[9618] in www-talk@info.cern.ch
RE: URL Expansion proposal
daemon@ATHENA.MIT.EDU (Paul "S." Wain)
Mon Jan 15 04:31:06 1996
Date: Mon, 15 Jan 96 09:17:02 GMT
From: pwain@acorn.co.uk (Paul "S." Wain)
To: www-talk@w3.org
In-Reply-To: <199601150454.UAA23791@homer.spry.com>
Resent-From: www-talk@w3.org
Ryan Grant wrote:
>
> Anyone with Spry's Mosaic 95 can type "ibm" and see a web page.
>
> Note that for any given "foo", "foo" is not and never ever will be a valid
> url, so this implementation trick does not conflict with the standards (nor
> should it be one).
If we are talking of tricks to make life simple, why not consider one
that in at least one WWW browser I know of. If you enter just the
host name as the URL (i.e. www.acorn.co.uk) you get pushed to the
URL staring "http://" hostname "/" (i.e http://www.acorn.co.uk/ and it does
add the trailing slash :).
This shouldnt bastardise things too much, and is purely a UI issue at this
point then? It doesnt need a new protocol and works in 'most' cases.
One thing that didnt get picked up by the previous debate was that it does
not allow for URLs of the form:
http://www.brunel.ac.uk:8080/
i.e. including the port number? What does that become? IBMs server in
Australia running on port 9999 becomes *IBM*au:8080? Why not just stick
with the URLs in the 1st place, they at least have some meaning after
all!
P.
(Not writing for Acorn, not working for ART).
