[tei-council] Preparation for F2F meeting

[lou]

We'll discuss this later, but I think it might be helpful to specify 
more exactly the scope of the proposed validation service. Does it give 
the punter anything beyond what they can readily get by running jing (or 
whatever) for themselves?

And when we say "valid", does it mean "valid against TEI-ALL"? "valid 
against TEI-ALL plus the schematron rules defined by the TEI"? 
semantically valid (TEI-conformant as defined in P5)"  i.e. respecting 
the intended semantics of TEI tags?

An ODD file itself has to be valid against TEI-ALL, in some sense, and 
also against the TEI-ODD odd. Beyond this set of constraints (readily 
checked by existing validators), there are some more recondite areas 
such as those which Roma tries to provide for in its so-called 
sanity-check, ensuring that (for example) every element retained in an 
ODD-derived schema is either referenced in some content model, or a root 
element.

And then there are  constraints which might be added in a pureODD world 
but which cannot be validated to the same extent by different schema 
languages.

I think this may be  an excellent area for future work! Thanks for 
proposing it!


On 16/11/14 14:46, Peter Stadler wrote:
> Am 16.11.2014 um 09:40 schrieb Martin Holmes <mholmes at uvic.ca>:
>
>> I think if you're offering validation, it's no use unless you provide
>> good feedback, so I would offer as much detail as possible.
> Yes, sure. I only wonder if that needs to be the first step or if we can’t start with valid/invalid.
> That simple feedback would already allow to use the validation service in various processing pipelines where you just want to filter for valid or invalid files. (That’s definitely something for machines, a human reader needs more details.)
>
> Best
> Peter
>
>