From: wweems@oac3.hsc.uth.tmc.edu (William A. Weems)
Message-Id: <9310201429.ZM17933@oac3.hsc.uth.tmc.edu>
Date: Wed, 20 Oct 1993 14:29:59 -0500
In-Reply-To: Terry Allen <terry@ora.com>
To: Terry Allen <terry@ora.com>
Subject: Re: Adding ISSNs.
On Oct 20, 7:54am, Terry Allen wrote:
> Subject: Re: Adding ISSNs.
> ISSNs point to the whole run of a journal, not individual issues.
>
>-- End of excerpt from Terry Allen
Would not ISSN designate the naming authority scheme identifier? Seems like we
have two possibilities
<URN:ISSN:1234-5679:J_THEOR_Obj_F54_1994>
or
<URN:NSDP:ISSN1234-56789:J_Theor_Obj_F54_1994>
where NSDP => National Serials Data Program of the Library of Congress.
>From the proposed URN draft we see:
< 3.3.5 The Opaque String
< The opaque string component of the URN is any string the Naming Authority
< wishes to assign to a given resource, subject only to the constraints of the
< character encoding scheme.
This would seem to imply that the naming authority for serials could include in
the opaque string information designating the volume or issue as well as a
specific article.
It seems that EACH article in a Journal will require a URN number because it is
the article that one cites when making a reference within a paper that is being
written. Thus, if one wants to have a hyperlink to the article written by
"Allen, et.al", the URI system MUST be able to find and display the article.
The ISSN space is also where we have to deal with the fact that a publisher does
in fact make judgment as to "sameness". As I write this, we are working with a
major scientific society that has been publishing for over a century.They would
like to begin publishing articles in Jan. 1994 using both Gopher (i.e. ASCII)
and WWW (i.e. HTML). Both the publisher and the authors agree that the
information content in both formats is the SAME. In order to allow referencing
to exist for decades, an article distributed in either the Gopher or HTML format
will probably have to have the SAME URN. The Gopher and HTML versions of the
serial will thus also need the same ISSN#.
Relative to this issue, Rob Raisch wrote:
<Food for thought: Assume a document in ASCII and the same document in
<Postscript. Do they contain the same "intellectual content?" What metric
<do we, (wearing our publisher hats), use to decide?
<Consider the following example. Is there a difference in "intellectual
<content" in these two sentences?
<Sentence one: The hammer hit the nail with enough force to shatter the wood.
<Sentence two: The hammer hit the nail with enough force to
< <ITALIC>shatter</ITALIC> the wood.
Operationally, a responsible publisher/archivist would not convert sentence one
into sentence two when say moving from ASCII to Postscript or HTML. They might,
for example use <ITALIC> to make the title of the article in an abstract stand
out from the authors and journal reference when they go from an ASCII to an HTML
version. This would not change content.
Another real world example is that say a scientific society that has been
publishing for decades wants to make earlier paper volumes available as
electronic test. Would the publisher have to get a new ISSN number? If the
publisher is careful in the conversion, would anyone say the information content
has changed within the context for which the articles were written? Ideally, we
might want to say that such conversion should not be made. But operational they
must be made if information is to be maintained in a current, readily usable
format.
Best Wishes,
Bill
--William A. Weems, Ph.D MSB 4.218 Director P.O. Box 20708 Office of Academic Computing Houston, Texas 77225 U. of Texas Houston-Health Sci. Cntr. Phone: 713-792-5277 Associate Professor, Physiology FAX: 713-794-5653 U. of Texas Medical School Houston wweems@oac.hsc.uth.tmc.edu