Okay, I think that makes sense (as I say, it's really no my area). So zero should be allowed.
I'll continue on that assumption. I'll most likely do a patch first before commit. Bob. 2008/5/2 Michiel van der Wulp <[EMAIL PROTECTED]>: > Hi Bob, > > The intention of the current implementation is to allow the user to go from > a [non-concurrent composite state] to a [concurrent composite state] and > vice versa by adding or deleting regions. > IMHO this is a nice feature - the user often draws a composite state with > various contents, then discovers that he needs concurrency. > IIRC, if this does not work anymore, then that is regression. > > PS: I will not be able to read email until Sunday night. > > > Regards, > Michiel > > > ----- Original Message ----- From: "Bob Tarling" <[EMAIL PROTECTED]> > To: <[email protected]> > Sent: Friday, May 02, 2008 11:44 AM > Subject: Re: [argouml-dev] There have to be at least two composite > substates in a concurrent composite state. > > > > > > > Okay, I'll continue the enforcement but manage it elsewhere. I simply > > had no view of my own as I'm not so familiar with usage of this > > diagram. > > > > Should I enforce as it tries to at the moment or should I also enforce > > so that there is never zero composite states. > > > > e.g. On creation of the composite state automatically create 2 inner > > composite states and refuse to delete inner composite states if there > > are only 2 left (or an attempt to do so deletes the parent). Which is > > best or still allow zero? > > > > Bob. > > > > > > 2008/5/2 Michiel van der Wulp <[EMAIL PROTECTED]>: > > > > > Hi Bob, > > > > > > > > > > > > > The well formedness rule - "There have to be at least two composite > > > > substates in a concurrent composite state" - seems to be enforced by > > > > the application rather then by crtics. > > > > > > > > > > Yes, the WFR for 2 regions in a concurrent state is enforced. > > > > > > This to make it transparent for the user. The user is kept unaware of > the > > > regions as separate entities in the UML model. > > > BTW: Otherwise it would be very hard to implement: How would ArgoUML > have > > > to react if there were only one? Think about the common user's usecase > of > > > drawing regions. > > > > > > We (the creator of concurrent states and me) decided on this strategy > to > > > reach a solution that is user-friendly (and implementable). > > > > > > > > > > > > > It surprised my to find that the button to create a new Concurrent > > > > Region on a Composite State actually creates 2 if there are currently > > > > zero. > > > > > > > > > > Yes, that is a consequence of this line of thought. > > > > > > > > > > > > > So should I make my job easier by removing the enforcement of this > > > > rule or should I make this enforcement better by managing it in the > > > > model subsystem? > > > > > > > > > > Please, improve this enforcement. > > > This one is for usability. > > > > > > > > > > > > > The enforcement doesn't fully enforce as it currently allows zero > > > > composites, is that good or bad? Possibly it's a balance between > > > > enforcement and critics? > > > > > > > > > > Err.. that must be a bug. How did you do that? > > > Please note that the former checkmark for concurrency on the composite > > > state proppanel is intentionally removed! > > > > > > I see some simularity with swimlanes in an activity diagram, but there > the > > > pool is not drawn first (and independently), so it differs. > > > > > > Regards, > > > Michiel > > > > > > > > > > > > ----- Original Message ----- From: "Bob Tarling" > <[EMAIL PROTECTED]> > > > To: <[email protected]> > > > Sent: Friday, May 02, 2008 3:03 AM > > > Subject: [argouml-dev] There have to be at least two composite > substates in > > > a concurrent composite state. > > > > > > > > > > > > > > > > > > > > > > > > > The well formedness rule - "There have to be at least two composite > > > > substates in a concurrent composite state" - seems to be enforced by > > > > the application rather then by crtics. > > > > > > > > As you probably know its my opinion that we should seek some balance > > > > to enforce or hint different rules where it really aids the user to > > > > get the design right first time. > > > > > > > > So here is one to discuss. > > > > > > > > It surprised my to find that the button to create a new Concurrent > > > > Region on a Composite State actually creates 2 if there are currently > > > > zero. > > > > > > > > There is also the nasty hack to delete the remaining one if we get > > > > down to just one (that probably won't work under all circumstances). > > > > > > > > So should I make my job easier by removing the enforcement of this > > > > rule or should I make this enforcement better by managing it in the > > > > model subsystem? > > > > > > > > The enforcement doesn't fully enforce as it currently allows zero > > > > composites, is that good or bad? Possibly it's a balance between > > > > enforcement and critics? > > > > > > > > Bob. > > > > > > > > --------------------------------------------------------------------- > > > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > > > > > > > > > -- > > > > No virus found in this incoming message. > > > > Checked by AVG. > > > > Version: 7.5.524 / Virus Database: 269.23.7/1409 - Release Date: > > 1/05/2008 > > > 8:39 > > > > > > > > > > > > > > > > > > > > > --------------------------------------------------------------------- > > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > > > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > -- > > No virus found in this incoming message. > > Checked by AVG. > > Version: 7.5.524 / Virus Database: 269.23.7/1409 - Release Date: 1/05/2008 > 8:39 > > > > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
