reality check

[nubcnubdo]

reality check:

1) Linux market share at 1.34%
http://www.arstechnica.com/news.ars/post/20070903-linux-marketshare-set-to-surpass-windows-98.html

2) VL ranks 25th (3 month time frame) after flap at DW

[Triarius Fidelis]

Either that 1,34% figure or counter.li.org's estimate of the number of Linux users worldwide is incorrect. Assuming both are correct, the number of computer users worldwide is something like 2,16 billion.

Something isn't right. Which is it?

[nubcnubdo]

cited sources:

http://www.w3counter.com/globalstats.php

http://marketshare.hitslink.com/report.aspx?qprid=2

[Triarius Fidelis]

At least one of any of those sources (mine or ours) must be wrong.

[nubcnubdo]

I dont consider it in terms of right and wrong. Rather, I see the report as a statistical snapshot, based on a good-sized sample, with no obvious bias in methodology. I'm not sure that absolute values of market share can be known. Besides, this low value isnt that surprising, is it? What do you estimate, 3.x%?

This article at DesktopLinux says Linux market share is 3.4%
http://www.desktoplinux.com/news/NS6997734761.html

2) VL ranks 25th (3 month time frame) after flap at DW

One thing should be clear from this stat: We can't afford negative publicity of any kind.

[Triarius Fidelis]

I dont consider it in terms of right and wrong. Rather, I see the report as a statistical snapshot, based on a good-sized sample, with no obvious bias in methodology. I'm not sure that absolute values of market share can be known. Besides, this low value isnt that surprising, is it? What do you estimate, 3.x%?

Of course, nothing definite can be known by design. I like it that way. My guess is that it's somewhere between those two figures at 2,x% (Only a guess.)

[Toe]

My personal estimates would be like this:

Web servers: Major presence.
Other servers: Significant, but relatively minor.
Embeded: Strong presence.
Desktop: Pretty much irrelevant outside geeks.

[incognu]

Hmm ... I just noticed this:

The W3Counter's statistics, which are based on the web browser user agent text

(arstechnica)
and

Some ways not to estimate

Some measures are lower than others.

One in particular that is low is estimates based on the number of browsers identifying themselves as Linux.

(counter.li.org)

[nubcnubdo]

Good point, incognu. The fact that browsers of many distributions do not give a Linux ID may account for the low value for market share by W3Counter. As I recall, VL's browsers do not broadcast a Linux ID.

Sorry, hanumizzle, you were right about the low figure 1.34% being incorrect. Flawed methodology.

CNN Money: Linux finds home on more desktops
http://money.cnn.com/news/newsfeeds/articles/newstex/IBD-0001-19346172.htm

OTOH:
http://www.theinquirer.net/gb/inquirer/news/2007/10/08/mac-linux-fail-capitalise-vista

[Joe1962]

As I recall, VL's browsers do not broadcast a Linux ID.

Our SeaMonkey does, as of 1.1.2: http://www.vectorlinux.com/forum2/index.php?topic=3188.msg19773#msg19773

[easuter]

So does our Firefox package, since version 2.0.0.6. http://www.vectorlinux.com/forum2/index.php?topic=4010.0

[Triarius Fidelis]

Sorry, hanumizzle, you were right about the low figure 1.34% being incorrect. Flawed methodology.

CNN Money: Linux finds home on more desktops
http://money.cnn.com/news/newsfeeds/articles/newstex/IBD-0001-19346172.htm

Yeah, I think there's a larger market, since even the crass OEMs have joined in earnest.

[Triarius Fidelis]