Perhaps schema proponents can better articulate why they need an _official_
reworking of EAD in their schema language of choice?

I stress official. EAD is more than the DTD: in fact it is not even
the  DTD, it just so happens that the common currently shared formal
definition of EAD is expressed by DTD syntax (and not just for XML, you
will remember).  Like Clay and many others, I too use formal grammars (and
rules-based validation mechanisms) that are impossible to achieve express
in a DTD.  But really, who cares by what means these constraints are
enforced? My investment is in the markup, not the formal expression of it

A locally derived schema (RELAX NG, XML-verbose) is in heavy use in my unit
at Yale, and has more than realized the effort involved in crafting it.
But  it is used just for local purposes. It constrains an instantiation of
EAD that is specific to our needs, a subset of EAD proper. There are many,
many benefits to our use of this schema, but it remains a utilitarian tool
targeted for specific use. Schema benefits are all on the "processing"
rather than the "exchange" side of things: There is no mechanism for
formally validating an XML instance against a schema for instance. The XML
spec will tell you, amongst other things, that the root node (ead) must
match that of the document type declaration, and that the document type
declaration must contain a reference to a grammar that is expressed in DTD.
How to do this _portably_ with schema?

I'm a big fan of schemas (shemata), and would be further out to sea than
I'd be happy about without RNG, but I would imagine that our community
(which is exceedingly diverse) needs to better understand the nature of the
advanced constraining features that schemata allow for. Example: should we
impose a schema constraint that date@normal be ISO 8601? Then, as Mary Lacy
pointed out in a recent message to this list, how to express distant, BCE
dates? Now your Julian date@normal
<http://scienceworld.wolfram.com/astronomy/JulianDate.html> are invalid.
Creating an official schema is not going to be as simple as letting loose
one of the very many DTD to schema converters on the EAD 202 DTD; it's
going to require a lot of thought and analysis. Locally however, or for
consortial arrangements, schema are invaluable-- they are the mechanical
best practise guidelines, if you will.

just my $0.2.


St.

Stephen Yearl
Systems Archivist
Yale University Library::Manuscripts and Archives



At 05:24 PM 6/2/2004, you wrote:
>Thanks to Clay for bringing this up.
>
>I'd also be VERY interested in finding out the current status of an
>official EAD W3C XML/RelaxNG schema. Like Clay, I'd be willing to do what
>ever I could to assist in the effort.
>
>Terry
>
>On Wed, 2 Jun 2004, Clay Redding wrote:
>
>>Hi,
>>
>>I, as well as several list members and institutions, have now converted
>>the EAD 2002 DTD to XML Schema for local development with XForms,
>>InfoPath, OAI, JAXB, etc.  I'm curious where the provision of an
>>official EAD 2002 XML Schema stands.  I have heard that one is in
>>development, as well as a Relax NG schema.  When might we expect these?
>>
>>I'll be glad to lend a helping hand if possible to help accelerate the
>>process.
>>
>>Thanks,
>>Clay
>
>Terry Catapano
>Special Collections Analyst/Librarian
>Columbia University Libraries Digital Program
>212-854-9942
>[log in to unmask]