RE: [rights-examples] ln00x use case discussion

["DeMartini, Thomas" <Thomas.DeMartini@CONTENTGUARD.COM>]

Title: RE: [rights-examples] ln00x use case discussion

responses below after [TD].

thanks,
thomas.

-----Original Message-----
From: DuCharme, Bob (LNG) [mailto:bob.ducharme@lexisnexis.com]
Sent: Tuesday, July 16, 2002 1:06 PM
To: 'DeMartini, Thomas'; 'rights-examples@lists.oasis-open.org'
Subject: RE: [rights-examples] ln00x use case discussion

>                <proposed:principalClass>
>                        http://www.foo.com/platinumMember
>                </proposed:principalClass>

Does the "proposed" namespace prefix reference a namespace that I'd have to make up, or does it already exist, or as I'm guessing from its name, is it something that you're suggesting may be worth requesting in the spec to address this issue?

[TD] Yes, the latter: something I'm suggesting may be worth requesting in the spec.

(By the way, this is an issue for the spec committee: in all the ContentGuard-produced documentation I've seen, namespaces are referenced all over the place without ever being declared, with two serious consequences: 1. none of the examples in a document like ExampleUseCase.pdf are schema-valid, and 2. someone new to XrML reading one of the documents has no way of knowing where the various element type names are coming from. Not showing the declarations in every example is one thing, but not showing them anywhere--the string "xmlns" doesn't show up once in the PDF file--makes everything much harder to follow.)

[TD] I hear your pain.  In fact, I've gotten so sick of reading specifications that have examples that don't validate, that I hope it never happens here.  Unfortuneately, making sure that all the examples in a document are correct is a time-consuming process because you can only validate the examples as xml and getting the formatting to look good is never a simple cut-and-paste.  This is why the examples are dynamically pulled in.  Also, this allows the reader to have real live xml that validates at their fingertips without having to copy it out (another operation that always leads to errors).  As you read any of the documents, please also have at your side the companion .xml files (which *are* complete).  I know there is nothing in the documents pointing readers to these and as soon as I'm done typing this e-mail I will make the recommendation internally that we do that.  As for the documents we produce out of this SC, I agree, we should learn from this as well and be sure to point the reader at the full xml files to go along with the text (regardless of how much of the xml files actually gets included in the text).

Thanks,
Thomas.