[tei-council] Influential Blogs?

O'Donnell, Dan daniel.odonnell at uleth.ca
Thu Mar 25 18:09:18 EDT 2010



Sebastian Rahtz wrote:
> On 25 Mar 2010, at 21:21, O'Donnell, Dan wrote:
>
>   
>> Lou: this is exactly the conclusion you, Laurent, and I came to a month 
>> or so ago: AccessTEI is a project and it needs a specific instantiation 
>> to work with rather than a way of producing instances. This also frees 
>> us up to work on Tite without worrying about retraining people in India 
>> every time we make a change; and it allowed us to intervene by hand to 
>> remove the inappropriate examples.
>>     
>
>
> I probably agree with you, but to be honest I don't at all understand what you mean
> by "a specific instantiation to work with rather than a way of producing instances"
>
> can you say it again in words of one syllable for me?
>
> and what do you mean by "intervene by hand to remove the
> inappropriate examples"?
>   
Sorry Sebastian, I've not been expressing myself well all week. Let me 
try again: what I mean is that although in the accessTEI contract we 
specified the Tite ODD rather than a specific schema or DTD as being the 
reference for the programme (after some discussion in Council), it 
turned out that that is not really very practical if the ODD is not 
completely locked down. Like any project, AccessTEI wants its schemas to 
change more slowly than the ODD might, especially if we on council pick 
up and look at some of the ideas for development we've been discussing 
here over the last while.

In addition, the problems caused by defining an ODD by exclusion rather 
than inclusion turned out to be quite serious in constructing a training 
programme for the keyers in India: meaning (at the moment) that we 
needed to do some hand tweaking (removing examples, for example, that 
contained elements not found in Tite or used them in alternate ways when 
Tite restricts prefers one method over the other).

So what I meant was that, again like most of us when we start our 
projects, AccessTEI developed a schema and coding guidelines based on 
the way the Tite ODD was on the day it was generated (this is what I 
meant by an instantiation); but it won't be returning to the ODD to 
generate new schemas and documentation (what I meant when I said it was 
better to have "a specific instantiation to work with rather than [only] 
a way of producing instances") for a little while--presumably until it 
is felt that the changes in the Tite ODD are such as to make 
regenerating the ODD and retraining the keyers worth the effort 
involved. This means that we at Council are free to concentrate on the 
issues that would really improve things, without worrying about whether 
making changes in the ODD would require a retraining of the keyers. Some 
of the big things that would make it worthwhile revisiting the ODD and 
regenerating the AccessTEI schema are the things we've been talking 
about: better integration with the DLF coding levels, changing the way 
we make ODDs so that it could proceed by inclusion rather than 
exclusion, working on the way examples are included in ODDs that are 
subsets.

So while initially (at our last teleco, for example) I thought--with Lou 
and others--that it might be necessary to do some work on and locking 
down of the Tite ODD in order to start the Apex training programme, on 
reflection and discussion, it became clear that in actual fact there was 
nothing we needed to do to the ODD--all the issues involved tweaking and 
cleaning up artefacts of the generation process in response to the 
specific demands of the training programme for the keyers. So baiscally 
we generated the schemas from the canonical ODD and then did the kind of 
by-hand removal of examples that don't work so well in Tite that we were 
discussing at Council when we raised the problem a couple of weeks ago 
or so.

-dan
> --
> Sebastian Rahtz      
> Information Manager, Oxford University Computing Services
> 13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
>
> Sólo le pido a Dios
> que el futuro no me sea indiferente
>
>
>
>
>   

-- 
Daniel Paul O'Donnell
Associate Professor of English
University of Lethbridge

Chair and CEO, Text Encoding Initiative (http://www.tei-c.org/)
Co-Chair, Digital Initiatives Advisory Board, Medieval Academy of America
President-elect (English), Society for Digital Humanities/Société pour l'étude des médias interactifs (http://sdh-semi.org/)
Founding Director (2003-2009), Digital Medievalist Project (http://www.digitalmedievalist.org/)

Vox: +1 403 329-2377
Fax: +1 403 382-7191 (non-confidential)
Home Page: http://people.uleth.ca/~daniel.odonnell/




More information about the tei-council mailing list