Message-Id: <199404221840.MAA17609@idaknow.acl.lanl.gov>
To: Terry Allen <terry@ora.com>
Subject: Re: Yet more URI/URC
In-Reply-To: Your message of "Thu, 21 Apr 94 19:37:41 PDT."
<199404220237.TAA14099@rock>
Date: Fri, 22 Apr 94 12:40:12 -0600
From: rdaniel@acl.lanl.gov
> Terry Allen writes:
> [Mitra asks] about pathological cases, and on reflection
> I see an issue here about URCs vs URNs. To what extent
> ncan URCs be URNs through being designators for collections?
>
> The pathological cases I can see involve annotations, commentary,
> and tendentious compilations. I won't suggest any particular
> offensive example, but imagine a URC that couples "all-that-you-
> hold-holy" with "all-that-is-most-offensive-to-you". In
> the present world, such a thing would have to be a published
> work, such as a book, to have any meaning. But if anyone can
> make up URCs, then mere bibliographic juxtaposition could
> conceivably have the same effect as actually printing up
> an anthology. Where is the border between bibliography
> and collective work?
This is really a much more subtle question than I thought at
first. Initially I thought that collections typically have
commentary describing particular articles are included. That
new content makes them a resource that gets a URN. There will be
some associated URC(s) so we can find the resource. The URCs may
contain a "References:" field, but it won't have the commentary
so it is not the same as a collection - it is just a bibliographic
listing. So no problem, URNs and URCs are distinct.
But then I though some more about Terry's example of tendentious
(that's "biased" to ex-Okies like me :-) compilations. Yesterday we
talked about compilations that were lists of URNs that cited the
original URN. This is pretty neutral - we didn't assume any editorial
axe was being ground. But there is nothing to stop groups that oppose
the view of the original URN from providing a URC that has a
"Related-Material" field pointing solely to refutations of the original
URN. There is no commentary here, but there is clearly an editorial
bias and you do not have any information to clue you into what is
happening until you read the refutations. This seems to be what Terry
is worried about - editorial content masquerading as impartial content.
Of course, how do we find these extra URCs? If the publisher of the
original resource has control over the list of IP addresses that
"publisher.uri.int" maps onto (and I think they should), these extra
URCs for a URN will not typically show up. So, the extra URC has to have
its own URN. This takes it out of the realm of an impartial service
provided by the system and puts it back into the world of a resource
provided by a publisher that may have a particular agenda to advance.
So this pathalogical case doesn't worry me too much. Any
counter-arguments out there on the list?
> Alas, I fear there are copyright issues here, and that no
> answer we give will resolve the matter. But it's worth
> thinking about in advance.
Copyright issues are something that has been on my mind lately.
There are a couple of things that lessen, but don't eliminate,
the problem. First, you have probably heard of the secure versions
of Mosaic and httpd being developed as part of the CommerceNet
program. Being able to tell who is accessing material should ease the
concerns of copyright holders somewhat since they will be able to
bill people or deny them access. Second, these collections are not
(IMHO) contrary to the long-term interests of copyright holders. The
collections include other works by specifying their URN. If someone
wants to read the cited work, they have to get it from somewhere,
presumably somewhere controlled by the publisher so that a bill
can be generated. What if people just copy big chunks of the original
resource rather than provide a link to it? I would think that
searches based on relevance feedback could locate many of these
occurances, and the copyright holder then has legal recourse. What
about pirate servers? Sorry, no magic bullet here. However, by
definition they will have limited accessibility.
Ron