[tei-council] What did we decide?

[David J Birnbaum]

Dear Lou (cc Council),

Thanks!

 > where does apostrophe mark a plural in English? /It disambiguates 
phrases like "the dogs' paws" and "the dog's paws"

/Call me an orthographic or linguistic pedant, but the "s" marks the 
plural in "the dogs' paws" and the apostrophe marks the possessive in 
both "the dog's paws" and "the dogs' paws". That is, although the 
relative place of the apostrophe and the "s" distinguishes the singular 
from the plural, it is not the apostrophe, but the "s", that marks the 
plural.

 > "Parentheses and other marks of suspension" should omit "other" /Why? 
"suspension" is being used in a linguistic sense here

/This isn't the meaning of suspension that occurred to me when I read 
it, but I understand now.

 > /"Element foreign" has some "include" notes /Sorry, I dont understand 
this comment

There were places (here and elsewhere in my sections) where there were 
notes that said something like "Include xxx" (with a title or number or 
something in place of the xxx). That seems like a clumsy sort of link; 
shouldn't we either actually include the text or word the links as 
something like "more information about xxx is available at yyy"? 
"Include xxx" sounds like an instruction in a processing environment, 
not part of guidelines or documentation.

 > Stray leading backslash in "\list = element list { list.content, 
list.attributes }" /Not a stray: RELAXNC needs it
/
Oops. Sorry!

 > do we need separate data.TruthValue?? and data.xTruthValue classes? 
Isn't this like specifying language where it has to be European, i.e., a 
formal restriction where user error is unlikely? /Not sure I understand 
this analogy /

I think what I meant was that we might consider having a single class 
that accepts booleans, "unknown", and "inapplicable", and that we don't 
have to have a separate boolean-only class because users who want the 
more restricted boolean-only option are likely to select "unknown" or 
"inapplicable", and don't need to be protect from it by the schema. I 
think the point of the analogy was that a project that requires us to 
specify the language of some text doesn't usually provided a narrowly 
constrained list of languages, and trusts instead that the user does not 
need to be protected from picking an inappropriate one, and that one 
might take the same attitude toward the two boolean-like classes. I 
don't feel strongly about this, but it did seem as if we were using two 
classes where one would probably meet our users' needs.

Best,

David
/
/Lou Burnard wrote:
> Thanks David.
>
> If you check
> http://tei.oucs.ox.ac.uk/trac/TEIP5/ticket/33
> and
> http://tei.oucs.ox.ac.uk/trac/TEIP5/ticket/54
> you'll see that I've at last got round to your helpful comments on 
> those chapters at least.
>
> L
>
>
> David J Birnbaum wrote:
>> Dear Lou (cc Council),
>>
>> My memory about this is a bit foggy, but I thought someone suggested 
>> in Berlin that since TEI conformance requires an ODD, the Guidelines 
>> should say nothing about modifying the TEI otherwise than through the 
>> use of an ODD. That is, although other types of modification can be 
>> implemented by someone perverse enough to try, the Guidelines 
>> shouldn't get involved in (or even acknowledge the existence of) such 
>> non-conformant endeavors.
>>
>> I don't remember whether we agreed on this or whether it just floated 
>> around the table, slipped out under the door, headed down in the 
>> pater noster, and disappeared across the square, but whatever the 
>> disposition, I think it would 1) make our lives easier and 2) make 
>> the lives of those who wish to modify the TEI easier.
>>
>> Best,
>>
>> David
>>
>> Lou Burnard wrote:
>>> I thought we had some discussion in Berlin about what the Guidelines 
>>> should say about the feasibility of constructing a TEI schema by 
>>> non-ODD methods, e.g. by lumping together references to TEI dtd 
>>> fragments within a doctype, or building a relaxng schema by 
>>> referencing the TEI Relaxng modules.
>>>
>>> I cannot however find any reference to it in the minutes (instead 
>>> they seem to document all sorts of minor corrections which we not 
>>> only decided but dealt with in the meeting itself)
>>>
>>> Before I go ahead and rewrite history, here's what I believe the 
>>> party line to be:
>>>
>>> 1. TEI Conformance (whether implicit  or algorithmic) requires that 
>>> there be an ODD. Therefore if you don't have an ODD, you cannot be 
>>> TEI conformant.
>>>
>>> 2. Nevertheless, we do create TEI modules as distinct schema/DTD 
>>> fragments, with published names (someone was actioned to review the 
>>> names I think? sebastian?) and document said names and modules.
>>>
>>> So, if you're smart, you can go ahead and build yourself a TEI 
>>> aschema without an ODD. But we don't promise it will always work -- 
>>> at release 1.n we might (for example) move an element from one 
>>> module to another.
>>>
>>> Argal, the discussion in ST.xml#STIN should refer only to ODDs? 
>>> (which is where I came in)
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> tei-council mailing list
>>> tei-council at lists.village.Virginia.EDU
>>> http://lists.village.Virginia.EDU/mailman/listinfo/tei-council
>>
>>
>