On 13-11-23 03:01 PM, Saašha Metsärantala wrote:
> I have not been able to be active on this list for a while, but I am
> back now :-) and I hope to be active again!
>> If a file validates, then it is perfect.
> Maybe ... if the schema is perfect, which it seldom is! Let's keep in
> mind the following fundamental insight applying to any non-trivial XML
> vocabulary I know of: An instance XML document which does not validate
> is suboptimal, but an instance document needs more than just validate to
> be considered optimal (except in extremely trivial XML vocabularies, but
> I do not consider that TEI is trivial).
> There is a huge number of XML vocabularies. When I come in contact with
> a new XML vocabulary, I almost always feel disappointed about not
> finding a SHORT summary presenting this vocabulary for XML experts. The
> TEI was no exception at all: I did not find any one-page-long document
> called something similar to "A short introduction to TEI for XML
> experts" (assuming a good knowledge of XSLT, xs: datatypes, namespace
> techniques, customizations, etc.). Such a document would be extremely
> welcome by (thousands of) XML-acquainted people who want to check
> whether (and maybe discover that) TEI may be useful for their purposes.
I wonder if Wikipedia would be a good place for such a thing. The
current TEI page:
is still a slightly odd compromise between detailed instances of tagging
and a quick gloss of some high-level technical info (mea culpa -- I see
I was working on it myself a while ago).
But in all honesty, it's hard to capture the purpose, range and
technical sophistication of TEI in a single page. The fact that
customization is so front-and-centre in our toolset arises out of the
complexity and range of humanities documents, and some understanding of
the userbase and range of needs is really essential; someone who is an
"XML expert" but not a humanist would most likely be puzzled by this
without substantial background information.
> This "TEI for XML experts" page should give the information its title
> suggests and also contain two links:
> 1) one link to a set of appropriate example instances (TEI documents)
> 2) one link to the schema(s) of the vocabulary (TEI in this case)
> The set of example instances should contain a MINIMAL number of
> realistic TEI documents written by TEI *EXPERTS* for exactly this
> purpose and each of these TEI documents should be as short as possible.
> Nonetheless, they should contain all the most significant use-cases and
> show best-practice for EVERY (non-deprecated) element and attribute of
> the vocabulary. In other words: Reading (the XML source of) these TEI
> instances should give any XML expert a good knowledge to start using the
> vocabulary. Of course, I am aware that some XML vocabularies (TEI among
> others) contain quite many elements and attributes, which makes it
> difficult to limit the size and number of example instances.
> Nonetheless, any redundancy should be avoided here.
> The schemas are useful not only to automatically validate documents.
> Reading (the XML source of) a well-written XML Schema is also the
> easiest way to acquire a deep and accurate knowledge of an XML
> vocabulary, but this reading needs to be combined with the one-page-long
> "TEI for XML experts" summary and the well-written example instances.
> Let's keep in mind that the XML world contains much, MUCH more than only
> ONE vocabulary and the possibility to combine these vocabularies is one
> of the very good reasons for choosing XML as a meta-format!