Mark Olsen <[log in to unmask]> writes:

> I agree completely with Lou and Grace and would again propose to
> the TEI that it consider creation of a fairly large test sample
> of materials -- old and new, lite and heavy, simple and complex,
> SGML and XML -- to serve as a test suite for developers.  Having 
> worked with TEI encoded materials from many projects, I can attest 
> to the considerable variability of TEI conformant documents, between 
> projects and even between documents from individual projects.  
> This is particularly true in terms of header contructs and objects
> like notes, pointers, etc.  I suspect that the collection created
> here from those projects I know of, contacted, and got a favorable
> reply is far from a representative sample -- almost all Americans.  
> I think the TEI and users of TEI encoded resources would benefit greatly 
> from as wide a collection set as can be reasonably assembled.  

I wholeheartedly second this suggestion.  On practical reasons, maybe
we could use the TEI Wiki as a move forward to collect such sample
files (or maybe even pointers to them) with a short description of the
language used, nature and purpose of the project that created it
(customization files if any) etc.  James, where would be a good place
to put this?

This could then periodically be harvested and made available in a
convenient bundle or test-suite for interested parties.

All the best,


 Christian Wittern 
 Institute for Research in Humanities, Kyoto University
 47 Higashiogura-cho, Kitashirakawa, Sakyo-ku, Kyoto 606-8265, JAPAN