[tei-council] P5 use of rend in examples

Piotr Bański bansp at o2.pl
Mon Nov 14 04:04:03 EST 2011

On 14/11/11 05:52, Martin Holmes wrote:
> On 11-11-13 05:34 PM, Sebastian Rahtz wrote:
>> I have said it before, but I say it again, I think this is Wrong Wrong Wrong. @rend is defined
>> as an unbounded set of whitespace-separated tokens.  Usage like "font-style: italic", using two tokens
>> which must (implicitly) be conjoined to make up the desired effect, is simply abuse in
>> my interpretation of @rend. I regard it as saying "you can describe the look of this beast
>> using as many words as you like from a vocabulary", not "write a sentence describing the look,
>> using the vocabulary".  As a processor of this stuff, how do I reconcile<hi rend="font-style: italic">
>> with<hi rend="bold underlined red">?
> You never do both in the same project, and the encoding description can 
> specifically state that the project uses CSS code in its @rend attributes.
>> I don't suppose I will make you and Syd change your minds, however :-}
>> Depressingly, the Guidelines say "Although the contents of the<att>rend</att>  attribute are free
>> text", which is not technically true  - it is
>> <datatype maxOccurs="unbounded">
>>          <ref name="data.word"/>
>> </datatype>
> There's nothing that says that data.words must be interpreted as 
> discrete, with no relationship to each other, is there?

Only indirectly, by mentioning "tokens" in the description, and the most
straightforward interpretation of the term 'token' in this context would
be "font-style: italic". But there is no strong intuition behind that
that would preclude arguments to the contrary, I'm afraid.


More information about the tei-council mailing list