[topicmapmail] Re: Document Object Identifiers/CrossRef

[W. Eliot Kimber]

Daniel Rivers-Moore wrote:

> One other point - You talk about what happens when URLs change because
> the owner of a resource rebrands and so wants the resource to be
> referenced via a new URL. You say "it can be very difficult to find that
> resource's new location".
> 
> More than "difficult", I'd say. Rather "impossible in the general case".
> What I mean by this is that there is *no algorithm* for finding it -
> even if everybody involved obeys all the rules and all the norms of good
> practice.

I was thinking about locating things by Googling them--if the thing is
to be found it can probably be found that way--but that is a purely
human activity, not an algorithmic process.
 
> You have clearly spent some time studying the DOI material, so perhaps
> you know the answer to the following question: Does the DOI paradigm
> provide an algorithm for finding something after ownership of it has
> changed? And a related question: Does the DOI paradigm provide norms of
> good practice for owners of resources to follow when the ownership of a
> resource changes hands? In other words, is there a specified way of
> leaving a 'forwarding address' when you pass ownership of a resource to
> someone else?

The DOI mechanism says "there must be an infrastructure that enables
managers of names to maintain their mapping to resources" (to paraphrase
from memory). CrossRef is the (or a?) reference implementation of this
mechanism. The DOI Handbook is very clear in saying "it is the
responsibility of the name owner to maintain the mapping". Who the DOI
manager is will vary from instance to instance. For example, a publisher
might manage DOIs for many articles, where the article resources are
maintained by 3rd parties. If those third parties move those resources,
they would have to notify the publisher who would then update the DOI
mapping. Or a article author might manage the DOI for her article and
update it as new resources become available (for example, when the
article is published on a journal publisher's Web site).

As far as I know the DOI mechanism only defines one level if
indirection, although I haven't done enough reading to see if they
explicitly allow or disallow one DOI to mapped to another, either
directly or indirectly through a DOI-based URL, although I suspect that
there's not much point to that sort of indirection because if you've got
enough info to map one DOI to another you've got enough info to map the
first DOI to the second DOI's resources.

Cheers,

E.
-- 
W. Eliot Kimber, eliot@isogen.com
Consultant, ISOGEN International

1016 La Posada Dr., Suite 240
Austin, TX  78752 Phone: 512.656.4139