[tei-council] New discussion document on 1.0 release priorities

James Cummings James.Cummings at computing-services.oxford.ac.uk
Thu Jul 20 02:52:36 EDT 2006

Sebastian Rahtz wrote:
> Christian Wittern wrote:
>> * While I am undecided on whether we need a module dependency, I think
>>   we need to avoid producing invalid schemata whenever feasible.  This
>>   includes detecting of (accidental?) deletion of essential elements.
>>   My students here got a schema from Roma yesterday that happily
>>   indicated TEI as the start element, although the element TEI was
>>   deleted (together with the complete textstructure module).   
> There is a law of diminishing returns here, I think. Catering
> for accidental deletion of <TEI> may be going too far.....

While I think I want basic module dependency, I'd worry about us doing 
element level dependency for every element. Is it catering for 
stopping this in specific that we should be worried about?  Or is it 
validation of the resulting schema that we should concentrate on?  It 
just seems better to let the person create whatever schema they want, 
pipe it through jing or something somehow, and if there are errors 
pass them back to the user for them to fix themselves.


More information about the tei-council mailing list