I suspect that you're being ironic. Unfortunately there may be many
readers who know so little history that they might literally believe
you.
Have a look at
http://www.w3.org/History/19921103-hypertext/hypertext/WWW/Status.html
and see if you can discern dominance of any particular OS. NeXT,
VAX/VMS, IBM VM/CMS, Macintosh, all feature in different ways.
(P.s the claim made in that page that the VM/CMS browser was
non-existent was not really true. Some of us supported a port of the
W3 client to VM/CMS: it wasn't very good, but it certainly wasn't
"nonexistant"(sic) as that document claims. In fact, see
http://www.w3.org/History/19921103-hypertext/hypertext/WWW/LineMode/vm-cms/Overview.html
But that's all a long time ago now)
But the MIME and HTTP specifications are not just a private affair of
the WWW - they have proper IETF standards-track specifications about
their behaviour on the Internet. They are carefully defined to be
rather OS-agnostic. *If* the claim had been true that they were
unix-centric, do you suppose they would have defined the canonical
newline representation to be CRLF? Surely it would have been LF
alone, just as it is in unix?
I know you're joking, really, but...
We would be entitled to scold /anyone/ who uses filename extensions in
their URLs. To anyone who takes theory as more important than
practice, it's clear that they have no business being there at all.
(If you /really/ care about network bandwidth, you should be aware
that most browsers nowadays support gzip compression, and (x)HTML
lends itself very well to such compression. You could save file space
too, in that way).