Re: [Ref-Links] Do you need to know content-type?

[Mark Doyle <doyle@aps.org>]

Greetings,

On Tue, 02 Mar 1999, "Tim Ingoldsby" wrote:

> a. have all links first come to an HTML "intermediary" page, where users
> linking in from outside can perhaps see header information (title, authors,
> abstract), and select among full text options (e.g., HTML, PDF, PostScript,
> etc.). This is the strategy being used by Academic Press, among others.
>
> b. have the business partner construct full text link choices on their site
> to each of the full text content formats available for linking.
>
> Obviously, choice (a) does not require a "content-type" field in the DOI
> metadata, but choice (b) does, under most circumstances. I tend to favor
> having choice (b) available to application developers. After all, users who
> have just seen the header information on an A&I site, and have decided that
> they now want to see the full text of the article, will only be irritated
> by being linked to a page on the publisher site that shows them exactly
> what they have just viewed on the A&I/aggregator site all over again. I do
> agree with Eric that there are other ways to achieve this, but even so I
> don't think we should exclude using a "content-type" field in the DOI
> metadata from being one option available to publishers.

The American Physical Society has opted to require outside linkers (through  
a very simple URL - http://publish.aps.org/linkfaq.html) to go to our  
intermediate pages (which contain at least the title/authors/abstract and  
links to all deliverables) instead of right to full-text deliverables. This  
allows us to preserve our 'brand' in some sense because users will see the  
presentation of the article as the APS has designed it. They will see all  
available options for deliverables (in addition to the HTML and PDF that Tim  
mentioned, their could be SGML/XML or, for archival material, individual  
page images). They will see links to errata, comments, replies, and  
companion papers, and links to the table of contents which can serve to put  
the article in context. In general, the idea is to give users a pathway into  
the journal and if they want to navigate around a bit, they can do so  
(adding back in the serendipity factor familiar to researchers who discover  
an adjacent article of more interest than the one they are looking for).

Providing links from the outside directly to deliverables is not as  
scalable because outside agencies will have to maintain links for all  
deliverables (or short change the user by artificially limiting their  
choice). It diminishes the value of having the articles collected together  
into a coherent collection.

I think all of this outweighs the 'irritation' of the user who has to make  
one more click.

So, I would say that the need for content-type metadata is highly dependent  
on the business model a publisher adopts. It is not crucial to the model  
that APS has adopted.

Regards,
Mark

Mark Doyle
Research and Development
The American Physical Society
doyle@aps.org

------------------------------------------------------
Ref-Links maillist  -  Ref-Links@doi.org
http://www.doi.org/mailman/listinfo/ref-links