>Martin Bryan wrote:
>> Re comment in 1. Introduction that "Someone should be able to store DTD's
>> without going through modelling processes" I wholeheartedly endorse this.
>> Writing a DTD is a modelling process. It should not be necessary to
>> duplicate this effort using UML.
David S. Frankel replied:
>I would like to defend the Technical Architecture document on this point.
>It says: "A mechanism for deriving XML syntax from UML models in a
>consistent manner is a requirement."
>UML is a vastly richer medium for modeling the semantics of business data
>and processes, as compared with UML. This is not a criticism of XML. It is
>just an acknowledgement that UML and XML were invented for different
>purposes. Deriving an XML syntax from a UML model in a consistent manner
>means you *don't* repeat the modeling process; rather, you use the superior
>modeling medium to model the domain and then generate the XML syntax
>according to a well-defined pattern.
I didn't think Martin was objecting to deriving XML syntax from UML
models, but to a *requirement* to use UML. (Correct me if I missed
your point, Martin.)
UML is fine if a company has somebody who knows it,
but why not have simpler mechanisms for those who don't?
E.g. write DTD's by hand, or provide a fill-in-the-blanks
interface as I thought Core Components was developing?
-Bob Haugen
Logistical Software LLC
