Re: Object identifiers -Reply

[Norman Paskin <N.Paskin@DOI.org>]

I believe that we may also usefully learn from the music industry, who
have dealt with the issue of Agreements as part of their Common
Information System (see CIS: A Collective Solution for Copyright
Management in the Digital Age: by Keith Hill; Copyright World issue 76,
Dec 1997).  The CIS system is one which has been the subject of a
tremendous amount of work yet is, to those not intimately involved,
relatively poorly documented (as those involved admit!).  I have been
having discussions with Godfrey Rust, Keith Hill and others about what
we can learn from their work, and the CIS group is well aware of DOI
work.  The CIS includes the development of an International Standard
Agreement Code (an identifier) linked to an Agreement Information File (an
archive).  I do not know if this has any merits for the publishing world but
we certainly ought to look at it.

 

>>> Helen Henderson <Helen_Henderson@classic.msn.com> Mar 16,
1998 >>>
John Erickson proposed an agreement registry or server which would
hold many 
to many relationships between agreements and registered objects with
the 
implication, I think, that the existence of an agreement for a particular 
object could be used so that permissions for access to the object can be 
verified. This sounds very much like the BODOS system that BIDS already
has in 
place and allows a aggregator to provide access to objects to a variety
of 
user sites which may have different agreements with individual rights
owners 
objects available on the aggregators system. There is no reason why this

agreements/permissions database can not be used by different
aggregators.

Agreements may exist to cover DOIs which do not yet exist and the
system will 
need to either be kept up to date or the definition of the DOIs covered by
the 
agreement generic enough for the system not to block access to a site
which 
has an agreement covering an newly released object.

Outside the relationship between owner and user, the same problem of 
agreements covering many objects is likely to exist between trade
partners.  
Here I think that the management of agreements has to be under the direct

control of both parties on their own separate systems. Which need to
allow for 
agreements covering multiple objects some of which have yet to be
produced.




----------
From:  John Erickson
Sent:  Monday, March 09, 1998 18:46
To:  'Sally Morris'; Kelly Frey; 'edward@plato32.demon.co.uk'; 
'chris.zielinski@alcs.co.uk'; 'charles@bic.org.uk'; 'cclark@cla.co.uk'; 
'brian@bic.org.uk'; 'lynette.owen@awl.co.uk'; 'n.paskin@elsevier.co.uk'; 
'skpassoc@internetmci.com'; 'jpeterson@eiq.com';
'mark@markbide.co.uk'; 
'Dafydd.Wynphillips@alcs.co.uk'; 'robertom@pira.co.uk'; 
'ihinds@copyright.com'; 'Helen_Henderson@classic.msn.com'; 
'godfreyrust@dds.netkonect.co.uk'; 'renato@dstc.edu.au';
'lisld@ukoln.ac.uk'
Cc:  'Niso-doi@vms2.macc.wisc.edu'
Subject:  RE: Object identifiers

Hi!

Sally writes...

> As we think about how to express 'agreements' which contain the
applicable
> subset of an object's terms and conditions in the context of a particular
> relationship with a particular user (or group of users), another issue
has
> raised its head, as follows.  If (as I think we all agree) certain
> agreements may cover more than one object - several objects from one
owner,
> from more than one owner (eg National Site Licences) or even a
complete
> repertory (eg CCC/CLA blanket licences), then the identification of
> individual objects needs to be done in such a way that it is possible and
> easy to identify of group of objects to which one agreement applies.    

This is an interesting question. Essentially this is the case of an
object claiming a membership in a class, or family of objects. What
defines that class? To make this work, I think there must be an
agreement registry somewhere that defines many-to-many relationships
between agreements and registered objects.

1. In the case of a 'registered object,' many agreements are associated
with a particular object, enabling the owner to have many agreements for
many different license groups. Individual users or entire classes of
users would be members of such a license group. 

2. An individual agreement might have a family of objects associated
with it, each individually registered.

Thus the database supporting agreement management would contain, in
the
agreement record, a manifest of objects covered under the agreement.
And, likewise, an individual object record would have a manifest of
agreements that pertained to the object. At the usage level, the terms
for a particular instance of use would be defined by the intersection of
the user's credentials, their usage (including setting, exact use, etc),
the object itself and the covering agreement. 

This assumes the availabity of one or more services that can perform the
role of 'agreement server,' to which object names and user credentials
can be presented and candidate agreements returned. This abstraction
requires everything to be 'registered' with a particular service.
Resolution might be particular to name space --- if a publisher assigned
a DOI using a particular internal namespace, the object might
automatically fall under a standing agreement. Or resolution could
transcend namespace. A bridge between URN and DOIs/Handles, forged
right
now, might ensure the viability of that...but that's another discussion!

Have a GREAT one!

John

| John Erickson, Ph.D.   VP-Rights Technologies
| Yankee Rights Management
| 999 Maple Street              802-649-1847(V)
| Contoocook, NH 03229          802-649-2193(F)
| http://www.ybp.com/yrm      jerickson@ybp.com