MSIE slipping further?

[Gerald Combs]

Oren Beck wrote:

> HOW does it "look" to the tracking tools if someone is using WINE ?

The same as if they were running under Windows.  Most web log analyzers
parse the contents of the "User-Agent:" string that is passed from the
browser to the server for each request.  E.g. Firefox on my laptop (when
booted to XP) sends

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) \
    Gecko/20041107 Firefox/1.0

The format of this string is a de facto standard, AFAIK.

> The question that comes to mind  is accuracy . Though many
> browsers state that they "can" identify as "BrowserX.X" does it
>  really- show properly to your tool ? Same query as to OS identity .
>  I do wonder if some misidentification is skewing results in unpredictable
>  directions . Example being if I am using Opera installed in a Win9X
> machine
>  then use Knoppix to run Opera- too lazy to import bookmarks for example ?

I have no idea how accurate these statistics are.  They're highly
dependent on the notion that the overwhelming majority of users _don't_
fool with their "User-Agent:" settings, and that browsers don't lie
about themselves by default.  Some web analysis tools use JavaScript
tricks to try to ferret out extra information about the user's
environment (e.g. screen resolution).  You could even use some sort of
passive OS fingerprinting to identify the client, but it might not be
very useful.  I have a feeling that compared to the number of MSIE
users, the number of people running Opera under Wine under Knoppix is
several digits to the right of any useful rounding error.

At any rate, most people who have the need or desire to change their
"User-Agent:" string do so in order to make whatever oddball browser
they're using look like MSIE.  This makes MSIE's drop in numbers
doubly-damning, IMO.