[tei-council] Conformance inconsistency

Syd Bauman Syd_Bauman at Brown.edu
Fri Oct 26 11:16:14 EDT 2007


> Thanks for the fast, interesting reply Conal. 

I realize there are follow-ups to my post of last night to which this
is a reply -- I haven't even read them yet. First, I would like to
apologize to Council for the convoluted logic in that post. I
realized (at some god awful time in the wee morning) that this was
perhaps one of the stupidest lines of reasoning I have ever publicly
traversed.

I was complaining about the definition of algorithmically conformant
based on some bizarre notion that Conal could differentiate James's
myRend= attribute from my myRend= attribute because (of course) they
are in different documents. And they are. 

But the whole point of algorithmically conformant is that if Conal
wants to avoid this sort of headache, James & I should be sending him
our documents *after* *we* have applied the transformation -- we
shouldn't be making Conal do it. He should never see a myRend= at
all.

That is, the point of algorithmically conformant documents is so that
I can have my nuanced, pithier, more precise, or whatever encoding in
the privacy of my own project, or even among some friends. But as
soon as I want to engage in blind interchange, I can (and should)
automatically convert my documents into what I had hoped we would
call "strictly conformant", i.e. is conformant without requiring
transformation.


This doesn't change my underlying concern, BTW. I still think that
the definition of algorithmically conformant needs to exclude the
possibility of stuff added into the TEI namespace, or, far better, we
need to stop dictating what people do in the privacy of their own
projects, and remove the namespace purity rule from algorithmic
conformance.



More information about the tei-council mailing list