Message-Id: <9403021527.AA28001@expresso.bunyip.com>
From: Peter Deutsch <peterd@bunyip.com>
Date: Wed, 2 Mar 1994 10:27:23 -0500
In-Reply-To: Tim Berners-Lee's message as of Mar 2, 10:47
To: timbl@www0.cern.ch, uri@bunyip.com
Subject: Re: URN and DNS namespaces
Hi all,
There's something that's been ringing gently in my ears for
a while now and I've only now realized what's causing it...
[ Tim wrote: ]
. . .
> . . . No suppose that when we actually do the URN
> convertion we build in a special case that URNs of the form
>
> urn:/internetdns/com/pubcom/mkt/98123/182346
>
> instead of being resolved by
>
> pubcom.com.internetdns.urn
>
> would be resolved by
>
> pubcom.com
Every time I look at an example in this discussion for using
DNS in URNs I have to perform a mental flip of the FQDN
(as above) to get the URN. I just realized that this is
exactly what I had to do with British email addresses and
most people think this is fairly bogus (the Brits because
we have it wrong, the Yanks et. al. because the Brits have
it wrong and the lines form on the left to pick a side on
_that_ fight).
Now, without deciding whether the British or the rest of
us are correct, I believe that _requiring_ users to do
this flip for URNs is a "Bad Thing (tm)". URNs are
supposed to be user-friendly, DNS is already widely used
and at least partially understood and if people want to
perform the "guessing" method of resource discovery we
decrease their chances of success by requiring the mental
arithmetic.
It seems to me sort of like we're asking them to perform
one of those "skill testing questions" to claim the prize
under the bottle cap and I think that wrong. So, without
touching upon any of the rest of the issues in this
debate, I request and suggest that we ask proponents of
this resource discovery method to use the format:
URN:182346/98123/mkt/pubcom/com/internetdns/
instead of:
urn:/internetdns/com/pubcom/mkt/98123/182346
I can see the problems of of having the string grow "into"
the URN: prefix, but feel that this is outweighed by the
increased ease of use for naive users. Everything in
life's a tradeoff and I think the added complexity for
users is just not justifiable here.
. . .
> The abstract namespace should be kept separate from the resolution
> mechanism, but that doesn't stop an existing DNS space being adopted
> as part of the abstract URN tree.
Hear, hear. And in that case, I'm suggesting that we not
require users to perform mental gymnastics.
> This would also mean that we could actually try the mechanism out.
Hear, hear (2).
- peterd
--
---------------------------------------------------------------------------
"The future belongs to neither the conduit or content players, but
those who control the filtering, searching and sense-making tools
we will rely on to navigate through the expanses of cyberspace."
- Paul Saffo, (_Wired_: March,1994)
---------------------------------------------------------------------------