Scott,
It basically comes down to the fact that not all UML tools
support it. I think we should still recommend it, just not make it
mandatory.
jim
From: Scott Came
[mailto:]
Sent: Thursday, March 24, 2005
11:27 AM
To:
Subject: RE:
[legalxml-intj-exmndr] Recap of issue on domain modeling
Yes, this more accurately reflects what we agreed to.
Jim,
perhaps you can offer an argument as to making XMI optional?
> My
minor suggested tweaks are included below in bold.
>
>
> ---
Ellen
>
> The information transmitted is intended only for the
person or entity to
> which it is addressed and may contain confidential
and/or privileged
> material. If you received this in error, please
contact the sender and
> delete the material from any computer.
>
>
>
> ________________________________
>
>
From: Scott Came [mailto:]
> Sent: Thursday,
March 24, 2005 11:13 AM
> To:
> Subject:
[legalxml-intj-exmndr] Recap of issue on domain modeling
>
>
> Here is what I took as consensus from our discussion this morning.
I
> agree with this; Tom, since you were the other active participant in
the
> discussion, could you please post back with your agreement as well,
or
> suggested modifications. Others feel free to chime in as well. I
will
> then make sure this gets into the Domain Modeling section in the
MNDR.
>
> (Again, keep in mind, we're saying what's necessary for
compliance with
> the MNDR, not that the whole world should do what we
say.)
>
> 1. A domain model must be created.
>
> 2.
The domain model must be at least one of: a spreadsheet, a
> concept map,
or a UML class diagram. (What these things are, in detail,
> will be
described in the document.)
>
> 3. The domain model should be a UML
class diagram.
>
> 4. If the domain model is a UML class diagram,
then the diagram must be
> supplied in XMI and image (JPG or PNG) format.
The image format must
> not contain symbols or notation that are not part
of standard UML class
> diagrams.
>
> 5. Nothing in this
specification precludes a domain model from being
> documented, published,
or displayed using diagramming notation other
> than standard UML,
provided that the domain model also be supplied
> according to the
provisions of rule #4.
>
> 6. The domain-GJXDM mapping artifact
must not rely on (replaces
> "reference ") any domain model concept or
element that is not
> represented in standard UML notation. Proprietary
concepts may be
> included in the mapping artifact as long as they do not
determine the
> mapping results.
>
> Thanks,
everyone.
> --Scott
>
>
