[tei-council] invalid examples allowed in Guidelines?

Laurent Romary laurent.romary at inria.fr
Mon Jan 10 06:28:25 EST 2011

I think that we sjhould not break up egXML too much, but make it more  
powerful to cope with various similar use cases. As a preliminary  
point, we have the same problem in the ISO work, where we would like  
to state that examples could be validated against the schema that the  
ODD spec is defining (within the same document), and this shoxs that  
we should not focus only on the issue of the TEI namespace per se.  
This leads me to identify two aspects:
- indicating that one should or should not try to validate an example.  
Mimicking the its:translate attribute (cf. http://www.w3.org/TR/its/),  
we could think of something like validate="yes" (resp. "no") when we  
want to indicate that an example is to be validated by whatever  
- indicating against what to validate. We could have the simple  
mechanisms of providing (not changing, thus not using @xmlns) the  
reference namespace for the example within @ns (we do have this  
attribute already, why not use it)

Le 9 janv. 11 à 01:45, Sebastian Rahtz a écrit :

> As you or may not know, every <egXML> in the Guidelines source
> is checked at build time against the P5  schema. This is a good  
> thing, of course.
> While playing with validation of P5 again, I found (reminded myself)  
> that 26
> files have incomplete examples, along the lines of:
>     <msDesc xml:id="DN17">
>        <!-- ... -->
>      </msDesc>
> which is not valid because it misses the mandatory child  
> <msIdentifier>.
> Unfortunately, I cannot see a way to flag some instances <egXML>
> as being deliberately incomplete.
> There are four choices:
>   a) fix the examples so they are minimally valid
>   b) fix the build script so that this class of error is thrown away,
>        which has the disadvantage of  masking genuine errors;
>   c) leave the error messages in place and remember to ignore them  
> each time
>        (very confusing for the unwary)
>   d) convert all these class of examples to CDATA in <eg>
>        so they are not checked at all. This seems a bad precedent.
> We were (until today) in a state of b).
> Option a) is possible, but makes for some over-verbose examples
> (eg
>   <fileDesc>
>         <titleStmt> <!-- ... --></titleStmt>
>         <editionStmt> <!-- ... --></editionStmt>
>         <extent> <!-- ... --></extent>
>         <publicationStmt> <!-- ... --></publicationStmt>
>         <seriesStmt> <!-- ... --></seriesStmt>
>         <notesStmt> <!-- ... --></notesStmt>
>         <sourceDesc><!-- ... --></sourceDesc>
>    </fileDesc>
> would need expanding a lot)
> This was discussed a few years ago but since I have the patient open
> on the operating table, I'd welcome some current options on the
> proper way to proceed.
> --
> Sebastian Rahtz
> Information and Support Group Manager, Oxford University Computing  
> Services
> 13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
> Sólo le pido a Dios
> que el futuro no me sea indiferente
> _______________________________________________
> tei-council mailing list
> tei-council at lists.village.Virginia.EDU
> http://lists.village.Virginia.EDU/mailman/listinfo/tei-council

Laurent Romary
laurent.romary at inria.fr

More information about the tei-council mailing list