Re: [XSCH, ALL] some detailed comments from datatype note review

Evan,

Thanks for your comments. 

[...]

> Detailed comments:
> 
> - Section 1.3: Editorial/presentation issue - In the definition for an
> "OWL datatype interpretation" are the words "for each supported
> datatype URIref u w.r.t. D" intended to be subscript?  They rendered
> this way on every browser I tried.

They are intented. To improve the visual appearance, we could rephase it as

"LV (only) contains PL and  the value spaces for each datatype in D."


> - Section 1.4: In the definition for a "unary datatype group" the term
> "primitive base datatype" is used.  What is the qualifier "primitive"
> meant to convey here?  It seems to me that these are merely datatypes
> in the group which are not derived from other datatypes in the
> group. "base datatype" seems sufficient to convey this.  The current
> wording could be interpreted to denote XML Schema primitive datatypes,
> which is inconsistent with the example.

Each datatype is the  base datatype of its derived datatypes; e.g., in Example 1A, integer is the base datatype of humanAge. A primitive base datatype in a unary datatype group can be different from a primitive datatype in a type system because we only support datatypes in the datatype map D. In other words,  it is possible that only part of the derivation tree are supported; hence, the top datatype in a supported sub-derivation tree is the primitive base datatype in a unary datatype group.

Similar discussion can be found in Section 3.2.

 
> - In the definition for "unary datatype expressions" the text reading,
> " the set of G unary datatype expressions," looks incorrect.  Should
> it read, "the set of unary datatype expressions for G,"?

OK,  will change as you suggest.

 
> - Example 1D.  Cool.  Where and how can someone use this in OWL DL
> descriptions?

OWL does not support user-defined datatypes, which is the motivation of Section 2. A solution to extend OWL DL to support user-defined datatypes should cover

1) a standard way of referring to an XML Schema user defined simple
type with a URI reference, and

2) a formal framework of combining SHOIN with user defined datatypes so that the combined language is still decidable.

We will further revise Section 2 to cover the 2) point.

Greetings,
Jeff

--
Dr. Jeff Z. Pan  ( http://DL-Web.man.ac.uk/ )
School of Computer Science, The University of Manchester


 
> - Section 2.3.  Suggest adding a transition after the first
> paragraph.  Something like: "There are some issues with this
> solution."
> 
> - Section 3.5: In this section the term "primitive-equality" is used
>  to (I think) refer to equality as described in section 3.4.  If this
>  is true, then the term should be introduced in section 3.4 and used
>  consistently thereafter when referring to that concept.
> 
> - should the subsection entitled "Using eq in RDF and OWL" be better
>  titled "The Semantics of Using eq in RDF and OWL"?
> 
> - There is still a note to the editor in this section, "@@@ todo
>  datetime stuff - I think they are all incomparible should check."
> 
> *****
> 
> Evan
>

Received on Sunday, 16 January 2005 14:25:28 UTC