Re: Proposal to close ISSUE-82: UML diagrams

From: "Elisa F. Kendall" <ekendall@sandsoft.com>
Subject: Re: Proposal to close ISSUE-82: UML diagrams
Date: Wed, 16 Apr 2008 11:51:32 -0700

> I am on holiday this week with my family, and was unable to call in
> today, but find this completely unacceptable.  

> The UML should, at a
> minimum, be correct UML given that there are folks in this working
> group willing to assist to make it correct, including but not limited
> to me.  

Sure, the UML should be valid UML, I wasn't proposing that it not be.

As far as I can tell, making the UML valid involves, the suggestions
that Conrad made
1/ removing the <<set>> stuff
2/ moving the association names to the middle of the links
3/ changing dashed lines to solid lines
4/ changing <<list>> to  ordered
However, I'm not a UML expert so I was suggesting that a UML expert take
a role in making the diagrams be valid UML.

> Peter Haase (who created the metamodel in the first place) and
> I have discussed working together to revise that as well. 

Revising what?  

> I am also willing to assist in educating folks in the working group
> about the purpose of a metamodel, and why it is ideally suited to the
> purposes specified in the FS&SS.  

Sure, go ahead.

> Leaving it as is will only create
> confusion among UML savvy folks who are looking to this working group
> for leadership. 

Leaving what as is?

> Elisa

peter




> Peter F. Patel-Schneider wrote:
> 
> >I propose to reject alignment between the UML diagrams in the Syntax
> >document and the OMG ODM, as the purpose of the diagrams in FS&SS is
> >very different from the purpose of the OMG ODM.
> >
> >I propose to accept that the diagrams may not be not ideal UML and
> >invite the raiser of the issue to make specific suggestions for changes
> >to them.
> >
> >Peter F. Patel-Schneider
> >Bell Labs Research

Received on Wednesday, 16 April 2008 20:29:06 UTC