Announcement

Collapse
No announcement yet.

A Linux distro by Microsoft?

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    A Linux distro by Microsoft?

    No, I didn't read anything making the claim, or even predicting it. I read this article about Microsoft breaking into the list of TOP contributors to the Linux kernel.
    http://www.wired.com/wiredenterprise...gn=Interesting

    Of course, they had no choice. It was contribute or be sued for violation of the GPL. Microsoft has been getting increasingly bold about taking code from GPL projects to include in their OS. After several Windows apps and utilities were shown to contain GPL code Microsoft was forced to abide by the GPL and contribute the changes to the GPL code back to the projects they were "borrowed" from. So, either there was a LOT of "borrowed" code, or Microsoft has turned over a new leaf and decided that if they can't whip Linux then they should join Linux.

    The contributions by Microsoft has been beneficial to both sides.
    Greg Kroah-Hartman tells Wired, is that the code originally contributed by Microsoft in 2009 has now been whittled down to less than half its size. “When it first was released by Microsoft, it was about 20,000 lines of code. Now it is 7,000 lines, and supports more devices, [including] mice and newer releases of the Hyper-V system,” he says. “Merging their code into the kernel tree caused it to get smaller overall, making it easier to maintain, and have less bugs.
    “Pretty big proof that getting the code into the main kernel tree was the right thing to do.”
    The Linux kernel dev crew takes 20K lines of code from Microsoft's dev crew and whittles them down to 7K lines but also ADDS support for several new devices and newer releases of the Hyper-V system? That says volumes about the comparative skill levels of the two groups of coders.

    But, the main reason why I posted this story: I've mentioned in the past, both here and other places, that Microsoft is not seeing the forest for the trees. They laid a HUGE egg with VISTA, and by all accounts are getting ready to lay a bigger one with Win8. Their smartphone market share is dropping below 1%, making them essentially invisible in that market, even after buying/hijacking an OEM that held over 40% of that market, and turning a silk purse into a sow's ear. What if Microsoft either created its own Linux distro and then made version of its office products to run on it, or, made versions of its office products that ran on, say, Debian based distros, and got out of the OS business altogether?

    With their monopoly on the PC OEM desktop they could jam their Winux distro down OEM throats just like they jammed Windows. Imagine PCs flowing out of DELL, HP, Toshiba, ACER, ASUS, Lenovo and other major PC makers that were preloaded with "Winux". Better yet, imagine PCs flowing out of those OEMs that were pre-loaded with Debian and contained native demos of Office, Word, Excel. The user could select the DE they wanted to use and it would be installed during the customer setup: enter the name, password, timezone and DE choice. For tablets and smartphones the consumer could be offered Plasma-Active-Two. Microsoft could establish a "Debian Store"/repository and offer GPL apps, proprietary apps, sell movies and music, or charge for listening or viewing them. Microsoft does not support their OS unless given a credit card number, and most warranty problems are handled by the OEMs, so moving to Linux wouldn't change their OS support expense by much, if at all, unless people didn't mind giving MS their CC number get educated about Linux, even when such help is freely and widely available on the web. The OEM would tailor their hardware to work with Debian, writing video and audio drivers as an LGPL3 kernel module.

    What business plan could Microsoft adopt that would benefit both them and Linux?
    "A nation that is afraid to let its people judge the truth and falsehood in an open market is a nation that is afraid of its people.”
    – John F. Kennedy, February 26, 1962.

    #2
    Originally posted by GreyGeek View Post
    What business plan could Microsoft adopt that would benefit both them and Linux?
    Azure.

    Yep, you read that right. AWS is great, but they really need some competition. Azure easily matches AWS's durability and reliability. So if I need to run a collection of Linux-based server workloads, having cloud diversity is:

    * good for me (and by extension for Linux), because I can continue with my production builds and achieve even greater availabilty
    * good for Microsoft, because they can accrue revenue from me that they previously didn't

    Comment


      #3
      A monopoly on the (slowly dying) home pc desktop market? As tablets increase in numbers, there will be different demons to battle here, such as data providers and device manufacturers, perhaps?

      Comment


        #4
        Originally posted by SteveRiley View Post
        Azure.

        Yep, you read that right. AWS is great, but they really need some competition. Azure easily matches AWS's durability and reliability. So if I need to run a collection of Linux-based server workloads, having cloud diversity is:

        * good for me (and by extension for Linux), because I can continue with my production builds and achieve even greater availabilty
        * good for Microsoft, because they can accrue revenue from me that they previously didn't
        From the Wikipedia:
        The Windows Azure platform includes five services — Live Services, SQL Azure (formerly SQL Services), AppFabric (formerly .NET Services), SharePoint Services, and Dynamics CRM Services
        So, if Linux Azure was added to their services, offering a virtual Linux PaaS, you think they could attract Linux users?

        But, couldn't any cloud service began offering a Linux virtual platform? Ten years ago IBM installed over 4000 virtual SuSE Linux desktops on a single Z90 mainframe in just under 30 minutes, merely as a test. They said they could have installed 10,000 in under an hour. If some cloud service, using Linux, also installed PostgreSQL, LibreOffice, and other productivity apps, and created virtual remote desktops that only required an Internet connection to log into, there would be no need to use anything but a dumb terminal, or maybe a laptop with a 16"+ screen, dvd burner and USB memory stick, say 64GB for local backup.

        While I no longer consult, nor even program, how many home users would be willing to pay for a virtual desktop connection and per GB monthly data storage? I'm using about 100GB of my 287GB Kubuntu partition. Probably less than half of that is my actual data. So, using 50GB, and the fact that I probably spend 10-12 hours a day on my laptop, RackSpace would be willing to charge me $175/month, plus ISP charges. That would put me at $225/month, instead of $52/month using my own computer which, if I went somewhere, I'd have to take along.
        "A nation that is afraid to let its people judge the truth and falsehood in an open market is a nation that is afraid of its people.”
        – John F. Kennedy, February 26, 1962.

        Comment


          #5
          Azure has a VM role in beta now. It supports multiple guest operating systems, not just Windows. When it comes out it will compete against AWS EC2.

          Cloud isn't primarily about saving money, although if you develop appropriately you can: remember, true compute clouds don't run the meter when the VM instances aren't running.

          Where cloud shines is in availability, durability, and security. Most enterprises couldn't come close to duplicating what the big cloud providers build every day. If I were CIO of someplace, I'd put all my IT in the cloud.

          Comment


            #6
            Just saw this:
            http://it.slashdot.org/story/12/04/0...-what-can-i-do
            "A nation that is afraid to let its people judge the truth and falsehood in an open market is a nation that is afraid of its people.”
            – John F. Kennedy, February 26, 1962.

            Comment


              #7
              Please see the post below about admin passwords on cloud stuff, I don't want to "step on" GGs post.

              However, about the thread itself.

              I can see the possibility where MS would take PCLINUXOS, completely rebrand the buttons, etc. etc. to MS. and then do the "sources hijack" outlined below, and provide a "dual boot" MS 7 whatever, and.... the MS linux...

              woodsmoke

              Comment


                #8
                Originally posted by GreyGeek View Post
                The Linux kernel dev crew takes 20K lines of code from Microsoft's dev crew and whittles them down to 7K lines but also ADDS support for several new devices and newer releases of the Hyper-V system? That says volumes about the comparative skill levels of the two groups of coders.
                . . .

                What business plan could Microsoft adopt that would benefit both them and Linux?
                Could those 13K lines have been aimed more at future MS benefit? I'm just suggesting that perhaps they could have been discarded, not on their quality, but on lack of present usefulness to Linux.

                Comment


                  #9
                  Regarding the issue Jerry found... the original source material makes for an interesting read. If you're interested in the details, I'd suggest re-reading the Slashdot posting, expand all the comments, and look carefully. The original poster ignored emails from the provider, thinking they were spam. One commenter wondered why the developer would want to retain ownership of the customer's domain, an interesting question. The letter, and the accompanying new contact appointment form, appear to indicate the collapse (wholesale transfer) of a fundamentally broken process (weak validation). Actually, since we don't have any further record of follow-on communication, it's difficult to speculate more. This is quite an edge case, and highlights one of the many issues surrounding domain name hosting and ownership -- that's the real issue here, not so much whether the cloud is good or bad.

                  Regarding cloud computing in general... Yes, individual providers will have incidents. Humans aren't perfect. Traditional on-premise IT certainly isn't incident-free, either. When a cloud provider has a problem, it makes the news because, well, you know, it sells. News, by definition, is something that doesn't happen very much. Why doesn't it make the news when Joe's Auto Repair gets a virus and then cleans up? Because incidents like this happen thousands of times a day all across the world. When something bad happens at AWS (or Rackspace, for that matter), it's rare enough that it's newsworthy. Also, while the total number of incidents at a provider might appear to be high, my own observations lead me to conclude that the average number of incidents per customer is much lower than the average number of incidents per non-cloud on-premise IT shop.

                  Comment

                  Working...
                  X