[tei-council] @calendar and @datingMethod

[Lou Burnard]

The @scheme attribute used to be a nice simple @TEI attribute with 
values "yes" or "no", which distinction is, in practice, all we've ever 
needed it for.

If you look at the spec for <att>, you will see that @scheme is indeed 
defined as data.enumerated, with a suggested list of values, none of 
which we actually use in the text of the Guidelines (<cough/>). It also 
contains the interesting note "A namespace prefix may be used in order 
to specify the scheme as an alternative to specifying it via the scheme 
attribute: it takes precedence". I think this is intended to allow you 
to say e.g. <att>html:href</att> rather than (as we currently do) <att 
scheme="html">href</att>, and to warn you that nonsense like <att 
scheme="TEI">html:href</att> will not work. It doesn't tell you where 
you define that namespace prefix, of course, nor does it need to.

The spec needs fixing hoiwever since one of the examples cheerfully says 
"<att scheme="XX">style</att> is not among them" (i.e. the TEI global 
attributes) so I will add some of the values we do use in the Guidelines 
to the list at the same time.

How, out of curiosity, do you think  translated names for attributes 
should be handled? e.g. in the spec for att.damaged, the Spanish 
translator has decided to translate the content of the <att> element as 
well as the surrounding text, resulting in things like   "<p> La 
etiqueta <gi>daño</gi> con el atributo <att>grado</att> debe ser 
utilizado..."

I think this is just plain wrong, not only because it's inconsistent 
with the practice of other translators, but also because it implies that 
there is somewhere an attribute called "grado", which there isn't.  I 
think this should read:

<p> La etiqueta <gi>damage</gi> (daño) con el atributo <att>degree</att> 
(grado) debe ser utilizado..."




On 21/05/14 10:29, James Cummings wrote:
> On 20/05/14 21:58, Martin Holmes wrote:
>> Great. We should probably do this as part of the Test build, really. But
>> it's conceivable that we might use <att> or <gi> to talk about things
>> from other namespaces.
> Should this take account of the @scheme attribute then? On that
> note I've wondered a couple times whether @scheme should either
> not be data.enumerated or be replaced or complemented with @ns
> for an actual namespace.  Though I suppose you might want both.
> You might want to be able to say:
>
> ===
> <p><gi ns="http://www.tei-c.org/ns/1.0"
> scheme="TEI">persName</gi> allows many elements inside it but in
> <term>TEI-Simple</term> <gi ns="http://www.tei-c.org/ns/1.0"
> scheme="TEI-Simple">persName</gi> has been changed to only allow
> forenames and surnames.</p>
> ===
>
> (Ok, a fictitious example but both these different 'schemes'
> would be in the same namespace.)
>
> -James
>
>