Re: Relative URI or relative URI reference

• From: Roy T. Fielding <fielding@gbiv.com>
• Date: Wed, 18 Aug 2004 12:50:20 -0700
• To: Elliotte Rusty Harold <elharo@metalab.unc.edu>
• Cc: "Clive D.W. Feather" <clive@demon.net>, uri@w3.org
• Message-Id: <D6504C45-F14F-11D8-A2AE-000393753936@gbiv.com>

> This is exactly the confusion I'm militating against. The idea that 
> "relative URIs" are not a subset of "URIs" is simply bad terminology 
> and it needs to be fixed.  Any normal person is going to say, of 
> course a relative URI is a URI.
>
> Either let's redefine "URI" so that a relative URI is a URI or let's 
> stop talking about relative URIs. The current path is just too damn 
> confusing.

No, I have no power over time and that is the only way to change
conversational use.  There is such a thing as "Relative URIs".
They are those things that you put into relative URI references
when you don't use a URI.

There is no reason for the specification to change just to satisfy
one side of a pedantic argument that has nothing whatsoever to do
with the protocol being defined.  The only thing that matters is
that Relative URI is a noun with a defined meaning, because that
noun is associated with an ABNF rule that defines its syntax.

....Roy

Received on Wednesday, 18 August 2004 19:50:09 UTC