Re: Registrar

• Mail folder: WWW Talk Jul-Oct 1993
• Next message: David Martland: "MOSAIC and embedded images... b/w vs colour"
• Previous message: Dave_Raggett: "Re: Figure overlays in HTML+"

Date: Wed, 14 Jul 1993 05:19:23 -0500
From: Ed Krol <krol@ux1.cso.uiuc.edu>
Message-id: <199307141019.AA20407@ux1.cso.uiuc.edu>
To: raisch@ora.com, timbl@nxoc01.cern.ch
Subject: Re: Registrar
Cc: ccoprmm@oit.gatech.edu, terry@ora.com, uri@bunyip.com,
        www-talk@nxoc01.cern.ch
Status: RO

Rob writes:

>URNs do not refer to living documents.  URNs refer to static products. 
>Just as the URN is forever, the product to which it refers is forever, and
>if there is some reason to change it, that change creates a new product.
>If the product is for some reason no longer available, then the URN maps
>to zero URLs.

I think this is slightly overstated.  URNs refer to documents
which will be static products.  You really open a can of worms
if you say this this strongly.  Does this mean that the URN working
paper cannot have a URN?  I like the print world analogy.  Sometimes
a book is reprinted with minor corrections and sometimes it becomes
a 2nd edition.  In these cases the reprints with corrections
would have the same URN and the 2nd edition would have a different
one.  The author and publisher (URN authority) determine which is 
the case.

The only difference in the two cases is that "minor corrections"
can occur only with some difficulty (and it is obvious from
the text, by convention on the printing history page, which version
you have).  In the electronic case, changes can occur willy nilly
and there is no history.