W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2004

Browser statistics Re: fixed font size

From: Charles McCathieNevile <charles@sidar.org>
Date: Thu, 15 Apr 2004 11:09:52 +1000
Message-Id: <99A4F1E4-8E79-11D8-913D-000A958826AA@sidar.org>
Cc: IG Group <w3c-wai-ig@w3.org>
To: "Paul Davis" <paul@ten-20.com>

Hi Paul,

I am curious about the site stats you look at - although I don't 
necessarily mean to dispute your facts.

I have noticed that many browsers identify themselves as IE (Opera does 
this by default, and many others make it easy to do so). Simple playing 
around suggests that this is to get around browser-sniffers that 
exclude other browsers for no terribly good reason.

Likewise, there are many techniques for stats gathering that do not 
work neatly. Proxying means that straight log checking is fraught, and 
the use of such proxy-aware techniques as images drawn from a CGI 
gateway and not cached by proxies assumes that users who do not get 
images are statistically insignificant.

There was a W3C activity trying to characterise the web and its usage, 
and people like Len Kasday were involved because of the importance of 
any such data to accessibility questions (and the importance of 
questions like the above in the way data are gathered). They gave up, 
in part I believe because it turned out to be an apparently intractable 
problem.

Cheers

Chaals

On 9 Apr 2004, at 19:28, Paul Davis wrote:

> It is a fact according to site stats that I look at that 97% of 
> browsers
> that visit sites I monitor are I.E. in various stages of evolution. 
> (sic.
> congrats to Microsoft)
>
>
--
Charles McCathieNevile                          Fundación Sidar
charles@sidar.org                                http://www.sidar.org
Received on Wednesday, 14 April 2004 21:13:00 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:13:32 UTC