TOOS Stalls – GNU/Linux Thrives In North America In 2015


SeeTop 7 Desktop OSs in North America from 1 Jan to 14 Apr 2015.
I love it when, despite a preponderance of M$’s units on retail shelves of legacy PCs, GNU/Linux takes a huge share of the market. That’s happening now in North America, where I and oldfart live. People here finally know they have a choice and are making that choice more than ever before.

I guess now the trolls will boast that M$’s rate of growth for the flagship is higher than GNU/Linux on the desktop, but we know the gap on the share of page-views is barely moving. There’s no way “8.1” will ever reach 50% in the next couple of years when GNU/Linux will likely reach over 10% share and people are clinging to “7” just as they did with XP. You know what happens now? “Positive feedback”, all kinds of people and organizations springing forth to support GNU/Linux in the market: Consumers, IT people, retailers, ISVs, OEMs, … all of those folks who used to be M$’s slaves. They are free at last. Expect growth of GNU/Linux desktops to ramp up in 2015, The Year Of The GNU/Linux Desktop. Merry Christmas to all.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology and tagged , , , , , , , , . Bookmark the permalink.

11 Responses to TOOS Stalls – GNU/Linux Thrives In North America In 2015

  1. LinuxGentlesir says:

    Pro-tip. Rambling on for 5 posts in a row doesn’t make you look clever or improve the persuasiveness of your arguments. It does however make you look psycho. Less is more DrLoser.

  2. DrLoser says:

    Oh heck, why not?

    The five easy steps to creating a Linux daemon.

    Go ahead, guys … knock yourselves out!

    (And just to make this as trivial as possible … assume that all the relevant functional code of the daemon is already written. All you have to do is to turn what would otherwise be a standalone process into a daemon!)

    After this, we may proceed to atomic operations and memory barriers.

    Or perhaps not.

  3. DrLoser says:

    (For Mozilla read Bugzilla. I apologise.)

  4. DrLoser says:

    Then again, this is an interesting topic. How would you define a troll?

    Exhibit One: “A troll has no interest in, or knowledge of, the subject matter at hand.” In this case, Linux (either on the server or on the desktop).

    I’ve worked with Linux on the desktop for several years, Robert. I wasn’t entirely convinced by Linux on the server, but that’s because I am old-fashioned and come from a Unix background — thus, preferring Solaris.

    I don’t qualify as a troll on Exhibit One. Now, let’s proceed to the rather more contentious Exhibit Two, shall we?

    Exhibit Two: “A troll has less knowledge of the subject at hand than the OP.”

    Gonna be a tough one that for you, Robert. Let’s narrow it down. Do I have more knowledge than (say) you of the following elements of Linux?

    1. Debugging the network stack. Let’s make this simple: using iptables and traceroute and so on to switch between scenarios, we both have experience. I think we’re about even on this one.
    2. Using the Bash command-line. Sorry, Robert: I’ve been doing this for far longer than you have.
    3. Writing our own server daemon. Been there, done that, Robert. I can explain Steven’s “five easy steps,” if you want. You have never done it.
    4. Running a full-scale security audit over every open port and every SysV-Init script (restarts included). Done that, Robert. You? Not.
    5. Writing a substantial multi-threaded C/C++ application such as a Web-Crawler on Linux. I have done that, Robert (on CentOS, but does the Distro matter?). You have not. In fact, you are totally incapable of the task.
    6. Creating a three-step boot process to go from the kernel to a ramdisk to a fully-fledged bootable system. I have done this, Robert (admittedly with QNX, but on a Posix basis). You have not.
    7. Writing a message-passing system for embedded Linux devices (TCP and UDP and a balancing algorithm between the two). I have done this, Robert. You have not.
    8. Testing. Oh, wait, I’m talking to Robert Pogson. A man who apparently holds the sincere belief that bitching to somebody via Mozilla is somehow a valuable contribution. (It isn’t.) I have done this, Robert. You have not.

    Of the two of us, Robert, I see you as the troll and me as the trolled party. I certainly cannot see a single substantial point upon which you trump my Linux expertise and can therefore reasonably call me a troll.

    But it’s all good. Chuckle!

    😉

  5. DrLoser says:

    Yep. I’d say 90% or so of DrLoser’s comments are designed to make people look bad instead of to inform the conversation…

    As a general rule, Robert, I merely quote people here and then analyse what they say. It’s my way of “informing the conversation.”

    It’s hardly my fault if what most of your contributors say shows them up to be totally clueless (irrespective of OS preference). I can only respond to whatever dim-witted nostrum is put in front of me.

    Oh, and either way … rubbish definition of Troll. Try again.

    … like this one:“What we’re seeing here, Robert, is quite a lot of variance. But it’s variance against established market positions in the USA.”

    Considering that I quoted max/min/median for both sides, and considering that (unlike you, I might add), I didn’t cherry-pick a range or even dispute that Linux has a high-spot and M$ has a low-spot … or even try to hide those facts … I would say that I was brutally honest, Robert.

    There is quite a lot of variance there. Feel free to contradict that with statistical analysis.

    But at the end of the day, in the USA, the numbers for M$ come out at ~79% and the numbers for Linux come out at ~2%. Any normal person would look at those two base figures and notice a fairly major and medium-to-long-term discrepancy.

    You won’t accept this, I know, but leaving variance aside, that is a fairly dispassionate conclusion.

  6. DrLoser wrote, “We’re not trolls”.

    Troll: “The well-constructed troll is a post that induces lots of newbies and flamers to make themselves look even more clueless than they already do, while subtly conveying to the more savvy and experienced that it is in fact a deliberate troll.”

    Yep. I’d say 90% or so of DrLoser’s comments are designed to make people look bad instead of to inform the conversation, like this one:“What we’re seeing here, Robert, is quite a lot of variance. But it’s variance against established market positions in the USA.”

    Variance is indicated by a sequence of random values dancing around some mean. This is not what we’ve seen in 2015 in the webstats of StatCounter for GNU/Linux on the desktop. The mean has increased from ~1.7 to about ~3.3. That’s nearly doubling in a single quarter. Just look at the damned graph, for pity’s sake. Don’t you believe your own eyes?

  7. DrLoser says:

    I guess now the trolls will boast that M$’s rate of growth for the flagship is higher than GNU/Linux on the desktop, but we know the gap on the share of page-views is barely moving. There’s no way “8.1” will ever reach 50% in the next couple of years when GNU/Linux will likely reach over 10% share

    Fair enough. Let’s examine these propositions in order.

    1. We’re not trolls, Robert. Just people who disagree with you — horrid, innit? Also, each and every one of us has used Linux on the desktop for several years. Some of us are actually capable of dealing with systemd. Not much of a troll profile, then.
    2. Not sure where “boasting” comes in. But, OK, we can examine the relative movements in web-page views (on the stipulation that it means anything).
    3. “The gap in page views is barely moving.” True, but not in the way you mean, Robert. See below.
    4. “There’s no way that 8.1 will ever reach 50% in the next couple of years.” True, but who cares? 8.1 doesn’t compete against Linux. 8.1 competes against Windows 7.

    OK, a simple statistical summary for simple folk, as we find on this site. Let’s just sum up all Microsoft desktop systems and compare them to Linux. And let’s pick the three obvious measures.

    All M$: Minimum 73.41. Maximum 82.31%. Median 78.7%.
    Linux: Minimum 1.29%. Maximum 3.19%. Median 1.92%.

    M$ minimum on March 25th. M$ maximum on Jan 1st.
    Linux minimum on January 15th. Linux maximum on 11th April.

    It took a bare ten days for the M$ maximum to dip below the M$ median (77.8% on January 11th).

    It took a bare two days for the Linux maximum to dip down to the Linux median (1.93% on 13th April).

    What we’re seeing here, Robert, is quite a lot of variance. But it’s variance against established market positions in the USA.

    Linux isn’t going anywhere fast there, in terms of page views.

  8. ram says:

    They are a bit pricey, with memory and SSD they are coming in at close to 1k$ wholesale. On the other hand, they represent alot of grunt and are less costly than the so called “gamer” machines. Personally, I’m going to wait a while for the price to drop.

  9. ram wrote, “Intel has released their i7 NUC boxen. No fans, tiny, and only uses 28 Watts. It is designed for Linux.”

    I thought they were designed to maintain Intel’s lifestyle. They are much more efficient than the old ATX cases but still too expensive IMHO. The Little Woman and I have discussed paths to upgrading but haven’t made any decision yet.

  10. ram says:

    Intel has released their i7 NUC boxen. No fans, tiny, and only uses 28 Watts. It is designed for Linux. Intel even has instructions on how to make the boxen in media servers/stations — they drive 4k monitors and have 8 channel sound. This will certainly help Linux uptake worldwide.

  11. DrLoser says:

    Nothing like being whupped into the ground by Apple, is there, Robert?

    On your chosen field, OSX varies between 14% and 20%. And Linux varies between ~2% and ~3%.

    Can I interest you in a FreeBSD derivative? It has two unique selling points:

    1. It doesn’t come with systemd.
    2. It whups the ass off Linux in page views.

Leave a Reply