[tei-council] specDesc spec vs. reality

[Sebastian Rahtz]

On 9 Dec 2013, at 12:29, Lou Burnard <lou.burnard at retired.ox.ac.uk> wrote:

>>> Regarding (b) I don't think requesting an inherited attribute should be regarded as an error. We don't make that distinction elsewhere (e.g. I can modify an inherited attribute in a elementSpec) and it is certainly not the way I remember things being done before.
>> I am inclined to agree, unfortunately, that if you specify atts=“xml:id next resp” it should go and get those from attribute classes.
> 
> Good. Not unfortunate at all! Please make it so Mr Spock.

one day, when I feel strong. perhaps we should add that to the objectives for TEI Nudge :-}

> <specDesc atts="a b c" withValList="a b">
> 
> defaulting to "no valList display”.

y’know, I’d feel inclined to make display of closed valLists be the default, and forget about
optjons. we barely use them in the Guidelines anyway.

open valLists I’d suggest just ignoring

> If a valList appears in a content model, then  I assume it will always be displayed in the doc: this doesn't affect what happens in a <specDesc> though, which is blessedly silent on the topic of content models.
> 
> While we're looking at this, I have long wanted some way of grabbing <egXML>s from the specs for display in the descriptive prose (or indeed vice versa)

better make them identifiable then….

Sebastian