Re: [jats-list] XSD and @noNamespaceSchemaLocation

["Kevin Hawkins kevin.s.hawkins@xxxxxxxxxxxxxxxxxx" <jats-list-service@xxxxxxxxxxxxxxxxxxxxxx>]

If we take the argument of the unnamed editor of the XSD specification
that we can't trust a document to tell you it's valid, we also can't
trust the document to identify its format more generally.  That is, a
document might claim to be JATS green but actually be blue, or it might
claim to be JATS but actually conform to a pre-JATS NLM DTD, etc.

So if I am suspicious of documents, as this unnamed editor of the XSD
specification is, I wouldn't rely on the URI of the DTD as a magic token.

--Kevin

On 9/25/15 4:21 PM, Maloney, Christopher (NIH/NLM/NCBI) [C]
maloneyc@xxxxxxxxxxxxxxxx wrote:
> Okay, thanks!  I added this comment:
> http://www.niso.org/apps/group_public/view_comment.php?comment_id=601
>
> Ibd just also respond to the point about the bland of the liarsb.
>
> There are other reasons for a document to identify itbs schema, other
than
> to allow for validation. The identification lets consumers know what can
> and canbt be done with the file. In that respect, the URI of the DTD (the
> system identifier) or of the XSD should be considered, not as the address
> of a schema document, but rather as a canonical identifier, indicating
> (just like a "magic numberb does) what the file format is.