Find the answer to your Linux question:
Page 1 of 3 1 2 3 LastLast
Results 1 to 10 of 29
Here is my informal statistical research of linux distro reliabilty based on ratio of frequency hits via Google search of KEYWORDS with syntax: # of DISTRO NAME LINUX PROBLEMS hits ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
  1. #1
    Linux Newbie
    Join Date
    Jan 2005
    Posts
    146

    LINUX DISTRO RELIABILTY SCORES


    Here is my informal statistical research of linux distro reliabilty
    based on ratio of frequency hits via Google search of KEYWORDS
    with syntax:
    # of DISTRO NAME LINUX PROBLEMS hits /# of DISTRO NAME hits

    ratio expresses relative reliabilty of distros
    gentoo linux 875000/3550000=0.2464
    fedora core linux 667,000/2000000=0.3335
    ubuntu linux 267000/2290000=0.1166
    debian linux 3200000/11400000=0.2807
    slackware linux 563000/2920000=0.1928
    knoppix linux 227000/1820000=0.1247
    suse linux 2280000/9310000=0.2449
    mandrake linux 755000/4570000=0.1652
    windows 98 7520000/59300000=0.1268


    All things being equal this experiment can be considered scientific
    and accurate enough for analyzing end use satisfaction and reliability

    Ubuntu can be considered the most reliable and excellent in enduser satisfaction.

    *Win98 was included in the analysis but cannot be considered accurate or fair since statistically, linux users tend to post more often on the net and are poorly served by manufacturer's drivers and support.

  2. #2
    Just Joined!
    Join Date
    May 2005
    Location
    Down by the old oak grove
    Posts
    34
    Not 100% airtight but an interesting method nonetheless. A poll would probably be more accurate. Also, you forgot RedHat.

  3. #3
    Linux Guru bryansmith's Avatar
    Join Date
    Nov 2004
    Location
    /Ontario/Canada
    Posts
    2,619
    Quote Originally Posted by Darkbackwardz
    Not 100% airtight but an interesting method nonetheless.
    I agree. It is intersting but with some distros (like Gentoo for example), there is more done by hand and therefore there is likely more problems/questions.

    Bryan
    Looking for a distro? Look here.
    "There can be no doubt that all our knowledge begins with experience." - Immanuel Kant (Critique of Pure Reason)
    Queen's University - Arts and Science 2008 (Sociology)
    Registered Linux User #386147.

  4. #4
    Linux Newbie
    Join Date
    Apr 2005
    Posts
    235
    what was the first distro you ever heard of?
    for me it was red hat, which is almost the same thing as fedora. if neone else agrees that they first heard of red hat, it may mean that all the noobs jump strait into red hat / fedora (as did I) and therefore resulted in a slightly higher ratio. one last thing to worry about. i just did the same test windows and got the ratios for xp, 2000, me, 98, and 95 in order: .14, .18, .27, .13, .2
    so i got curious just cause i hate to c windows doing better than linux, so i went thru the first 20 of each link. from what i found the windows problems were much more serious, and i made a somewhat vague separating line and determined that 14/20 of the windows problems were serious. whist 18/20 of the linux problems were just general inquiries asked like noobs like me.

  5. #5
    Linux Engineer
    Join Date
    Mar 2005
    Location
    Where my hat is
    Posts
    766

    Re: LINUX DISTRO RELIABILTY SCORES

    Quote Originally Posted by Dalani
    Here is my informal statistical research of linux distro reliabilty
    based on ratio of frequency hits via Google search of KEYWORDS
    with syntax:
    # of DISTRO NAME LINUX PROBLEMS hits /# of DISTRO NAME hits
    And those of us that do a quick search on Google because the individual who has asked the question is too lazy to do the same? The numbers will not reflect the accurate use of the distribution.
    Registered Linux user #384279
    Vector Linux SOHO 7

  6. #6
    Linux User
    Join Date
    Jan 2005
    Location
    Arizona
    Posts
    288
    That Windows 98 score has to make you wonder what effect marketing might have on the results. I'd think that at some point number of advertisements (distributor, reseller, copy/paste "review"/advertising scam, or even gossip) proportionally surpasses the number of users it recruits, and thus the number of problems resulting. Or am I missing something else more obvious that might cause this? Because, uhh... does anyone here actually think Win98 is more reliable than Debian?

    Maybe it's that users are more active in development in OSS. A corrolation of that would be higher scores (higher as in better) for commercial distributions like Libranet, Linspire, and Xandros.

    Arch .1589
    Linspire .1786
    Libranet .4740
    Xandros .0909
    Lycoris .1438

    I don't think it's accurate enough to form any hypothesis on that.

    Some other difficulties, distributions that refer to themselves with 'Linux' in their name, like Arch and Debian, likely alter their score.

    Maybe you need to include, in addition to 'problems'; 'issue', 'difficulty', 'complication', 'quandary', 'conundrum', and 'bugaboo'. ...you get the point.

    And again, a distribution which starts you off with a base console prompt and a package manager is going to have a lot more "problems" than one which boots you into a fully configured desktop environment.

    Everything has to start somewhere, do you plan to refine it?
    Michael Salivar

    Man knows himself insofar as he knows the world, becoming aware of it only in himself, and of himself only within it.
    --Goethe

  7. #7
    Linux User
    Join Date
    May 2005
    Location
    SA, TX
    Posts
    333
    where do ya get the time to do stuff like this

  8. #8
    Linux Newbie
    Join Date
    Jan 2005
    Posts
    146
    Maybe you need to include, in addition to 'problems'; 'issue', 'difficulty', 'complication', 'quandary', 'conundrum', and 'bugaboo'
    You have to understand that this method is relative not qualitative. ie it's accuracy is based on a comparison of similar though different items- in this case Linux distros. The word 'problem' is as likely to be used by any distro user as another. As long as all queries are done with the same keywords except the distro name.

    It's true that hand built Linux distros will not fare well in this test then 'commercial' desktop distros like Mandrake or Linspire. That's the whole point: ease of use and reliability (less frequent 'problems') will be apparent from these results.

    caveat emptor

  9. #9
    Linux Newbie
    Join Date
    Aug 2004
    Posts
    220
    the first thing a stat class teaches you is that if you gather data improberly (basicly anything but an SRS) then your conclusions well be wrong. as seen by this

    I did the same thing but decided to use "bugs" instead of "problems" just because.

    microsoft 0.03
    windows 0.05
    fedora 0.11
    slackware 0.15
    mandrake 0.15
    debian 0.2
    knoppix 0.2
    redhat 0.21
    suse 0.21
    gentoo 0.24
    ubuntu 0.25

    notice how windows, fedora, and mandrake are near the top (notice how fedora is near the top but redhat is not)

    this is because the search "windows" found windows os and windows (glass), fedora = hat , mandrake = plant, and harry potter refrence

    but the biggest problem whould be all the pages that have a small program that includes the text "works on windows 98/xp" but no complains becuase they would like you to think their product is perfect.

    this is my guess as to why windows has "fewer bugs" then every linux distro

    but nice effort

    edit: ya I know that you but linux after every one (I'm assuming that you have AND in there as well) but you did but "fedora core" instead of "fedora" which is likely to decrease the number of "user" pages instead of offical pages which include know problems etc.

    any way I was just point out WHY windows scored so high

  10. #10
    scm
    scm is offline
    Linux Engineer
    Join Date
    Feb 2005
    Posts
    1,044
    Most Windows users wouldn't have the wit to use Google to figure out problems - they'd go to a computer store, or phone Micro$oft.

Page 1 of 3 1 2 3 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •