[tei-council] Re: 'token' classes

[Lou Burnard]

Yes, precisely. I am hoping to draft a revision of the section of 
chapter ST concerned to address this point a little more firmly later on 
today.

  Laurent Romary wrote:

> I also agree with this option, but I would suggest to tackle it from  a 
> more semantic point of view (instead of pure syntactic constraint  of 
> having white spaces or not in the type attribute).
> 
> What we want (maybe...) is to consider that the type attribute  contains 
> a code or identifier (not necessarily in the sense of xml)  that is 
> related to a predefined classification of a given object. We  should 
> state that strongly somewhere so that it gives encoders the  instruction 
> to actually +think+ of the general picture in which a  given value for 
> type will be considered. Typically, the possible  types of division in a 
> corpus (<div>) the various grammatical  features he wants to use in a 
> dictionary (<gram>) etc.
> 
> Laurent
> 
> Le 18 oct. 05 à 10:41, James Cummings a écrit :
> 
>> Syd Bauman wrote:
>>
>>>> I think we can proceed with two datatypes (Syd's data.name(s) and
>>>> data.token(s), for the sake of argument) and add a third one later
>>>> after review?
>>>>
>>>
>>>
>>> I think that's a good idea. If users all over the world scream for
>>> the capability to put spaces into their type= attributes, we can
>>> rethink this. If you & Christian are the only two, you know how to
>>> customize TEI yourselves ... :-)
>>>
>>> I think Lou will be vindicated, and most no one will really think
>>> it's horrible to say "collar,dog" as opposed to "dog collar".
>>>
>>
>> Just to say that I think having no spaces in type attributes (or  only 
>> having
>> spaces in similar attributes when they are real separate tokens (or  
>> whatever))
>> is a good idea.  I am happy to be convinced otherwise, but it seems  a 
>> perfectly
>> reasonable restriction which will stop people doing <div type="The  
>> chapter where
>> she gets killed"> or something silly.
>>
>> -J
>> -- 
>> Dr James Cummings, Oxford Text Archive, University of Oxford
>> James dot Cummings at oucs dot ox dot ac dot uk
>>
>>
> 
> 
>