Re: Updated Requirements Document ready for review

Bernard,

Thank you for your careful review and comments. I apologize for the
delay in responding to you. Please see my replies to your individual
points below.

Jeff

Bernard Horan wrote:
> 
> Jeff
> 
> I'd like to suggest the following changes:
> 
> 2.1 Web Portals
> First para, change to:
> "A Web portal is a web site that provides information content on a
> common topic, for example a specific city or domain of interest. A web
> portal [allows...]" -- this avoid the cardinality confusion.

OK.

> Third para, IMO the additional content is a little woolly. In particular:
> "...can provide a terminology for describing content and axioms...",
> "This ontology could include definitions that state things..." and "When
> combined with facts..." (I don't think there's a definition of 'facts'
> anywhere.)

This content was arrived at as a clarification during an exchange on the
public comment list. Would help if we defined some of the terms in the
"What is an ontology?" section? How could I change it to make it less
"wooly?"

> 2.2 Multimedia collections
> 2nd para, last sentence: unnecessary comma.

Thanks.

> 3rd para, first sentence: missing "be" between "would" and "of".

Ok.

> 3rd para, fourth sentence: change "we want to be able to infer" to "it
> should be possible to infer" or some such.

Ok.

> Although, TBH, I preferred the original.
> 
> 2.3 Corporate web site management
> Why have the final bullets been removed?

In general, the "requirements" bullets were removed from use case
sections because, 1) not every use case had such bullets, 2) the ones
that did typically listed requirements that were just as applicable to
other use cases and 3) when listed, the requirements were incomplete. In
general, I tried to communicate the same information as an example of
the use case. If you feel I left something important out, then please
suggest additional text.

> 3.2 Ontology Evolution
> Change first para to:
> "An ontology may change during its lifetime. A data source must specify
>   the version of an ontology to which it commits."

Ok. But I think I change "must" to "should."

> 3.7 Compatibility with other Standards
> How would one measure "reasonably compatible"?

I meant to suggest that compatibility should be pursued unless there are
good technical reasons to avoid it. I would welcome any phrasing that
would better communicate this idea.

> R13. Attaching Information to statements
> Last sentence: this seems out of place here, as it's neither a
> requirement nor a justification.

It is a suggestion of a possible way to satisfy the requirement, not
unlike the mention of the Dublin Core element set in R5.

> 
> regards
> 
> Bernard
> 
> Jeff Heflin wrote:
> 
> > I have attached an updated version of the requirements document in HTML
> > format. New text is in red, and deleted text is indicated by strike
> > through. Comments, especially from the official reviewers, are welcome.
> >
> > Jeff
> >
> >

Received on Monday, 17 June 2002 14:38:58 UTC