[tei-council] <quotation>

[Kevin Hawkins]

Below ...

On 12/11/12 9:49 AM, Gabriel Bodard wrote:
> Aha, I see. Thanks.
>
> Is this "convention of the annotation" then something we might want to
> take advantage of, if we want to make default attribute values useful to
> at least some users?

I want to clarify what this "convention of the annotation" actually is. 
  It sounds like syncRO has implemented a non-standard extension to 
RELAX NG in <oXygen/>.  While proprietary standards are sometimes 
adopted by a community at large and may even be come part of the 
non-proprietary standard, I don't think the TEI should yet tether itself 
to something not adopted elsewhere, especially if it isn't even openly 
documented.

> I have to say that I'm with Wendell, however, and don't like the idea of
> default attributes anyway anyhow. If I don't specify @marks on
> <quotation>, that means I am not specifying whether some or any or all
> quotations marks are preserved or not. (What assumptions you want to
> make about that is another question.) The default value of all missing
> attributes should be @*='unspecified', in other words.
>
> (I suppose a default value on a _required_ attribute might make sense,
> if only as a convenience for data-entry.)

I have always assumed that a default value for a required attribute is a 
convenience for data entry, but I still don't understand what a default 
value for an optional attribute actually means.

--K.