Linux strong #3 in OS market - Linux

This is a discussion on Linux strong #3 in OS market - Linux ; On Dec 6, 7:39 am, Linonut wrote: > * cc fired off this tart reply: > > If you want to discredit sites like the one in the OP, that's fine. > > But there are good stats out from ...

+ Reply to Thread
Page 5 of 5 FirstFirst ... 3 4 5
Results 81 to 95 of 95

Thread: Linux strong #3 in OS market

  1. Re: Linux strong #3 in OS market

    On Dec 6, 7:39 am, Linonut wrote:
    > * cc fired off this tart reply:
    > > If you want to discredit sites like the one in the OP, that's fine.
    > > But there are good stats out from sites that aren't affected by one OS
    > > or the other. As in, using Linux doesn't predispose you to going to,
    > > or avoiding, a site.

    >
    > So put up some URLs, dude.
    >



    http://www.komar.org/halloween/2004/stats/

    Note the paragraph explaining bias *towards* open source products.
    Again, a little old, but what can you do.


    http://nixtechnica.blogspot.com/2006...g-systems.html

    People read digg regardless of OS don't they? "How to host a website
    on any computer" was the website that was dugg. *Any computer.* Newest
    and best percentage I've found!



  2. Re: Linux strong #3 in OS market

    On Thu, 6 Dec 2007 14:19:38 -0800 (PST), cc
    wrote:

    >On Dec 6, 7:39 am, Linonut wrote:
    >> * cc fired off this tart reply:
    >>
    >> > Real numbers isn't the point, the stats are a good estimate, which is
    >> > what a stat is supposed to be.

    >>
    >> Show me some calculations that indicate the stats to be good estimates.
    >>
    >> Just calculate me a friggin' alpha-value for cripesake!

    >
    >I don't have access to all the data. Damn closed source statistics! Of
    >course there is the possibility that it is all some Microsoft
    >conspiracy and the numbers aren't real at all, but entirely made up.
    >But I prefer to trust people like Google and the BBC.
    >
    >
    >> > If you want to discredit sites like the one in the OP, that's fine.
    >> > But there are good stats out from sites that aren't affected by one OS
    >> > or the other. As in, using Linux doesn't predispose you to going to,
    >> > or avoiding, a site.

    >>
    >> So put up some URLs, dude.

    >
    >http://www.currybet.net/cbet_blog/20...r_agents_2.php
    >
    >BBC before this supposed Microsoft infestation and Linux lock out that
    >I see posted in random spots here sometimes. So yes, it's two years
    >old, but from an unbiased site. How would the BBC's homepage be biased
    >one way or another? "The figures may, however, mask a *slightly*
    >higher use of Linux" (emphasis mine). There's a good explanation that
    >follows.



    The main point is that no matter where you look, geek bloggers
    excepted, the numbers clearly show Linux at less than 1 percent.

    It's no secret and all the waffling in the world will not change that.

  3. Re: Linux strong #3 in OS market

    * cc fired off this tart reply:

    >> Just calculate me a friggin' alpha-value for cripesake!

    >
    > I don't have access to all the data. Damn closed source statistics! Of
    > course there is the possibility that it is all some Microsoft
    > conspiracy and the numbers aren't real at all, but entirely made up.
    > But I prefer to trust people like Google and the BBC.


    See below.

    >> > If you want to discredit sites like the one in the OP, that's fine.
    >> > But there are good stats out from sites that aren't affected by one OS
    >> > or the other. As in, using Linux doesn't predispose you to going to,
    >> > or avoiding, a site.

    >>
    >> So put up some URLs, dude.

    >
    > http://www.currybet.net/cbet_blog/20...r_agents_2.php
    >
    > BBC before this supposed Microsoft infestation and Linux lock out that
    > I see posted in random spots here sometimes. So yes, it's two years
    > old, but from an unbiased site. How would the BBC's homepage be biased
    > one way or another? "The figures may, however, mask a *slightly*
    > higher use of Linux" (emphasis mine). There's a good explanation that
    > follows.


    Well, the title is very misleading:

    Operating systems visiting the BBC homepage

    The author says /nothing/ about possible confounding factors. 0.41%?
    Maybe. Or maybe it is 10x that.

    I notice it includes OS/2 Warp, with a ridiculous number of 0.001%. So
    where are the other < 0.04% OS's?

    --
    Tux rox!

  4. Re: Linux strong #3 in OS market

    * flatfish fired off this tart reply:

    > The main point is that no matter where you look, geek bloggers
    > excepted, the numbers clearly show Linux at less than 1 percent.
    >
    > It's no secret and all the waffling in the world will not change that.


    It won't change the numbers, but neither are you sure exactly what
    scenario the numbers support.

    Hell, in my own SSH logs (before I obfuscated the port number), and when
    I exposed a web server, I would see many systematic attempts to access
    Windows-only resources and perform logins. So many that one would
    assume these accesses were automated.

    So accesses don't always mean "users using OS Foobar accesses web-site".

    Yet this obvious possibility is simply ignored in the present argument.

    That being said, I do believe that Linux usage is a small fraction of
    Windows usage.

    But so what? It is obvious that Linux is viable and used in signficant
    numbers, even if the consumer user ignores it for the most part.

    --
    Tux rox!

  5. Re: Linux strong #3 in OS market

    * cc fired off this tart reply:

    > On Dec 6, 7:39 am, Linonut wrote:
    >> * cc fired off this tart reply:
    >> > If you want to discredit sites like the one in the OP, that's fine.
    >> > But there are good stats out from sites that aren't affected by one OS
    >> > or the other. As in, using Linux doesn't predispose you to going to,
    >> > or avoiding, a site.

    >>
    >> So put up some URLs, dude.

    >
    > http://www.komar.org/halloween/2004/stats/


    6.78% Linux
    >
    > Note the paragraph explaining bias *towards* open source products.


    Yeah, that's a little strange though. However, the follow-on link there

    http://www.komar.org/faq/slashdot-ef...004-halloween/

    is very interesting. I tend to believe it here -- the geek network
    emphasized the numbers. How much? Dunno.

    > Again, a little old, but what can you do.
    >
    > http://nixtechnica.blogspot.com/2006...g-systems.html


    9.73% Linux

    > People read digg regardless of OS don't they? "How to host a website
    > on any computer" was the website that was dugg. *Any computer.* Newest
    > and best percentage I've found!


    Indeed. I do think, as one poster notes, that technical users
    contribute to the percentage here, and that the stats have nothing to do
    with PC/Browser share.

    No word on checking for confounding factors.

    --
    Tux rox!

  6. Re: Linux strong #3 in OS market

    On Dec 6, 7:03 pm, Linonut wrote:
    > * cc fired off this tart reply:
    >
    > > On Dec 6, 7:39 am, Linonut wrote:
    > >> * cc fired off this tart reply:
    > >> > If you want to discredit sites like the one in the OP, that's fine.
    > >> > But there are good stats out from sites that aren't affected by one OS
    > >> > or the other. As in, using Linux doesn't predispose you to going to,
    > >> > or avoiding, a site.

    >
    > >> So put up some URLs, dude.

    >
    > >http://www.komar.org/halloween/2004/stats/

    >
    > 6.78% Linux



    News to me! Oh, wait, I read it.


    >
    >
    > > Note the paragraph explaining bias *towards* open source products.

    >
    > Yeah, that's a little strange though. However, the follow-on link there
    >
    > http://www.komar.org/faq/slashdot-ef...004-halloween/
    >
    > is very interesting. I tend to believe it here -- the geek network
    > emphasized the numbers. How much? Dunno.
    >
    > > Again, a little old, but what can you do.

    >
    > >http://nixtechnica.blogspot.com/2006...-operating-sys...

    >
    > 9.73% Linux



    Correct again.

    > > People read digg regardless of OS don't they? "How to host a website
    > > on any computer" was the website that was dugg. *Any computer.* Newest
    > > and best percentage I've found!

    >
    > Indeed. I do think, as one poster notes, that technical users
    > contribute to the percentage here, and that the stats have nothing to do
    > with PC/Browser share.
    >
    > No word on checking for confounding factors.
    >



    There were none.


    You can dismiss them outright, yet you can't seem to find reasons why.
    So I have a question: What do you think the % of desktop users running
    Linux is? Don't give me the Kelsey Bjarnason BS answer of "we can't
    know for sure!" I don't know how many golf balls can fit in a school
    bus, but I can make a good guess. You use Linux, you work with Linux,
    you have friends that use Linux and friends that don't use Linux,
    you're an expert, so give me your guess. I obliged in giving you links
    while you snip relevant portions of my posts, so it's the least you
    could do.

  7. Re: Linux strong #3 in OS market

    flatfish wrote:
    >
    > The main point is that no matter where you look, geek bloggers
    > excepted, the numbers clearly show Linux at less than 1 percent.
    >
    > It's no secret and all the waffling in the world will not change that.


    Actually, no. The numbers I see from across the industry place it at
    somewhere between 2 and 6 percent, depending on how you dice it and
    how you define a 'desktop'. Even the most conservative estimates
    never fall below 2. The webstats posted here claiming less than 1
    likely suffer from some serious sampling error. They probably
    aggregate traffic from mostly one region (the US) and could suffer
    demographic problems among the sampled site. We cant know without
    the methodology and data being released.

    Thad


  8. Re: Linux strong #3 in OS market

    * cc fired off this tart reply:

    >> No word on checking for confounding factors.

    >
    > There were none.


    How do you know? No discussion of them doesn't mean there were not any.
    In fact, it is likely that the lack of discussion means the author did
    not bother checking them out and correcting for them.

    > You can dismiss them outright, yet you can't seem to find reasons why.


    I've listed a half-dozen reasons.

    > So I have a question: What do you think the % of desktop users running
    > Linux is? Don't give me the Kelsey Bjarnason BS answer of "we can't
    > know for sure!"


    I it not a bull**** answer. There is no way of knowing.

    My /guess/ would range from 1% to 3% for /consumers/ only.

    For tech people, my guess would range from 20% to 50%.

    > I don't know how many golf balls can fit in a school
    > bus, but I can make a good guess.


    Maybe. But you're better off making a calculation based on
    closest-packing for spheres.

    In other words, use some /science/.

    > You use Linux, you work with Linux,
    > you have friends that use Linux and friends that don't use Linux,
    > you're an expert, so give me your guess. I obliged in giving you links
    > while you snip relevant portions of my posts, so it's the least you
    > could do.


    My guesses don't mean squat.

    Web-browser stats are almost as bad.

    --
    Tux rox!

  9. Re: Linux strong #3 in OS market

    * thad05@tux.glaci.delete-this.com fired off this tart reply:

    > flatfish wrote:
    >>
    >> The main point is that no matter where you look, geek bloggers
    >> excepted, the numbers clearly show Linux at less than 1 percent.
    >>
    >> It's no secret and all the waffling in the world will not change that.

    >
    > Actually, no. The numbers I see from across the industry place it at
    > somewhere between 2 and 6 percent, depending on how you dice it and
    > how you define a 'desktop'. Even the most conservative estimates
    > never fall below 2. The webstats posted here claiming less than 1
    > likely suffer from some serious sampling error. They probably
    > aggregate traffic from mostly one region (the US) and could suffer
    > demographic problems among the sampled site. We cant know without
    > the methodology and data being released.


    Unfortunately, unless it's on Wikipedia, information on the web is not
    subject to peer review.

    People are free to make up anything on the web. Just ask Microsoft and
    their "Get the Facts" anti-Linux propaganda.

    There is also another confounding factor. There are probably quite a
    few dual-OS users. How do you count them? Do they use one OS for
    browsing, and another for programming? One for play, the other for
    work? One for business, one for personal use? What's the mix?

    --
    Tux rox!

  10. Re: Linux strong #3 in OS market

    Linonut wrote:

    >some worthless troll wrote:
    >>
    >> You can dismiss them outright, yet you can't seem to find reasons why.

    >
    >I've listed a half-dozen reasons.


    Well, one can't be a good wintroll if one isn't a bald-faced liar...


  11. Re: Linux strong #3 in OS market

    Linonut wrote:

    > I it not a bull**** answer. There is no way of knowing.


    Thus statistical sampling is employed.


    > My /guess/ would range from 1% to 3% for /consumers/ only.


    Is this a /scientific/ /guess/?



    > For tech people, my guess would range from 20% to 50%.


    ROFL! You're so full of **** (and slightly dumb if you really believe the
    proportion of developers/programmers/tech support doesn't closely match the
    proportion of users of the platform they work on).

    You weren't using much /science/ when you pulled this guess from down low.



    > In other words, use some /science/.


    Is that anything like /cola advocacy/?



    > My guesses don't mean squat.


    Especially wild-ass farts like 50%. LOL@Numbonut!

    Now if you had said 2% to 5% of Linux/OSS developers and "advocates" use
    Linux exclusively I might agree - but even that's pushing it. Just ask the
    local cola cabal how many never use Windows, and you'll see.




    > Web-browser stats are almost as bad.


    Web browser stats across a range of sites is by far the best statistic on OS
    usage.




  12. Re: Linux strong #3 in OS market

    * DFS fired off this tart reply:

    > Linonut wrote:
    >
    >> I it not a bull**** answer. There is no way of knowing.

    >
    > Thus statistical sampling is employed.
    >
    >> My /guess/ would range from 1% to 3% for /consumers/ only.

    >
    > Is this a /scientific/ /guess/?


    Do you have a reading comprehension problem, DFS?

    >> For tech people, my guess would range from 20% to 50%.

    >
    > ROFL! You're so full of **** (and slightly dumb if you really believe the
    > proportion of developers/programmers/tech support doesn't closely match the
    > proportion of users of the platform they work on).
    >
    > You weren't using much /science/ when you pulled this guess from down low.


    Do you have a reading comprehension problem, DFS?

    >> In other words, use some /science/.

    >
    > Is that anything like /cola advocacy/?


    Do you have a reading comprehension problem, DFS?

    >> My guesses don't mean squat.

    >
    > Especially wild-ass farts like 50%. LOL@Numbonut!


    Why not? That number fits in with my group.`

    > Now if you had said 2% to 5% of Linux/OSS developers and "advocates" use
    > Linux exclusively I might agree - but even that's pushing it. Just ask the
    > local cola cabal how many never use Windows, and you'll see.


    Who said anything about "Linux exclusively"?
    Do you have a reading comprehension problem, DFS?

    >> Web-browser stats are almost as bad.

    >
    > Web browser stats across a range of sites is by far the best statistic on OS
    > usage.


    Nope. You may want to qualify that statement.

    Hint: HTTP isn't the only protocol used by operating systems.

    Pretty silly. "cc" asks me to guess, and then "DFS" gets on my case
    about my guess not being scientific.

    What an idiot.

    --
    Tux rox!

  13. Re: Linux strong #3 in OS market

    On Dec 7, 8:05 am, Linonut wrote:
    > * cc fired off this tart reply:
    >
    > >> No word on checking for confounding factors.

    >
    > > There were none.

    >
    > How do you know? No discussion of them doesn't mean there were not any.
    > In fact, it is likely that the lack of discussion means the author did
    > not bother checking them out and correcting for them.


    How do you know?


    > > You can dismiss them outright, yet you can't seem to find reasons why.

    >
    > I've listed a half-dozen reasons.


    All variations of "You can't prove it!" I say web stats are a good
    indicator, and valid statistics. You say they're not. I give an
    explanation why they are, links to unbiased sites. You claim that I
    can't 100% prove that they're unbiased. I offer up that Linux users
    browsing/Total OS users browsing = Linux users/Total OS users. You
    snip it. Basically I've offered an explanation for all my opinions,
    and the best *you* can do is say, well prove it. Well it's impossible
    for me to prove it because I don't have access to the data. But you
    can't come up with one legit reason why these web stats are so horrid
    other than the fact that *I* can't prove them 100%. You can't prove
    them wrong either, but I don't bring up that ****. I have results from
    unbiased sites, and sites biased *towards* Linux. You've got nothing.

    > > So I have a question: What do you think the % of desktop users running
    > > Linux is? Don't give me the Kelsey Bjarnason BS answer of "we can't
    > > know for sure!"

    >
    > I it not a bull**** answer. There is no way of knowing.



    Of course there isn't, that's why it's a bull**** answer. It's common
    knowledge that we can't know 100% for sure what the usage is. I asked
    what you thought.

    > My /guess/ would range from 1% to 3% for /consumers/ only.
    >
    > For tech people, my guess would range from 20% to 50%.


    So what...5-10% total? That's what I go with.


    > > I don't know how many golf balls can fit in a school
    > > bus, but I can make a good guess.

    >
    > Maybe. But you're better off making a calculation based on
    > closest-packing for spheres.


    Yes. Get an estimate for the size of a golf ball (not all are going to
    be exactly the same), get an estimate for the size of the bus. Take a
    guess at how much space the seats fill up, then go. So you made an
    educated guess for consumers right? How did you come up with it? Based
    on experiences mostly I'll bet, but could it possibly be that the web
    stats factored in? I mean, if web stats were showing 20-25% Linux, I'd
    bet your guess would go up.


    > In other words, use some /science/.


    Did you use no science at all in making your guess?

    > > You use Linux, you work with Linux,
    > > you have friends that use Linux and friends that don't use Linux,
    > > you're an expert, so give me your guess. I obliged in giving you links
    > > while you snip relevant portions of my posts, so it's the least you
    > > could do.

    >
    > My guesses don't mean squat.


    That's horse**** and you know it. You work with Linux, and with people
    who use Linux. If it was more prevalent you would have gone with
    higher numbers. Observation is a key part of science, and you used it.


    > Web-browser stats are almost as bad.
    >


    They're better than your already good intuition. They have a much
    higher sample size, and are much more fact based.

  14. Re: Linux strong #3 in OS market

    * cc fired off this tart reply:

    > All variations of "You can't prove it!" I say web stats are a good
    > indicator, and valid statistics. You say they're not. I give an
    > explanation why they are, links to unbiased sites. You claim that I
    > can't 100% prove that they're unbiased. I offer up that Linux users
    > browsing/Total OS users browsing = Linux users/Total OS users. You
    > snip it. Basically I've offered an explanation for all my opinions,
    > and the best *you* can do is say, well prove it. Well it's impossible
    > for me to prove it because I don't have access to the data. But you
    > can't come up with one legit reason why these web stats are so horrid
    > other than the fact that *I* can't prove them 100%. You can't prove
    > them wrong either, but I don't bring up that ****. I have results from
    > unbiased sites, and sites biased *towards* Linux. You've got nothing.


    Bull****. You've shown no tables of data, no correlations with various
    factors, not even a frickin' T-test.

    (Neither have I, but the presence of bias is so well known in stats
    circles that I don't have to show it. It'd be like trying to prove
    that gravitational effects exist.)

    >> Web-browser stats are almost as bad.

    >
    > They're better than your already good intuition. They have a much
    > higher sample size, and are much more fact based.


    Didn't I just imply that?

    Statistics must be corrected for bias. Show me a site that does that,
    explains their methodology, and provides the raw data and the algorithms
    for deducing the source of the packets.

    --
    Tux rox!

  15. Re: Linux strong #3 in OS market

    On Dec 8, 8:15 am, Linonut wrote:
    > * cc fired off this tart reply:
    >
    > > All variations of "You can't prove it!" I say web stats are a good
    > > indicator, and valid statistics. You say they're not. I give an
    > > explanation why they are, links to unbiased sites. You claim that I
    > > can't 100% prove that they're unbiased. I offer up that Linux users
    > > browsing/Total OS users browsing = Linux users/Total OS users. You
    > > snip it. Basically I've offered an explanation for all my opinions,
    > > and the best *you* can do is say, well prove it. Well it's impossible
    > > for me to prove it because I don't have access to the data. But you
    > > can't come up with one legit reason why these web stats are so horrid
    > > other than the fact that *I* can't prove them 100%. You can't prove
    > > them wrong either, but I don't bring up that ****. I have results from
    > > unbiased sites, and sites biased *towards* Linux. You've got nothing.

    >
    > Bull****. You've shown no tables of data, no correlations with various
    > factors, not even a frickin' T-test.
    >
    > (Neither have I, but the presence of bias is so well known in stats
    > circles that I don't have to show it. It'd be like trying to prove
    > that gravitational effects exist.)



    Now how am I possibly going to do all that without hosting my own
    website? Again you go with "where's the proof." I already stated it
    was impossible for me to prove that at this time, so methinks you need
    a better argument.

    > >> Web-browser stats are almost as bad.

    >
    > > They're better than your already good intuition. They have a much
    > > higher sample size, and are much more fact based.

    >
    > Didn't I just imply that?


    "Almost as bad" is a negative connotation. Your predicition wasn't
    bad, so it should have been phrased "My guess is almost as good as web
    stats."

    > Statistics must be corrected for bias. Show me a site that does that,
    > explains their methodology, and provides the raw data and the algorithms
    > for deducing the source of the packets.
    >


    You're asking for the impossible. You assume they're wrong without
    providing a reason for that assumption. I assume they are correct and
    at least provide something to back it up.

+ Reply to Thread
Page 5 of 5 FirstFirst ... 3 4 5