xml-dev - Re: Gutenberg Project <longish>

[Rick JELLIFFE <ricko@allette.com.au>]

Sebastian Rahtz wrote:
> 
> Ann Navarro writes:
> 
>  > The project shouldn't require a Masters degree in DTD authoring, nor years
>  > of expertise in SGML or other markup technologies. It *should* be very easy
>  > to do.

I found TEI Lite to be very usable, much more so than DOCBOOK, which I
found to be unusable (too many tags, names from Mars, too little
documentation) at that stage. 
I know someone with years of SGML experience but who has a impairment
that makes it difficult to remember complex names: that person found TEI
very straightforward.

I had used TEI for a large dictionary project several years ago, but
that was just taking documents and using them. It was great; probably
the easiest DTD I have used in the last 12 years.

Last year I did a trial run of putting some Chinese poetry, Yuefu Shiji
(Wefu Shrzi), into TEI Lite. (Academia Sinica has a very large text
database of ancient Chinese text; the excercise was to prove that we
could quickly generate/convert the database into XML and TEI when we
needed to).   I found the TEI lite documentation to be pretty clear, and
TEI Lite was vey nice and terse to use.  The example can be seen under
http://www.ascc.net/xml/en/utf-8/i18n-index.html (you will need IE5)
click the "Yuefu Shiji" link.

It is hard for me to think of how to make a simpler, but useful, DTD
that TEI. Where TEI falls down for casual use is that it does require a
good deal of metadata: header information for saying who transcribed it
and cpyright and so on. But a slacker version can easily be made, or
authors can be just given the body elements and whoever puts it into a
repository can add the headers at registration-time.  But for actually
marking up text, TEI is very fast.

I would have thought that the last thing that people need is a new DTD
for text. People almost inevitably create a DTD that looks like the
things they are familiar with, in any case.  HTML's core is based on the
AAP/ISO DTDs around in the late 80s. TEI seems to have been a more
thoughtful, systematic and innovative initiative. 

I would be very interested in knowing why TEI Lite was rejected. A
project which says "we need to create a standard convention for
literature markup" (if that is what Project Gutenberg does) has to
justify why it rejects the currently existing one, which has been proven
useful for non-markup people to use for many years.

Rick Jelliffe

***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************