[tei-council] datatype issues (part 1)

Lou Burnard lou.burnard at computing-services.oxford.ac.uk
Sun Sep 11 07:58:50 EDT 2005


Warning: this post contains unintended innuendos. Oooo mrs.

James Cummings wrote:

>
>
> Is there a way to have someone put 'm' and it be understood in the 
> schema as iso 5218's '1'?  (Or put 35yrs and it be understood as P35Y 
> ?)  Or does this just defeat the point of standards...
>
Yes, it defeats the point of the exercise. Think dates. You can say 
<date>wibble</date> if you like, but  once you say <date 
value="xxxx">wibble</date> your value MUST conform to what the relevant 
standard says it should be.

Similarly, with the (one or two) places where sex rears its head **as a 
coded attribute value**.  It's supposed to be a normalised value: ISO 
has normalized in this particular way. If you really want to retain the 
ability to say sex="some-arbitrary-code-i-just-invented" that's a 
different attribute. We could have both sex and ISOsex I suppose, but 
maybe that's going too far.

Look at it another way. I have a very fine tool which knows how to 
process ISO sex values when it sees them, iff they are expressed in a 
conformant way. I will not be best pleased if it now has to recognise 
some arbitrary other set of values as well: in fact, I probably won't 
bother at all. It's worth my while as a developer to fit in with ISO's 
whims: I don't know whether or not the TEI is that important yet. 

My argument is that iff we are going to go to the trouble of normalising 
attribute values, and there is a pre-existing international norm, then 
we really have to have much better reasons not to follow it than to say 
"it's not intuitive". Says who? If you're arabic or chinese, why is "m" 
more intuitive than "1"? (or "u" than "0")?

We could have much the same argument about whether we should replace the 
required values of the xsd "boolean" datatype ((which are "true" and 
"false") with "0" and "1" or "yes" and "no" ...

Lou

p.s. you could in your ODD application redefine tei.data.sex as a 
different set of values, of course, tho I think you're making life 
difficult for others if you do








More information about the tei-council mailing list