On Feb 2, 12:35 pm, Roy Schestowitz <newsgro...@xxxxxxxxxxxxxxx>
wrote:
> ____/ Linonut on Saturday 02 February 2008 17:35 : \____
>
>
>
> > * Roy Schestowitz peremptorily fired off this memo:
>
> >> The World According to Linux
>
> >> ,----[ Quote ]
> >>| According to our statistics, 96.4% of computer users are running Linux.
> >>| Whoa! It looks like Linux has completely claimed the market. 0.6% of
> >>| computer users are running Apple; hmm, but I thought the Apple market share
> >>| was improving? Oh, and funny enough, 3% of computer users are running
> >>| Windows. Wait a minute! These stats are completely biased, right? Well, I
> >>| hope you're getting the point.
> >> `----
>
> >>http://linuxhow2.com/Feature_Articles/The_World_According_to_Linux.html
>
> >> Not everyone gets Linux preinstalled on a PC. Few people do. Case of point
> >> (new):
>
> > Maybe using website stats is a stupid way to generalize about the
> > entire PC market. If I happen to visit a site geared towards
> > programmers and web developers, I usually find that the number of
> > Linux users is much higher than usual. Maybe Linux users, like the
> > rest of the people on the Internet, visit those sites that interest
> > them most. Maybe web statistics are stupid. Maybe it's kind of weird
> > that so many Windows, Internet Explorer, users are visiting this
> > site. It's a Linux-fan site, guys!
>
> >> "We are implementing Linux in large govt projects across the country"
> >>http://dqweek.ciol.com/content/Columns/107072002.asp
>
> >> Linux-based Solutions
> >>http://www.dmnnewswire.com/articles/viewarticle.jsp?id=110125
>
> >> Dell leads the Linux workstation market
> >>http://www.thejemreport.com/mambo/content/view/362/
>
> >> Survey Says ? Linux Desktop Is Ever More Popular
> >>http://www.eweek.com/article2/0,1895,2235654,00.asp
>
> >>| Okay, now lets tie this all back in together. Novell claimed several months
> >>| back in a video ad that Desktop Linux users accounted for upwards of
> >>| 30,000,000 different people. That's 30 million. Recent statements made by
> >>| some Novell representatives indicate that they expect there are upwards of
> >>| 50,000,000 Desktop Linux users. Microsoft has never contested the number of
> >>| Desktop Linux users, and if anything the deal Microsoft signed with Novell
> >>| was tacit agreement that Microsoft believed those numbers to be accurate.
>
> >>http://zerias.blogspot.com/2007/11/desktop-os-vista-vs-linux.html
>
> >> Linux Users Base More Than doubled Over Last One Year: Survey
>
> http://www.sda-india.com/sda_india/psecom,id,22,site_layout,sdaindia,...
>
>
>
> >> Survey: Desktop Linux use grows
> >>http://weblog.infoworld.com/techwatch/archives/013576.html
>
> >> Desktop Linux on the Rise, Linux Foundation Reports
> >>http://www.eweek.com/article2/0,1759,2220549,00.asp?kc=EWRSS03119TX1K...
I followed your links. Found some other interesting links.
http://www.kace.com/pdf/KACE_VistaSurvey.pdf
According to the KACE survey, 44% of the companies surveyed are
looking to non-windows solutions, and 68% of those are looking at some
flavor of Linux.
http://tinyurl.com/23sc2s
Forrester things Linux is pretty hot.
<quote>
Linux is becoming a credible threat to Windows on the desktop, and
will grow over the next year as its distributors continue to work hard
at making it an enterprise-class offering, research group Forrester
predicts.
"Will desktop managers continue down the path of standardization on
the Windows platform, and will Linux not exist on enterprise desktops?
Not a chance," Forrester analyst Benjamin Gray says in a report
released Nov. 12 and titled, "How Windows Vista Will Shake Up The
State Of The Enterprise Operating System."
"The high volume of client inquiries for Linux on the desktop clearly
indicates its not going away any time soon. Expect Linux to experience
growth over the next year as the distributors work hard to make it an
enterprise-class offering,"
</quote>
> Their numbers are intended to demoralise users and encourage developers not to
> support Linux. In these studies, someone is typically paying the bills and
> particular sites donate their log files for a reasons. There's history to
> support this assertion.
Remember that browser based statistics only show who has the most
desktops attached to public IP addresses, not the number of desktops
accessing the site. Furthermore, many of these browser surveys look
for the Operating system in the first element of the browser
signature, which on Linux systems is usually "X11" not "Linux".
Windows and Mac systems list "Windows" or "Macintosh" as their first
element. Such surveys don't make any mention of the "Other" catagory,
about 80% of which has the X11 signature. The very small percentage
of Linux measured by these surveys is usually from browsers that have
put Linux in the first field, such as Konqueror. FireFox, for
example, lists X11 as the operating system. Here is a sample of
browser signatures.
http://www.enduser.co.uk/brsrhits.html
We seem to have come full circle. The topic has been widely discussed
on COLA, and now has made it out to a Blog. Maybe it will eventually
bubble up to a major sites or publication.
Here's another example.
http://www.ews.uiuc.edu/bstats/latest.html
Browser Platforms Hosts %
--------------------------------------------
1. Windows 2K (NT5) 2292 50.5
2. Windows NT 361 8.0
3. X11 215 4.7
4. Macintosh 164 3.6
5. Windows 67 1.5
6. other 1440 31.7
http://www.w3schools.com/browsers/browsers_os.asp
2007 WinXP W2000 Win98 Vista W2003 Linux Mac
December 73.4% 4.7% 1.3% 6.5% 1.9% 3.5% 4.0%
November 72.8% 5.1% 1.0% 6.3% 2.0% 3.3% 3.9%
October 73.9% 4.9% 0.9% 5.6% 1.9% 3.3% 3.9%
September 74.3% 5.4% 0.9% 4.5% 2.0% 3.4% 3.9%
August 74.4% 5.7% 0.9% 4.0% 2.0% 3.4% 3.9%
July 74.6% 6.0% 0.9% 3.6% 2.0% 3.4% 4.0%
June 74.6% 6.2% 0.9% 3.2% 2.0% 3.4% 4.0%
May 75.0% 6.5% 0.9% 2.8% 1.9% 3.4% 3.9%
April 75.8% 6.8% 0.9% 2.6% 1.9% 3.4% 3.9%
March 76.0% 7.2% 0.9% 1.9% 1.9% 3.4% 3.8%
February 76.1% 7.4% 0.9% 1.2% 1.9% 3.5% 3.8%
January 76.1% 7.7% 1.0% 0.6% 1.9% 3.6% 3.8%
Remember too, that since the survey just measures IP addresses, it
becomes easier to skew the result by adding more IP addresses to the
DHCP pool.
Then you have to weight based on the type of connection. Linux users
typically use NAT routers, either with their direct connection, or
through their ISP. Windows users often get DHCP addresses with each
reconnection. If a dial-up users dials every day, or a cellular phone
user reconnects only once per day, each Windows machine could have 30
different IP addresses. If the Linux machines are sharing a single
NAT router that keeps the same public IP address for over a month, you
could have 2000 Linux machines hiding behind only 1 IP address.
Some of these surveys also assume that there can be only one operating
system per IP address, so if an IP has BOTH Linux and Windows behind
it, the statistics could end up only measuring the Windows system.
Some surveys count the number of hits for each operating system. If a
Linux user uses a cache such as squid, the browser will get the
locally cached copy over and over, while the Windows users will be
getting hits for each GIF, include, or element of the frame.
If Windows tries to preload links, or ActiveScript pulls down lots of
different items, this can also skew the count.
|
|
