In writing the stylesheets for HTML and PDF formats for our EAD 2002
XML documents, I found the DTD rather un-helpful in forecasting the
possible children of DSC based on the @TYPE values. The values are
enumerated, but there is no attempt to validate the DSC based on this
Neither is there any help in the DTD for the children of the DSC.
There need be no components at all.
Therefore, I considered the DSC merely a wrapper element. The
children determined the transformation/generation of the new document
format. Thus the DSC/@TYPE value was not even considered.
Have you found some method whereby this attribute value can be used?
Library of Congress
--- Clay Redding <[log in to unmask]> wrote:
> Hi all,
> I'll soon begin working on new stylesheets for EAD conversion to
> formats, and I'd like to gather opinions on how some of you have
> on where to display series overviews (scopecontents, arrangements,
> physdescs, etc.). It seems to be a mixed bag in terms of how and
> everyone presents this data. Some use the dsc analyticover
> approach and
> place descriptions for all series prior to the full container list.
> Others use the dsc combined approach and describe the series only
> their subordinate contents are being presented to users. I've only
> used the latter approach, as this seems to be what most best
> documents and ISAD(G) suggest. No matter what, I'll continue using
> combined with multilevel description in the EAD. However, it makes
> sense many times for the series descriptions to be clearly laid out
> front for the user in the end display (via XSLT). But when does
> know to do this?
> In searching the list archives, this mail from Michael Fox in 2000
> strikes home:
> "With XSL stylesheets, it is possible to reorder and reuse data in
> more than
> one location in your presentation. Using the combined model, one
> extract series descriptions to create a separate list of series
> dates, quantities, scope and content descriptions, etc., as you
> wish, and
> then display (or not) the same data again with the details of the
> of each series. It is not necessary to have the data in a
> separate section
> or key it twice into two sections.
> We encode the data once in the combined model. In the past we have
> one integrated view for both web and print versions. We are now
> considering changing this so that the print version for larger
> at least will have a separate series listing as well. This is
> with past practice and seems to match better the linear reading
> pattern of
> print documents as distinguished from the non-sequential scanning
> electronic files."
> I remember hearing a rumor a few years ago about possible user
> looking into the issue of where series descriptions are presented
> their impact on end user cognition, but never learned the results
> these (if they existed). If these were not published, have there
> any recent developments or opinions on this issue in regard to
> usability for researchers? I don't recall much like this in the
> Any opinions on this matter are welcomed.
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around