W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2013

Re: Recommended URL structure - pages in the root itself seems latest

From: David Woolley <forums@david-woolley.me.uk>
Date: Thu, 24 Jan 2013 18:39:22 +0000
Message-ID: <51017FDA.5000907@david-woolley.me.uk>
To: 'wai-ig list' <w3c-wai-ig@w3.org>
Ganesh J. Acharya wrote:
> Now days many websites try to use URLs like 
> 
> example.com/car-model-112 <http://example.com/car-model-112> like 
> formats, (also this one seems to have lesser accessibility issues.)
> 
> instead of 
> 
> example.com/suv/car-model-112.html 
> <http://example.com/suv/car-model-112.html> etc
> 
> What are the recommended URL structures and why?

These are both relative URLs.  I am pretty sure that you actually meant
these to be //example.com/car-model-112, etc.  It would be strange to
have something looking like a host name in the local part of the URL.

Can you provide a little more context.  The only time end users should
see URLs is if they look at the address bar, so these should not appear
visibly in the body.  From that point of view using a URL that reflects
the logical structure of the information would seem better, to me.  I
would further say that //example.com/suv/ should produce a page
cataloguing all their SUV models, not a message saying directory listing
not allowed.  However, web sites are so bad at doing this that there is
no general expectation that it will work.

If a URL is expected to get transcribed into print media, I would
personally want the scheme prefix included, as well.  I would prefer
that emailed URLs also didn't omit the scheme.  In fact, I had a recent
case where my old copy of Thunderbird refused to acknowledge a URL in
one of the OPs formats, when included in HTML mail.

It is, however, true, that URLs in print, and HTTP ones in email, are
generally absolute, and browsers make a reasonably good attempt at
guessing the scheme from other elements of the syntax (although they
wouldn't be able to distinguish between a sip: and mailto: one, or a
sip: one without a user part and an http: home page.

Although current browser heuristics will probably work for mailto: and
http:, I think one is being lazy and relying on undocumented heuristics,
if one actually relies on that in formal documents.

The .html is really there for technical reasons, and from a human
accessibility point of view, I don't see a problem in dropping it.
However, if it is dropped, it becomes even more important to use a
proper absolute URL format.

Providing the scheme is present, there has never been a real need for
the domain to start with www.


-- 
David Woolley
Emails are not formal business letters, whatever businesses may want.
RFC1855 says there should be an address here, but, in a world of spam,
that is no longer good advice, as archive address hiding may not work.
Received on Thursday, 24 January 2013 18:39:57 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 24 January 2013 18:39:57 GMT