[tei-council] Fwd: Tite: header
kevin.s.hawkins at ultraslavonic.info
Tue Aug 10 17:30:28 EDT 2010
My opinion is that this should not be our concern.
When a vendor receives an object to be encoded, it comes with an ID
assigned by the customer or is assigned one as soon as it comes in the
door. This is used as all or part of the filename. The stages in their
workflow are maintained in their own systems, either by file locations,
a CMS, or version control system. Who the customer is is likely in a
database of orders, and who encoded on their end is something they
surely already track for quality-assurance purposes. These workflow
methods need to exist on their end outside of the TEI document -- and
likewise outside of any metadata format -- because they need these same
things for all the work they do.
The TEI's approach to metadata is that it's safer if metadata is
included in a file because it's less likely to ever be separated from
it. However, I think this only matters in a long-term archiving. For
any ongoing digitization or publishing operation, there are other,
better tools for managing this information.
On 8/10/2010 5:20 PM, Lou Burnard wrote:
> Here's a note I sent to Perry, Laurent, and Dan back in March about what
> I consider to be a really off the wall modification introduced into TEI
> I never got any response, which is fine, we're all busy. But I would
> like to know whether anyone else feels as I do, or whether I should just
> keep this particular opinion to myself.
> -------- Original Message --------
> Subject: Tite: header
> Date: Wed, 3 Mar 2010 15:03:40 +0000
> From: Lou<lou.burnard at oucs.ox.ac.uk>
> To: Perry Trolard<ptrolard at artsci.wustl.edu>, Laurent Romary
> <laurent.romary at loria.fr>, "daniel.odonnell at uleth.ca"
> <daniel.odonnell at uleth.ca>
> Why no TEI Header in TEI Tite documents?
> The spec at
> "Tite omits the<teiHeader> element as a convenience to transcribers.
> This departs from normal TEI practice, which requires<TEI> as the root
> element, containing<teiHeader> and text elements. In order to bring a
> document encoded in TEI Tite into adherence with the TEI abstract model,
> projects should add a teiHeader before engaging in post-transcription
> processing. "
> Why is it a "convenience to transcribers" to prevent them from adding
> useful metadata to the text they have described?
> In its absence, how will they indicate
> a) which text this xml document is supposed to represent or be a part of?
> b) when they did it?
> c) what's the status (preliminary, finished, proofed etc.) of this xml
> d) who's responsible for shipping it, paying for it, handling questions
> about it?
> Presumably they will still do those things somehow -- by file naming
> conventions, associated paper work, mutual convention, or other means...
> Why not use the tool at hand for the job?
> No-one is saying that a "tite" TEI Header (or any other) need contain a
> kosher bibliographic description -- for that job we apply the
> rabbinical skills of the cataloguers when the xml document is
> accessioned. But it can still contain all that is needed for conformance
> to the TEI abstract model, namely
> * a title
> * a source description
> * a responsibility statement
> The title might be "text number 9456", the source description might be
> "microfilm batch number 6181032" and the responsibility statement might
> be "Joes Garage" but that would still be better than nothing at all, surely?
> I think the decision to junk the header in this context shows a
> surprising lack of imagination about what it's meant for, and how it can
> be used.
> I suggest we should try a bit harder to propose some minimal header
> structure that would fit in with the work flow anticipated for Tite
> documents. My belief is that Apex would welcome such an idea, though I
> haven't yet asked them!
> tei-council mailing list
> tei-council at lists.village.Virginia.EDU
More information about the tei-council