[UPS] Open Archive metadata set DTD

[Mark Doyle]

Greetings,

> From: Carl Lagoze <lagoze@cs.cornell.edu>
> Date: 1999-12-20 15:19:38 -0500

> I finally got a chance to look at the DTD and I think its got a hole in it. 
> You have author with simply PCData but as spec'ed its structured - name and 
> organization.
>
> Take a look at the output of:
>
> http://www.ncstrl.org:8070/Dienst/Repository/1.0/Disseminate/xxx/001/%23oams 
> /xml
>
> and see if that makes sense to you.  It seems that the author "name" should 
> be required and the "organization" optional. Yes?

I am going to try and update the DTD this week.. I think what you suggest is  
fine, but doesn't reflect the discussion at the conference itself where I  
thought it had been decided that there would be no structure. Personally, I  
agree with what you say (but arXiv may have problems with its unstructured  
data). Opinions?

I am going to add the <p> tag for enclosing paragraphs of the abstract.

I also think we should allow people to add their own elements using an  
extender tag (suggesting <extension> with attribute "name") so that people  
can do things like:

<extension name="version">v1</extension>
<extension name="Subj-class">MSC 1234</extension>
<extension name="source">TeX</extension>

This may encourage people to use the DTD and the metadata scheme without  
having to force them to lump everything into unindentifed <comment> tags.  
Part of the convention will be that people agree not to use the extender tag  
is there is a core element that already covers the data being tagged. We  
could use <comment> as the extender element as well.

If a particular name becomes prevalent, it can be promoted to its own tag to  
formalize the vocabulary. Does that seem reasonable?

Also, how about an oams: namespace (mapped to a URL at http://openarchives.org)?

Cheers,
Mark