Re: [URN] Re: Relative URLs and URNs

On Thu, 30 Jan 1997, Roy T. Fielding wrote:

> > Thus spoke Keith Moore:
> > 
> >> If we disallow unencoded '/' within URNs for now, we can always add 
> >> relative URNs later when we understand them better.
> 
> I couldn't disagree more.
> 
> If '/' is disallowed in URNs, then a conforming parser will reject
> a URN containing '/'.  That means you will never be able to introduce
> relative URNs.  Creating barriers to technology just because it isn't
> understood is a mistake, particularly when the only time '/' is
> meaningful for *any* URI is when it is used as a base for parsing
> relative URIs.
> 
> If you aren't using relative forms, then there is no reason to encode '/'.
> If you are, then you don't want '/' encoded.  There does not exist a
> situation in which it is beneficial to disallow '/' in URNs.

There might not be any disagreement. Keith wants '/' disallowed
within actually existing URNs. The parser can (and should, in this
case) be more open. "SHOULD" is the best way to say it (contrary
to my message a few minutes ago, where I didn't see that "SHOULD"
was on puprose).

Regards,	Martin.