W3C home > Mailing lists > Public > uri@w3.org > August 2009

Re: [hybi] [Uri-review] ws: and wss: schemes

From: Jamie Lokier <jamie@shareable.org>
Date: Wed, 12 Aug 2009 05:49:21 +0100
To: Maciej Stachowiak <mjs@apple.com>
Cc: David Booth <david@dbooth.org>, "Daniel R. Tobias" <dan@tobias.name>, uri-review@ietf.org, hybi@ietf.org, uri@w3.org
Message-ID: <20090812044921.GA22368@shareable.org>
Maciej Stachowiak wrote:
> >You've misunderstood.  This would not apply to arbitrary hosts whose
> >name starts with "wss".  Please re-read
> >http://dbooth.org/2006/urn2http/
> 
> What hosts would it apply to? One specific one? If it applies to  
> exactly one host, then I object to making the scheme rely on a central  
> server. We know from experience that this doesn't scale.

There should be no requests to that server.
It's similar to XML namespaces, which should never be dereferenced.

But maybe you have a point.  The mere occurrence of
"http://wss.example.com/some/path" embedded in script in a web page
will probably cause some (non-browser) agents to attempt to fetch the
resource at that URL.  Cheap spidering.

  -- Jamie
Received on Wednesday, 12 August 2009 04:50:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 13 January 2011 12:15:42 GMT