Re: ActivityEdge semantics



Date view Thread view Subject view Author view Attachment view

From: Les Munday (baldrick@ureach.com)
Date: Mon 08 Sep 2003 - 18:55:36 BST


I don't know the answer to this, but if it's ambiguous then I
would not use it.

I am of the understanding that a decision fork may be joined by
having both paths converge on a single activity, instead of
using a decision box to merge the paths.

This would be consistent with creating a decision split with two
paths exiting the same activity.

If this is chanig in UML 2.0, it going to make life very
confusing and create a lot of re-work for existing UML
diagrams.

Les.

P.S. Hope you don't mind if I copy this and post to the Rational
UML users group.


________________________________________________
Get your own "800" number
Voicemail, fax, email, and a lot more
http://www.ureach.com/reg/tag


---- On Mon, 8 Sep 2003, John Daniels (jd@syntropy.co.uk)
wrote:

> Dear UMLers,
> 
> I have a question about the semantics of multiple transitions
(edges)
> leaving an action in an activity diagram.
> 
> In UML 1.5, drawing an action with multiple outgoing
transitions
> implied a decision point, as explained in the spec:
> 
> ----
> A decision may be shown by labeling multiple output
> transitions of an action with different guard conditions.
> ----
> 
> However, my brief reading of the UML 2 "final draft"
superstructure
> spec seems to suggest that the meaning has changed:
> 
> ----
> [4] When completed, an action execution offers object tokens
on all
> its output pins and control tokens on all its outgoing control
edges
> (implicit fork), and it terminates. Exceptions to this are
listed
> below. The output tokens are now available to satisfy the
control or
> object flow prerequisites for other action executions.
> ----
> 
> I take this to mean that drawing multiple outgoing
"transitions" from
> an action in UML 2 implies forking, not decision branching. Am
I
> right?
> 
> All the best,
> 
> --John
> 
> John Daniels
> 
> 
> 
> 
> To remove yourself from this list please mail
puml-list-request@cs.york.ac.uk
> with a message containing the word "unsubscribe".
> 
> 
> 

Date view Thread view Subject view Author view Attachment view