03.29.08
Boycott Novell Contradicts Lies, Damn Lies and Statistics
We previously offered and presented a geographical breakdown of Boycott Novell’s readership, old Web browsers breakdown, and a look at Linux usage over time. Herein we use data from the past 28 days, as interpreted by AWStats. None of this data was preprocessed in any way.
| Versions | Hits | Percent | ||
|---|---|---|---|---|
| Windows | 593749 | 49.7 % | ||
 |
Windows XP |
452012 | 37.8 % | |
 |
Windows NT |
2045 | 0.1 % | |
 |
Windows Me |
605 | 0 % | |
 |
Windows Vista |
96080 | 8 % | |
 |
Windows CE |
385 | 0 % | |
 |
Windows 98 |
3022 | 0.2 % | |
 |
Windows 95 |
153 | 0 % | |
 |
Windows 2003 |
19950 | 1.6 % | |
 |
Windows 2000 |
19408 | 1.6 % | |
 |
Windows 3.xx |
89 | 0 % | |
| BSD | 2267 | 0.1 % | ||
 |
OpenBSD |
312 | 0 % | |
 |
NetBSD |
146 | 0 % | |
 |
FreeBSD |
1809 | 0.1 % | |
| Linux | 427822 | 35.8 % | ||
 |
Ubuntu |
169142 | 14.1 % | |
 |
Suse |
38320 | 3.2 % | |
 |
Red Hat |
2096 | 0.1 % | |
 |
Mandriva (or Mandrake) |
6576 | 0.5 % | |
 |
Fedora |
28780 | 2.4 % | |
 |
Debian |
33768 | 2.8 % | |
 |
Centos |
1079 | 0 % | |
 |
GNU Linux (Unknown or unspecified distribution) | 148061 | 12.4 % | |
| Macintosh | 64972 | 5.4 % | ||
 |
Mac OS X |
64857 | 5.4 % | |
 |
Mac OS |
115 | 0 % | |
| Others | 104661 | 8.7 % | ||
 |
Unknown | 102756 | 8.6 % | |
 |
Sun Solaris |
857 | 0 % | |
 |
Symbian OS |
353 | 0 % | |
 |
Unknown Unix system | 336 | 0 % | |
 |
Sony PlayStation Portable |
228 | 0 % | |
 |
OS/2 |
52 | 0 % | |
 |
CPM |
28 | 0 % | |
 |
BeOS |
17 | 0 % | |
 |
RISC OS |
8 | 0 % | |
 |
AmigaOS |
8 | 0 % | |
 |
Irix |
7 | 0 % | |
 |
HP UX |
7 | 0 % | |
 |
WebTV |
4 | 0 % | |
This hopefully helps in combatting the infamous market share lie. █
Woods said,
March 29, 2008 at 5:21 am
OMG!
People are still using CP/M *and* using it browse the Net! That is just too cool
Hats off, kudos and much respect!
Victor Soliz said,
March 29, 2008 at 8:59 am
Some of us are forced to use XP if we want to use other computers than ours to browse the net, so that probably explains the big amount of XP visits.
tomek said,
March 29, 2008 at 3:01 pm
hmm… something must be in those stats…
I have a web site (about 5 to 10 thousand web hits per month) and my stats give very similar percentage share, especially when it comes to FF vs IE (when comparing to this: http://boycottnovell.com/2008/01/29/january-stat/) but also the Lin vs Win stats are similar.
On my site:
FF ~ 60% vs IE ~ 13%
and
Win ~ 64% vs Lin ~ 32 %
I thought that the numbers were caused by the specific audience of my site (a Free Software project), but if they are so similar to stats of another web site (also with a specific audience, but still a different audience), then it means something, I think.
Roy Schestowitz said,
March 30, 2008 at 10:02 am
A while ago it was explained that — among a variety of factors — niches and population samples play a tremendously significant role. Web statistics that are published by the ‘big surveys’ tend to just take into account the Web sites which reach a broad audience and whose Webmasters disregard privacy of visitors. You’;re missing out vast bits of the whole population. Mind the headline of this post (statistics being worse than damn lies).
CoolGuy said,
March 30, 2008 at 10:36 am
I hit couple a fifty refresh per day over here as I have no life (ubuntu).
Roy Schestowitz said,
March 30, 2008 at 10:52 am
You can probably subscribe to the comments feed or use E-mail alerts.
Would the following be helpful?
http://boycottnovell.com/comments/feed/
ZiggyFish said,
April 13, 2008 at 6:19 pm
What would be good is to see the Web stats for Google. Whether directly or indirectly(though Google Ads), most people will visit that server. Also Google is a natural site.
Roy Schestowitz said,
April 13, 2008 at 8:34 pm
AdSense suffers from the ‘issue’ that it gets blocked by those who are more Internet-savvy (usually Mac and GNU/Linux users) and Google users in general are said to be those who understand its advantage over Yahoo. So, even Google queries (query volume to take into account) and AdSense (tabbling and surfing frequency to take into account) are no good measures, but maybe something close to reality. There are other issues (some listed here) such as diversity of strings (HTTP headers), forging, bots querying Google, zombies (320,000,000 Windows PCs are zombies nowadays), Squid, etc.