> Accordingly, there is every reason for browser providers to have their
> (non-netscape, non-IE) browsers announce themselves as some version of IE.
> It doesn't any longer mean "who are you", it means "which HTML/DHTML
> dialect do you want".
>
> Because of that shift in meaning, site statistics based on browser
> identification DO reliably inform us of HTML/DHTML dialect preference,
> but DON'T reliably tell us anything about browser preference.

Except that, as I mentioned to somebody off the list, that in most cases you
can *still* figure out which browser it really is.  In the case of Opera,
for instance, which is probably the most dominant non-NS, non-IE browser,
version 5 sends back an identification string identifying itself as IE *and*
Opera.  The log analysis tools I'm using, at least, know about this, and my
stats seem to be quite consistent with what others are seeing.

Also, a substantial number of  ISPs add extra cruft to the ID string in OEM
versions of IE which only reinforce the reliability of the identification -
there's little chance that a third-party browser would falsely identify
itself as being on AOL because that would at best be pointless.

So, yes, a browser *could* fool the loggers, but I'm pretty confident that
they aren't to any practical degree.

Having said that, I'm curious if anyone has a list of browser plugins that
are affected by this decision.  So far I've heard that QuickTime and Squeak
are affected, though can anyone confirm the former?  Anyone know about
Flash?  I'm asking because I got a question from a reporter who managed to
put two-and-two together to see how this rather innocent-seeming change
could further advance the domination of Windows Media by throwing a serious
wrench into QuickTime and other media players that used the crossplatform
Netscape API.

-- Duane