In the Depths of the Cloud, Open Source and Proprietary Leviathans Fight to the Death
Jono Bacon Asked Google Home ‘Who Founded Linux?’ You Won’t Believe What Happened Next!
Red Hat's Women in Open Source Award Winners, 2017
Imagine an Android Phone Without Linux Inside
Linus Torvalds Talks to Debian Users
Mozilla Relents, Thunderbird Can Stay
Heed the Prophet Stallman, oh Software Sinners!
July 4th, 2016

Needed: A Linux Three in One Distro

If FOSS is to have a future, we must embrace both mobile and the Chromebook model and develop a distro that’s equally at home on a phone, a low resource cloud based computer and on a traditional PC.

Op-ed

When Linus Torvalds started work on Linux, his purpose wasn’t to reinvent the operating system. Just the opposite. His purpose was to build an operating system that was a lot like the already existing Unix. In other words, he embraced what was already being used.

Linux desktop, Chromebook, mobile distroAs development continued, refinements were naturally added that didn’t exist in other operating systems, many of which eventually ended up in other *nixes and even Windows, just as many new additions to Unix also ended up in the Linux kernel. But the original purpose was simply to build on what had gone before, not to create something radically different.

Likewise, when Richard Stallman and his friends started the GNU project, the vision wasn’t necessarily to reinvent computing, but to make sure that access to computer technology wasn’t solely in the hands of proprietary vendors. As with Linux, as development progressed, enhancements were added to components that were new and not necessarily available elsewhere, but that wasn’t the purpose. The purpose was to build a desktop computing environment that was free and open.

So far, that’s worked out well. Because of the work of the Linux developers and the GNU team, I now use a free and open desktop system that in most ways surpasses anything the proprietary folks offer. This has benefited the enterprise too, with data centers across the world filled with Linux serverss, not solely because Linux is free to use, but because in most cases it gets the job done better and cheaper than proprietary server operating systems.

Things are changing, however. We’re witnessing a paradigm shift in the way that people and companies are using computers. As we saw with medicine in the last half of the 20th century, computers are becoming more specialized.

This began with the coming of smart mobile devices. For some uses, phones and tablets now offers a much better way to perform tasks. I saw this the other day at a Target store when I asked an employee on the floor if they carried a particular product. Instead of walking me to a workstation to check inventory, or to the proper shelf location to see what we could find, she whipped out a mobile device that had access to that particular store’s inventory. Not only could she tell me they were out of stock, but that the store across town still had three remaining.

There’s also the move to the cloud. Both businesses and individuals are finding it more convenient and cheaper to not only store most of their data on the Internet where it’s available anytime and anywhere, but to access applications online and work through a browser, leaving most of the heavy lifting to be done by the server. Because cloud storage reduces the need for local storage, and web apps take much of the load off the client CPU, this shift has given rise to the low resource Chromebook.

These shifts do not negate the need for traditional desktops and workstations. But the traditional PC with all of its power and complexity is increasingly becoming overkill for most consumer and commercial needs.

Most free and open source software advocates are resisting these changes, especially the move to the cloud. We like to have control of our computing environment and especially balk at trusting all of our data to an online service. However, by ignoring the Chromebook phenomena, and its tie-in to mobile, we are assuring that free tech rapidly becomes a dinosaur and just another footnote in technology history books.

Mobile is almost completely proprietary. Sure, Android may run on the Linux kernel, but that’s about its only free aspect. It’s not developed in the open, the user community has absolutely no say in its development, and as Steven J. Vauhan-Nichols pointed-out recently, it’s made for passively consuming data. While there are totally free versions of Anroid available, jailbreaking a phone or tablet to install them requires a degree of technical competence that is beyond nearly all mobile users.

To a lesser degree, the same is true of Chromebooks. Besides having “Linux inside,” they’re otherwise only free in a kinda/sorta way, as they come out-of-the-box completely tied to Google’s ecosystem. And while Google’s cloud services aren’t bad as far as cloud services go, they’re still proprietary, with all that implies.

However, as much as most of us don’t like these newfangled devices, we must embrace them if we are to have any hope of keeping tech free.

What we need is a single free and open GNU/Linux operating system that works equally as well on mobile devices, web-centric laptops and desktops such as Chromebooks, and on traditional PCs. An operating system that can sip resources when installed on phones, tablets and chromebooks, but which will be able to take advantage of the greater resources available on traditional machines. When in mobile or Chromebook mode, it must be able to take advantage of Android apps and it must be easy for a desktop user to switch to either of these two “lesser” modes.

This would be a daunting project and not something that could be started by a couple of coders with no resources in a basement or garage. This distro would need it’s own built-in cloud infrastructure of online apps and storage, or the ability to connect to free and open source web apps and storage. It would also need it’s own repositories, not only for desktop apps but for completely free, open source and ad-free Android apps as well. It goes without saying that the whole kit and caboodle must be licensed under the GPL.

Ubuntu’s on the right track with the convergence model, but it needs to be expanded to include something similar to a Chromebook mode as well as the cloud infrastructure.

A Ubuntu version might not be a bad idea. Canonical has the resources for development, and after it’s operational the code base could be used by others to add functionality, much as Mint and others are already doing with desktop Ubuntu.

This needs to be done. If not, FOSS is in danger of becoming meaningless.

The following two tabs change content below.
Christine Hall has been a journalist since 1971. In 2001, she began writing a weekly consumer computer column and started covering Linux and FOSS in 2002 after making the switch to GNU/Linux. Follow her on Twitter: @BrideOfLinux

Latest posts by Christine Hall (see all)

20 comments to Needed: A Linux Three in One Distro

  • tracyanne

    Ubuntu’s on the right track ….

    Yes a Chromebook style version of uBuntu’s convergence paradigm would be good. The problem is ensuring users have a secure, and private, cloud storage.

    Additionally we need FOSS tools that replace the current proprietary Cloud Offerings.

    I’ll be purchasing an Ubuntu Phone in the next few weeks, I’ll let you know what I think is missing from the FOSS stack.

  • Greg Zeng

    “equally at home on a phone, a low resource cloud based computer or on a traditional PC”

    Defining “traditional PC” is difficult now. Is it capable of playing modern games? Over the decades since 1980, my “traditional PC” has changed.

    In today’s terms, my first Pc’s were Chromebook-types, with a acoustic-modem to Fidonet. Internet was not yet created then. Today my Dell-XPS-15 lightweight notebook has “only” three terabytes of motherboard-storage, + USB3 drives of several types. But is it a “traditional PC”? Without a better GPU-memory setup, it is troubled, running powerful PC games at medium to high resolutions.

  • jymm

    I guess I am a Luddite. I prefer a separate OS for computers, books and phones. As for convergence, how did that work out for MS?

  • Mike S.

    @Christine Hall,

    I think even before convergence, two other things are more important.

    On the mobile device side, we need open source software drivers for wireless chips, cellular modems, GPS devices, and so forth. I have tremendous respect for the ambitions and goals of the Replicant project, but right now if you use Replicant you only have a few phone choices and can only use half the features on those phones.

    On the cloud side, we need things like:
    – Yacy (distributed search to replace Google and Bing)
    – Storj (distributed user owned encrypted data storage to replace Dropbox and Google Drive)
    – Equivalents to Facebook, Twitter, Instagram, and all products like them that are run in a distributed, decentralized way. The Ethereum and Safecoin distributed digital currencies are both designed to try to provide a kind of distributed pool of general purpose computing resources, and maybe someone can figure out how to build a distributed social network on top of one of them.

  • Mike

    > “On the mobile device side, we need open source software drivers for wireless chips, cellular modems, GPS devices, and so forth. I have tremendous respect for the ambitions and goals of the Replicant project, but right now if you use Replicant you only have a few phone choices and can only use half the features on those phones.”

    This more than anything. Phones are nasty, closed devices.

    One thing I’ve often wondered, but never investigated, is how phone and tablet manufacturers keep the drivers closed while building them into the GPL licensed linux kernel. Surely they are ALL in violation of the GPL?

  • tracyanne

    @Mike I believe they aren’t because they don’t actually build their proprietary code into the kernel. They provide a stub (More or less an API) that is GPL compliant, which connects to their proprietary code, which is supplied separate from the kernel, and runs outside the kernel.

  • Warren

    Microsoft tried this and we got Windows 8 (which sucked) and then tried to back only a little bit away from this and we got Windows 8.1 (which sucked less but still sucked).

    Just having it open source won’t make it stop sucking. The apps for Desktop and apps for a tablet and apps for a phone are not going to be the same apps. There is no point.

    Imagine someone starts making a FOSS linux powered phone. They will be destroyed by patent shills, chief among them Microsoft and Oracle. As bad as the Google non-free non-open elements of Android are, Android is only as open as it is and therefore good for the OSS fans, because it has a big bear like Google behind it.

  • @Warren The comparison with Windows 8 doesn’t work. Windows 8 forced the mobile desktop environment onto the desktop user with the interface formally known as metro.

    As for a FOSS phone not being able to survive in a world of software patents and monopolists…how about Ubuntu phone? And Firefox OS (or whatever they called it) didn’t fail because of lawsuits.

  • tracyanne

    @Warren, what Christine said plus Ubuntu phone presents a fully Phone compatible UI… on a phone. When the phone (the High end version so far) is connected to a Monitor, keyboard and mouse, it presents a standard Ubuntu/Unity PC style desktop, and there is no reason why it could not present any other style of FOSS PC style desktop.

    Currently also the Ubuntu 10″ Tablet presents a standard tablet style desktop, when in tablet mode. In PC mode… when a mouse and keyboard are connected it presents a standard Ubuntu/Unity desktop… once again there is no reason why any other FOSS PC style desktop cannot be presented when in PC mode.

    As for applications Ubuntu has their API and design parameters for a phone app that morphs with the desktop change from Phone UI to Tablet UI to PC UI. In addition there is no reason why many FOSS desktop applications could not use this API, to become “Convergent”.

    Microsoft’s Phone apps do not compare with what Canonical have done here.

  • Jim L.

    This sounds suspiciously similar to the ‘we all need to embrace one OS and leave multiple distros behind!’ crap we see surface from time to time. The sheer nature of open source means that will never happen, nor should be want it too. Even if that is just my perception and the true point of the article is the ability to scale multiple platforms, I’m afraid this would seem to be something many people are already working on.

    I’m typing this from my convertible laptop in tablet mode running Debian Testing and Gnome 3.20. Gnome already works quite well on tablets/laptops that can act as tablets, and although there is plenty of room for improvement, it is improving all the time. KDE too is apparently working on scaling all the way from mobile to desktop. And there is of course Unity, already making phone to PC convergence work even as they (Canonical) suffer from NIH symdrome and produce things like mir and snaps.

    That’s making the assumption FOSS even needs a single OS or interface to span multiple platforms. Currently, even without Ubuntu convergence the vast majority of phones run a mostly open source OS, and while it is tied to Google, let’s not forget Google, while a corporation driven by profits, has opened sourced most of Android, Chrome, Chrome OS, etc. and is a top contributor to the Linux kernel. Fully open forks are available with strong user, and more importantly, strong developer communities behind them. Companies like Google use FOSS and open source key parts of their software because the open source model works, and that’s not going to change as less people use desktop towers.

    What we need isn’t “a single free and open GNU/Linux operating system that works equally as well on mobile devices, web-centric laptops and desktops such as Chromebooks, and on traditional PCs.” What we need is people to continue and expand the good work they’re doing contributing to open source projects, from making sure there are open source alternatives to android, to improving touch support in the major DEs, to fighting MS whenever they try and lock down consumer PCs. FOSS is to a greater extant than microsoft ever will be anything already present in every level of computing technology we use, from web servers to phones. Why this all needs to be tied in a single distribution, I don’t know, and luckily the nature of FOSS is that it can’t be tied in a ‘single’ distro. We don’t need a single thing that works ‘equally as well’ on multiple platforms. Remember a core part of the Unix philosophy: Do one thing and do it well. Don’t try and be everything at once.

  • Somewhat Reticent

    This will make sense when people do the same tasks on all 3 kinds of devices – which they never will …

  • Mike S.

    With respect to ‘convergence’, I think it’s important to remember two things:

    1. The average smart phone today is more powerful than a PC from 1999, and mobile apps are engineered to make better use of its hardware than typical desktop apps of that period. 17 years ago if you ran too many programs, your operating system crashed or used swap files and slowed to a crawl – even Linux would slow. Mobile apps support auto-suspend and auto-kill to make sure memory is managed properly. So the average computer user that doesn’t play advanced graphics games or edit videos really can use their phone for all of their computing.

    2. All of the attempts at convergence so far were half-baked. Windows 8 was a usability disaster. Firefox OS was a nice concept, but I maintain it failed because it wasn’t stable and didn’t have enough features when it went to market.

    The important thing about convergence on Ubuntu Touch, KDE Plasma, or GNOME is that it’s evolving slowly and nobody is forcing it on users. So maybe 99.9% of the computer-using population doesn’t care this year, or next year, or even in 2020. But in 2025 or 2030, we’ll all be using it and taking it for granted because the kinks will be ironed out.

  • Roland

    Didn’t Steve Ballmer have this idea a few years ago? “One OS” results in the equivalent of a Swiss Army knife: sometimes handy, but useless for doing any real work. Stop chasing “market share”, unless you’re a corporation.

  • “This sounds suspiciously similar to the ‘we all need to embrace one OS and leave multiple distros behind!’ crap we see surface from time to time.”

    I am not entirely sure that is what it sounds like. OS (or a platform) is not the same as a distro. For example, Debian is the core for the largest number of derivative distributions, but those derivatives are still Linux.

    Sure, if we said “we all need to embrace vanilla Debian on everything, no derivatives”, that would be different…

  • Jim L.

    I could be wrong, but something with this part:

    “What we need is a single free and open GNU/Linux operating system”

    That runs me the wrong way. What ever you want to define the nuances between distribution and operating system to be many, many people are going to use them more interchangeably than the ridged definition you’re giving them (where ‘Linux’ is an OS and distributions don’t qualify as independent OSes). If I say “what OS?” And someone says “Arch” that’s an equally if not more acceptable anwser than “Linux” and that’s how people use those terms colloquially.

  • Andy Cater

    Debian everywhere 🙂 – not Ubuntu because too many commercial pressures conflicting

    It’s already too late, potentially, for FLOSS to build cloud and infrastructure for the Chromebook revolution: Amazon, Google and Microsoft have done that and there isn’t a huge player that can break into that now.

    We could persuade Amazon / Google to be more FLOSS like and open if there were a compelling commercial reason for them to do it – but there isn’t – they can use FLOSS internally for free and do

  • Randal

    I am not the target audience for this. I still don’t have or use a “smart” phone. (use of prepaid, costs less then $100 a year) I also am not a big follower of volunteering my information to the cloud for commercial harvesting/targeting of me, while at the same time giving the NSA a central access point to personal data.
    Also not sure about security or access issues. (how weak are phone lock/unlock passswords? What about stolen devices; either phone or computers as I know in several cases? Who pays for this “free” not as in software, cloud? What about the difference in storage?)
    I think there are those that would like one desktop design, across devices, yet hardware designs are far enough apart to limit the realities of this. (where are the arm based laptops for the software to be designed cross platform? What about the lower memory or drive space of a phone? etc. etc. etc.)

  • Mike S.

    @Andy Carter,
    I think our best hope – and it’s a long shot – is something like Etherium or Maidsafe. They take the idea of the Bitcoin digital currency and run with it, building a distributing computing platform on top of the same computation processes that manage transactions in the currency itself. So *in* *theory*, we could all have that software running on our gadgets and run a distributed, private, user-owned, unable-to-be-censored, computationally-infeasible-to-be-data-mined equivalent to Google search, Facebook, Twitter, Ebay, and Pinterest on top of that.

  • Mike S.

    @Randal,

    Skipping the smart phone is useful, but the problem goes deeper than that. Your credit and debit card companies share information. Most grocery stores and discount clubs like Costco have a membership card that lets them track your purchases. Browser “fingerprinting” makes it easy for third party cookie services to track you even if you run add-blockers. Unless you use a VPN, your internet service provider has a record of every website you visit. If you do use a VPN, if it’s a popular VPN provider there’s a good chance the NSA is still logging your web traffic metadata. Unless you host your own email, use PGP encryption, and only communicate with other people that host their own and use PGP, your email is collected and data mined. Even your non-smart phone supplies location data to your wireless carrier whenever it’s on, because they can triangulate your position using the phone’s communication with three nearby cell towers.

    So the price of 1950s privacy today is incredibly high: no debit or credit cards, transporting your phone with the battery removed, shopping only at flea markets and stores that allow cash-only transactions, and basically not using the internet at all. If you must use the internet, use TOR or a VPN you set up alone and reinstall your complete operating system every day or two to prevent browser fingerprint technology from building a unique profile on your browser.

    And the problem is, the number of people with the knowledge and skills and financial resources to do all of those things (like pay cash on all purchases without borrowing anything) is too small to matter.

  • Henry Kurth

    Apple already has this now and it’s using a BSD license. When Verizon starts selling Ubuntu phones, let me know, I might buy one, but until then, when my contract on my Blackberry Classic expires in April, I’m buying an Apple 7.