Syd Bauman wrote:
>DO> P.S. In terms of modifying the dtd: why? I generally find it
>DO> useful to try and work within the dtd... and rarely have found
>DO> something that couldn't be done.
>Hmmm... I half agree. I generally find it useful to try to work
>within the "vanilla" TEI schema, but often find something that can't
>be done, at least not properly, without extending it. The TEI schemas
>are both designed for, and intended for, extension.
It's true; but I'd also suggest people think in more detail
sort of extension they mean. I think that there are three sorts
a) when you customize the TEI by making it smaller (select
remove attributes, constrain datatypes etc). Your documents
are 100% pure
b) when you create shorthands or renamings for editing or I18N
convenience (I'm writing
in another window using "<uList>" instead of <list
Your document is not pure TEI, but can become so with a
c) when you encode something which the TEI simply does not
cover at present
(my current example is <death>, parallel to <birth>). Your
easily be made into portable TEI.
I suggest the confusion is between b) and c). When you make
called <prefatoryRemark>, it looks at first like class c),
but actually its
a class b), as it could be transformed to <opener
in all likelihood.
Yes, I know I am teaching some of you to suck eggs here, but
I do want us
to distinguish between "customization" and "extension". They
are not the
Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
OSS Watch: JISC Open Source Advisory Service