications even if they are server based. At that point the phone becomes irrelevant. A PC will be able to support both local and remote software at a similar price point to the combination of phone, hub, mouse, keyboard and monitor. Of course at present these hub schemes are significantly more expensive than the PC solution since they require high-end smartphones.


                  IMO a expansive trend toward server-based complex application execution and phones as tethered PCs are orthogonal. What they do have in common is that the future of neither is assured.





                • hrlngrv

                  In reply to skane2600:

                  My perspective may be skewed by what I use at work and on my Chromebook.

                  Many large enterprises have been using Citrix, VMWare and perhaps a few other remote desktop clients for years, even decades, to provide workplace software to remote users.

                  Then there are services like rollApp, a subscription service which runs Linux desktop software remotely.

                  It's already here, just not universally used. Kinda like wifi 20-odd years ago when people were still using analog modems to connect to the internet.

                  The question is how often and in what circumstances do most people use complex software. If most use it only while tethered, then your orthogonality argument weakens. The main stat to find would be correlation between PC use and complex software use. I suspect there's significant positive correlation.

                  From another perspective entirely, I don't want to fill out tax forms on a phone. Some things not work-related still require large monitors. However, much of that requires only a browser to go along with the large monitor. Do I need even a mini PC to run a browser? Docking phones is a minority taste (as cell phones were in the 1980s), but they're able to do much of what most people do computing-wise.

                • skane2600

                  In reply to hrlngrv:

                  Yes, I've used Citrix and experienced its mediocre performance. That may be indirectly due to the primary philosophy behind adopting such a system - saving money. Thus the "workstations" used are quite weak and the desire to save money also extends to the capability of the servers. YMMV of course.


                  The orthogonality I was referring to was between tethered phones and server-based applications, not between tethered PCs and server-based applications. Although there's talk here (primary by you I think) about using tethered phones to run server-based applications, that's not really the way the vendors are pitching them.

                • hrlngrv

                  In reply to skane2600:

                  . . . that's not really the way the vendors are pitching them [tethered phones].

                  I realize that. I believe it's more likely docked phones would be used for remote desktops than for running desktop software locally on the phone.

                • skane2600

                  In reply to hrlngrv:

                  Most remote desktop use is in the enterprise but I don't see companies moving toward buying everyone who uses a computer their own smartphone particularly since tethering is a premium feature making those phones expensive. Much cheaper to buy a low-end PC or PC dongle if the limitations of remote desktops are acceptable.


                  I just don't understand some people's enthusiasm for running complex software on (or via) a mobile smartphone when that smartphone has to be operating in a non-mobile environment.

                • curtisspendlove

                  In reply to skane2600:

                  “I just don't understand some people's enthusiasm for running complex software on (or via) a mobile smartphone when that smartphone has to be operating in a non-mobile environment. “


                  I think it would be pretty sweet if we can get to the point of everything being “wirelessly” tethered.


                  I dream of walking into work and popping my phone down on a wireless docking pad, triggering it to connect to mouse, keyboard, and monitor, loading my “docked at work” environment, and I’m off and working.


                  Pop it on my home docking pad between 5-9pm and it loads my “side-projects” environment. Or after 9pm it just loads steam and preps Factorio for me. :)


                  I’d say the dream is more in line with reduce complexity, number of devices, and toss in a bit of machine learning for personalization.


                  That said, the logistics involved in making that sort of thing be possible and reasonable make it fairly unlikely.

    • Jack Smith

      In reply to NextWithoutFor:


      I already did. Replaced my Mac Book Pro with a Pixel book and could not be happier. We will see a lot more headlines like


      "At Apple CEO Tim Cook’s old high school, they are selling their MacBooks to buy Chromebooks"

  2. curtisspendlove

    Well...hell. This makes my pending web dev device upgrade more interesting. Now the hardware just needs to get bumped up a couple notches (primarily SSD space).

    • Jack Smith

      In reply to curtisspendlove:



      Agree. We do have five more high end Chromebooks in development to compete with the Pixel Book.


      "Here are five high-end Chromebooks in development to challenge the Pixelbook"


      Should be a lot more. I purchased a Pixel Book so I could use and been working for a while now.

      • curtisspendlove

        In reply to Jack_Smith:

        “Agree. We do have five more high end Chromebooks in development to compete with the Pixel Book.”


        Yeah. I don’t need anything *amazing* level.


        But I’m not willing to drop a grand on something I haven’t proven out.


        That said I’m also not willing to drop two grand on my usual “go-to” this year either.


        I almost pulled the the trigger on a Surface Pro LTE but I had a feeling to wait a bit.


        Im not opposed to changing my process to be a bit more cloud-focused either.


        I was was just thinking the other day it might be time to polish off vim in case I end up going mostly server-based.


        But I still don’t have 100% reliable connectivity yet. So close, but I want a local option just in case.


        :: pouts ::

  3. Alastair Cooper

    This makes ChromeOS a general purpose operating system and there are a huge number of applications on Linux. It finally makes ChromeOS a full competitor to Windows, macOS and desktop Linux.

  4. rameshthanikodi

    i'm not sure if this will change much, and I honestly doubt Google is going to let people run Firefox on ChromeOS. We'll see, lol

    • Bill Russell

      In reply to FalseAgent:

      They won't have any choice if they allow Linux programs. However, I believe that Google has learned from watching Microsoft and realizes they need Firefox, (as well as Bing), to avoid becoming too dominant. So, they pay Firefox to make Google the default search engine, keeping Mozilla well funded, and actually hope that that "puts the brakes on" Chrome's uptake a bit, which is probably beyond what they could have hoped. They don't want it to become as dominant as IE was and end up crippled by coming under regulatory scrutiny, like if Firefox went "out of business" per-se.

    • Jack Smith

      In reply to FalseAgent:

      You can already run Firefox on ChromeOS with this new functionality. But honestly it is not new. You could already use FF if you wanted using the Android FF app.

      Google does not favor their stuff. They have over 80% smart phone share with Android yet create all their services for iOS.





      • Jorge Garcia

        In reply to Jack_Smith:

        Because, as a die-hard Android fan myself, it's sad to say it, but iOS people are far more receptive/susceptible to advertisements and are likewise far quicker to open their wallets, so that's why Google pushes all their services right onto iOS immediately.

  5. Kurt Koch

    If LibreOffice runs easily and well on this, it would eliminate any need for a Windows machine for most people and small businesses.

  6. jimchamplin

    This is actually really great. I figure it won't be too hard from this point to root the OS and be able to compile code on it?


    Chrome OS/Chromium OS provides a really nice - if simplified - desktop environment that is nicely done, well-integrated with the tools that most users want most of the time, and is rather attractive. If I can install .deb packages and have access to the terminal, I'll be happy to get a Chromebook as my everyday machine.

  7. Rick Foux

    Theoretically, wouldn't this make installing Steam possible on a Chromebook? Sure, you could only play the Linux compatible games, but that still seems better than playing Android games.

  8. Daekar

    I sense a great disturbance in the Force, as if millions of users cried out in terror, and were suddenly silenced. I fear something terrible is about to happen.


    I feel like the Google Death Star is nearly completed.

  9. dcdevito

    This pretty much confirms my current machine to be my last legacy OS computer, I won't have a need for windows or macOS ever again. Interesting stuff

  10. jimchamplin

    One question left for me... will this be part of the FOSS Chromium OS, or only the preinstalled commercial version?

  11. Angusmatheson

    This sounds amazing. ChromeOS will have its current web extension type apps, all Android apps, PWAs, and all Linux apps. That seems kinda amazing. That being said - a new desktop OS is that what we need in a declining market. This allows everyone at google stop using Macs...but will normal people buy one or care? They will keep doing Facebook and Instagram on their phone.

    • Bill Russell

      In reply to Angusmatheson:

      we are past the days of the general purpose computer being the "device" that everyone had to use for the "killer app" of "the internet". MS still doesn't seem to be able to accept this even under the supposed drastic re-steering of the ship under Nadella. It was a case of being too successful and this spoiled them.


      Case in point, even years after the windows 8 "situation", my windows 10 pro workstation at work saying "your device needs to restart", and making known that there is a candy crush update via distracting pop ups.


      Companies like Google and Apple more or less understand more purpose built "niche" products are the way to go. Niche meaning more appliance like not continually beating the dead horse of finding the perfect "ONE DEVICE" for both work and "play" (I will not get into defining those terms, so do not argue that the a Surface fulfills that) in this day and age. Apple has been most understanding of this, and that is because it is the "correct" view that also happens to fit their business model of selling hardware at high margins.

      • hrlngrv

        In reply to Bill_Russell:

        I'd be a little more charitable towards MSFT. The consent decree which settled US v MSFT limited MSFT's freedom of operation for several years, and MSFT seems to have ossified into certain patterns during that time which were unhelpful when the truly nasty coincidence of the iPhone appearing just about the time the consent decree ended occurred.

        Then again there was Office 2007 and the no compromise forced march to a different UI.

        As for niche, perhaps for hardware. Not so much for OSes unless PWAs take over. If any PWA would run effectively the same under Windows and Edge, macOS and Safari, Linux or BSD and any browser, Android, iOS, and whatever else the last 1% of phones use, then OS would no longer matter except to the few who need to run certain specialized software packages.

        • Bill Russell

          In reply to hrlngrv:
          I'd be a little more charitable towards MSFT

          Yes, true. I actually for the first time in like a decade, I've started to come around a bit on MS. The Azure Sphere system is the first new-ish thing that actually makes sense from them in years (other than Azure itself), and, apparently that is because it has no "Windows" in it. Not that I am out to hate on Windows for fun, it just drives me crazy that they haven't just focused on being a professional workstation OS with all the consumer nonsense going on.

    • VancouverNinja

      In reply to Angusmatheson:

      No one is going to be using Chromebooks - after 7 years the Sun has already set on the OS. No idea why Google hasn't canned it yet - ivory tower mentality maybe?

  12. VancouverNinja

    Google is grasping at anything that will get those paper weights sold. ChromeOS has zero future. 7 years later and 1/3 or 1 percent marketshare - the definition of OS marketshare flop. I think Windows phone had a bigger market share when Microsoft closed it down. Beyond the handful of people posting on this site it has performed like a lead balloon.

    • Daniel D

      In reply to VancouverNinja: Where as Microsoft is doing cutting edge development in Windows of features we are all desperately after and bound to increase market share even further. Like a new snipping tool....


      • VancouverNinja

        In reply to Daniel_D:

        More like in the areas of AI, Mixed and Virtual Reality, 3D printing, Gaming and enhanced productivity solutions. Then we can lump on top of that the only OS system that is being unified across multiple types of computing devices e.g. HoloLens, Surface Hub, PCs, XBOX consoles, Azure Sphere, and most likely a new mobile device. Dishing on Windows 10 vs OSs that have less than 1% market share or less than 10% is a waste of time.

  13. Marciano Siniscalchi

    While Linux has made huge strides on the desktop and on laptops, it is still plagued by things like driver issues, poor font rendering, slow graphics performance, a general lack of support from vendors, and the constant need to tinker with the OS to make things work. (Don't get me wrong: I love Linux, and I'm running it on my laptop. But, I am being realistic.). By supporting Linux **apps** on ChromeOS, Google can remove most of these obstacles. So, it's clearly a win.


    BTW: for those interested, there is a somewhat hacky but solid alternative that works *right now* on any Chromebook that can run Android apps: Termux. Its main aim is somewhat different---it is meant to run Linux command-line apps, not GUI apps. However, for example, I can run a full LaTeX distribution, and a decent SciPy setup, on my Galaxy Note 8.

    • Bill Russell

      In reply to Marciano_Siniscalchi:

      As with "Android", I can say anything I want to about "Linux", the greatest thing in the world or a POS, depending on my personal bent. Obviously the point here is to have a more standardized version of "ChromeOS Linux" that would be known to be supported on its machine. Linux is a kernel, and does not render fonts, nor provide any GUI on its own. I do understand what people mean typically in the context of discussing "Linux", however. Remember though that I can have a computer with what appears to be running "Linux" (in the context we have here) but have no Linux in it and you wouldn't realize it. For example, FreeBSD running a Gnome or KDE desktop.

  14. mikeharris123

    This is great news for people looking beyond a secure environment and wanting to develop on Chrome OS. Hopefully side loading of Android apps is next outside of developer mode. Then more powerful Chromebooks with more memory that don't break the bank

    • VancouverNinja

      In reply to mikeharris123:

      Why would anyone want to program on a platform that has no marketshare? Are you referring to the K-4 market and educational developers? Seems like a waste of time investing any effort into such niche platform. Line for line of code a coder would be far more profitable on any other platform than this one.

      • dcdevito

        In reply to VancouverNinja:

        I don't think you're seeing the benefit here. No one writes app for "Chrome OS", it runs web apps, so this would allow for web devs to write for the web platform, the biggest fish in the pond so to speak. This will also allow for writing Android apps, another big fish in the pond. Google would now be giving a development option on their devices, something they couldn't offer as of yet.

        • VancouverNinja

          In reply to dcdevito:

          Well no one writes apps for Chrome OS period...why even bother? Again the money is only on two PC devices - Windows 10 and, now less than 10% but still the other option, OSX. I guess in the end it will have a tiny following of devs that want to go off the beaten path. I personally follow the money.

  15. ebraiter

    They are only adding Linux app support to ChromeOS because with the exception of maybe K-12, ChromeBricks are not selling. None of the tech people I know would bother with a waste in R&D like ChromeOS.

    Google stick with the Android OS.

  16. Waethorn

    I don't see this as being a much different take on Microsoft's "throw all the sh*t on the wall and see what sticks" when it comes to app development options. Linux has a good community. Lots of software is cross-platform, especially browsers. There's lots of IRC applications, a fair number of cross-platform office productivity suites, paint tools, and you have popular applications for Twitter, Skype, Spotify, Slack and so on.

  17. Daekar

    So... I guess the real question is, why would I run ChromeOS when I could just run my favorite flavor of Linux? I could already use all the Google docs stuff via Chromium or Firefox, or all the Office 365 stuff the same way.


    So.. what do I get out of choosing ChromeOS over Mint, Ubuntu, or ElementaryOS?

    • Skolvikings

      In reply to Daekar:

      My wife and 12 year old have been sharing a Chromebook for the last year. It's great for me! I don't have to worry about patching, they never ask questions...

    • dcdevito

      In reply to Daekar:

      Easy....

      1. You won't need to run a full distro, so use Chrome OS for most of the time
      2. Security - these apps, as with everything else on Chromebooks, will be sandboxed
      3. Cost - chromebooks are cheap


      Running Linux only is a chore

      • hrlngrv

        In reply to dcdevito:

        1 - You won't need to run a full distro . . .

        Ignores the possibility that some may want to run a full distribution.

        3 - Cost - chromebooks are cheap

        So are most low-end laptops these days. Even possible with some effort to find laptops with Linux preinstalled or with no OS preinstalled. Low cost is no longer a big distinguisher for Chromebooks.

        Your #2 is the big reason to use Chrome OS rather than Linux proper, but it's a trade-off.

  18. skane2600

    I guess it depends on what limitations, if any, are involved. If Android Studio was supported including Android device emulation, that seems to me to be the use case with the most value. I suspect that using a conventional Linux distro would still be the best way to develop Linux applications but on that point I'll defer to those who are actually developing Linux apps.

  19. davidblouin

    Interesting indeed, until Google do a WP8 by releasing Fushia that is.

    • raptor

      In reply to davidblouin:


      Android apps will run on Fuchsia. Besides, who doesn't want a real time micro kernel OS with a capabilities based security model and built using modern concepts. Fuchsia is the future.

      • davidblouin

        In reply to raptor:

        It's not a "who wants" situation but more of a "who cares" one...

      • skane2600

        In reply to raptor:

        Generally a real-time OS and a general purpose OS are designed with significantly different goals in mind. If an application uses garbage collection or the OS runs on hardware that uses a cache, you're pretty much out of a real-time environment.


        Contrary to the casual use of the term, a real-time OS isn't about speed it's about consistency. In the ideal case, each OS function call or interrupt is guaranteed to return in the same amount of time, every time. In the less than ideal case there's a guarantee that the timing will vary over a tightly-defined range. Caches and garbage collection add an uncertainty factor to the timing.


        Real-time software is on the wane because more and more real-time functions have migrated to hardware.

        • raptor

          In reply to skane2600:

          That's just not true. A real time OS will dedicate the proper CPU resources to ensure it meets its deterministic priorities regardless of whether an app is using garbage collection.

          Also, I never said a real time OS was all about speed. Of course it's about being deterministic. It's also the reason why RTOSs are used in mission critical applications. RTOS will never go away because they're absolutely critical. Fuchsia isn't just an OS for laptops and phones it's also an OS to embedded systems and I can think of a lot of embedded systems that require a RTOS such as Google's car autopilot systems or even Android automotive.


    • dcdevito

      In reply to davidblouin:

      Right, and how do you think they'll be able to port Android apps over to Fuchsia? Hmm, maybe containerization? Ya know, just like how they're doing THIS....

    • Jack Smith

      In reply to davidblouin:


      I do NOT if not efficient. There is zero proof it can be done efficiently. We already did this in 1992 and I am proud to be on comp.os.minix at the time.


      My take. We were all taught in the late 80s micro kernels are the only way and anything else was bad. It was about the time MS hired David Cutler away from Dec and what OS X became was built. Both used micro kernels.


      But we had this snot nose kid that reported to no one and was accountable to no one. So he decided to build a new kernel using a monolithic kernel. Well with loadable modules so some want to say hybrid. But really monolithic.


      That kernel today runs majority of the smart phones on the planet, run the cloud and all of the top 500 super computers on the planet. As well as run your Chromecast, Google Home, Echo, Google WiFi, Tizen, ChromeOS, Dot, Tivo and majority of other things on this planet.


      A big reason is because it is crazy efficient. Here is a great WikI on the debate back in 1992 and then again today.


      I get a charge when conventional wisdom turns out wrong.


      Search on.


      "Tanenbaum–Torvalds debate"

      • skane2600

        In reply to Jack_Smith:

        Do you really believe that Linux using a monolithic kernel is the reason for it's success? It's been about a quarter century since Linux was designed, perhaps the relative value of monolithic vs micro kernel should be revisited since we are no longer running Intel 386 processors with 1MB of RAM.

    • Jack Smith

      In reply to davidblouin:


      This will work fine with Fuchsia if they want. They are using a separate Linux kernel to support. What Google did could also be done on Windows and OS X.


      But I have major doubts Google will move to Fuchsia. They right now have the same kernel on GW, Android, ChromeOS, their cloud, Chromecast, Google Home, Internal network gear and everything else. It would seem odd to use a new kernel.


      But can not explain why they are doing Fuchsia. But also realize Google develops mostly in the open so experimenting at other companies is not in view like at Google. Would not be surprised if Apple and MS have not experimented in a similar manner and we just never knew about it.



  20. Bill Strong

    If this can run Wine, it is interesting. Also, there are a few important pieces of software that still don't work well in browsers, but are on Linux.


    Blender is one of those. Also, a lot of video/3d/movie industry tools work on Linux, with the exception of adobe software.


    And, lets not forget, virtual machine software such as VMWare, Virtual Box and Qemu all work on linux, offering the ability to run and test software on lots of platforms you might target, including Windows and OS X.

    • Jack Smith

      In reply to Bill_Strong:


      Yes Wine works fine. As well as Stream games. The only reason we still have Windows boxes is for gaming. That is the big one that will take a long time to get competitive with Windows.

    • hrlngrv

      In reply to Bill_Strong:

      . . . virtual machine software such as VMWare, Virtual Box and Qemu . . .

      All of which need kernel support, which is unlikely to be found in Chrome OS kernels which almost certainly won't be user-configurable for quite a while longer if ever.

      • maethorechannen

        In reply to hrlngrv:

        If this is supposed to be "for the developers", it'll have VM support.

        • hrlngrv

          In reply to maethorechannen:

          If this is supposed to be "for the developers", it'll have VM support.

          Maybe, but most Chrome OS users (kids in primary and secondary schools) would have no need for VM support. I suppose Chrome OS could gain the tools necessary to load/unload kernel modules, but they're not there at the moment. And if it becomes possible to load ANY kernel modules, what becomes of Chrome OS security?

          • skane2600

            In reply to hrlngrv:

            Yes, I was wondering about the security issue too. Adding Linux support seems to be the exact opposite of the philosophy underpinning the Chromebook.

            • hrlngrv

              In reply to skane2600:

              . . . the security issue . . .

              Need to delve into what distinguishes Chrome OS from Linux. First and foremost is Chrome OS's disk partitioning scheme. Chrome OS has 3 sets of separate kernel and rootfs partitions on the main drive, with only one set mounted by default and mounted read-only. The others exist to be mounted read-write when needed for updates.

              In theory, Windows could work similarly except that Chrome OS kernels fit in 16MB partitions, and Chrome OS rootfs fits in 1GB partitions. OK, also Windows currently can't have C: read-only; it'd need to be rewritten so that the registry and perhaps a lot more would be stored on another read-write volume.

              The other thing which distinguishes Chrome OS is that apps are installed on a per user account basis. There's no system-wide software installation for all users. That provides better security, but it's inefficient for multiple user account machines.

              While it's possible to mount-bind additional configuration directories under /etc, /etc/fstab is on a read-only partition, so can't easily be changed. Meaning changing configuration would require either user-initiation or automatic processing of, most likely, /usr/local/etc (not standard) in order to add kernel modules. Putting any automatic, boot-up processing under /usr/local/etc, so on a read-write partition, would compromise system security. Whoopie that Chrome OS checks its kernel and rootfs before completing start-up, but if anything could be added as kernel modules automatically, that'd mean key loggers, ransomware encryption, etc could be loaded as kernel modules.

              No doubt I lack imagination, but I don't see how anything could be added at kernel-level without compromising Chrome OS security.

        • Jack Smith

          In reply to maethorechannen:


          This uses a VM already. Then a machine container on the VM. You can then use Docker if you want or whatever as you have a second Linux kernel that is separate from the ChromeOS Linux kernel.


          But really just about anything. I have a PB which already has this functionality. So Wine works and Stream games and Docker and can not think of anything that has not at this point.


          Biggest issue is they have not yet released GPU acceleration but looking at the commits they are working on it.


  21. hrlngrv

    . . . After all, Linux isn’t exactly well known for its vibrant apps ecosystem. . . .

    Linux is the #1 OS on all device types other than microcomputers, and there's software running on all those devices.

    As for general use software, I figure there's at least 3 times as many FOSS software titles as apps in the MSFT Store. Aside from IE, Edge, and Safari, there are Linux versions of all the other top browsers.

    Besides, Chrome OS uses a Linux kernel and, in developer mode, gains the POSIX toolset already. Also, it can't run Windows software without wine or equivalent, so it's much easier for it to run Linux software than Windows software. Finally, it's not as if there were millions of productivity-focused Android apps. Yes, the label makes it appear that only developer tools would be provided. Fine. Download source code tarballs for general use software, then use make to configure, build and install it. Would the Linux subsystem which ran make, gcc, etc not be able to run the software it builds?

  22. pbsie

    You know what'll happen next? Microsoft might decide to try to one-up it... ;) ;)

    • Jack Smith

      In reply to pbsie:


      Well the latest OS released by MS is Linux based. What would be interesting if MS enabled exactly what Google has done here. They could do the same thing. Same with OS X. It is so weird that we are all mostly using the same hardware but we make the hardware different by running different OSs. How bizarre.


      Google has enabled the VM on the chip so you are running a second Linux kernel separate from the Linux kernel that is with ChromeOS. This will also make moving to Fuchsia easier if that is truly the intention. I have major doubts that will happen.


      Then on top of the VM Google is using LXC/LXD machine containers. NOT to be confused with using Docker which is application containers. Then on top you can use Docker if you want.


      This would allow instant Linux applications. So you click, container comes down, spins up, do what you need, disappears without a need for an install and should be very safe.


      You could have those exact same containers running on Windows using the same technique but need the virtual drivers and GUI forwarding built into Windows.


      Then even the same on OS X. Not saying it is "good business" for Apple and MS to enable but purely on a technical level.





      • Boris Zakharin

        In reply to Jack_Smith:

        MS already allows running Linux apps on Windows, just no bundled GUI. You can still install a 3rd party X server and then you can run GUI apps as well. The difference is Windows mostly doesn't need this because you have all the dev / power user tools you need directly on Windows. ChromeOS does not. This is why this is a big deal.

        With MacOS you already have a UNIX-like environment including native X if you want. So while it can't run Linux apps per se, it doesn't really need to.

      • jimchamplin

        In reply to Jack_Smith:

        On macOS, just install MacPorts, or compile things from source.

  23. red.radar

    i wonder how usable the android versions of office are on the chrome book....

  24. SupaPete

    If it really runs full Linux apps soon, that is indeed very powerful.

    Because while the app/game list on Linux misses many things one gets on macOS/Windows, it still of course has WAY more (as in any at all) desktop class apps and games iOS and Android don't have.

    Just as example many of my dev tools are available for Linux.

    Also more desktop games than one might think meanwhile thanks to Steam etc having a bunch at least.


    If this Linux stuff support works well, that would be a big win for chromeOS.


    I feel like Google did some clever moves there with ChromeOS/Chromebooks over time. They went from a so limited state that many people like me discredited them for anything but on the couch web browsing usage over to becoming sophisticated enough in usage, maintenance and software available that one could recommend them for using as office computer in schools and even some offices (mostly those using google apps and services and the goolge office/docs apps) and now with them getting android and linux app support and ideally progressing how well those run quickly, yeah, that's starting to become something interesting to more and more users/use cases.


    It could then essentially become the most comfortable to use linux device, too.


    As at the same time both Apple and MS are pushing to limit what one can do and run on their desktop OS more and more, well, the differences between the OS regarding apps and games regular users would run become less and less.



    As a developer, sure, i can still run way more on macOS, windows and Linux than a chromebook.

    But for an average user, they get intentionally for money lust reasons of the platform holders pushed more and more to buy all on the app store of the platform holder on both macOS and Windows, so yeah, then not that different to chrome os to many regular users anymore.


    I just made a build of a desktop game of mine recently to show to some selected people and i was surprised when i realised for the mac build they could not run it on their mac with latest macOS right away, nor in intuitive way at all.

    They had to click the app icon to bring up a message saying it's blocked and they have to go to system preferences/ security and privacy settings and there then it showed a message that one could grant an exception to run that one app not from the store.

    Apple seems to meanwhile have removed the option completely to allow apps from anywhere, in security and privacy settings it only allows to choose allow apps downloaded from

    -the app store

    and

    -app store and identified developers.

    For any other app it does not run it and one has to manually go into the settings there and click the option to manually allow running that app after having clicked the app icon so it even appears as option there in the security and privacy settings to grant an exception for.


    Not cool at all.


    So yeah, with both MS and Apple going too far like that regarding how they want to limit their regular users with the pseudo argument of security, yeah, the lines become more and more blurry whether regular users can actually still do that much more on the desktop OS of those two soon.


    A while ago on Windows there was also the big argument of best cutting edge graphics/game tech with direct x and also most best and most advanced games on Steam etc, but there, again, MS is doing their best to screw that up with their push for the store and UWP apps since when a full UWP app it can't use all the features of the OS and graphics side, and when a containerized desktop app as uwp app, that is cumbersome and convoluted to even just deploy.


    So yeah..MS and Apple really have to change their direction to not screw up any left advantages for regular users while the competition is making their offering more and more sophisticated, basically MS and Apple are playing right into Google's hand.

Leave a Reply