Showing Tags and Constraints in Activity Diagrams

[tim]

Hello,

we employ specialized Acitivity Diagrams for modeling various things ... For this, we have defined a lot of constraints and tags ... that can be entered, but are not shown in the diagram. Generally, I noticed that tags and constraints are rendered for some elements, while they are not for others.

Specifically, we need to show constraints on Actions, Pins, and ParameterNodes. Furthermore, displaying tagged values in Actions and Activities would be very useful as well. Could that be added ? Would be great!

WBR
Tim

[«Midnight»]

Let's see if I can get you started.

displaying tagged values in ... Activities would be very useful...

Try this:

• Create an Activity and add a tagged value; perhaps a constraint as well.
  
• On the diagram properties check Show Tags (and perhaps Show Requirements and Show Constraints - you know the drill).
  
• Right-click the element and choose Advanced Settings / Use Rectangle Notation. Take a moment at this point to wonder why you cannot do this from the Element main menu; just another of the imponderables of the universe.
  
• Voila, you can see your tagged values, and constraints if applicable. The class glyph will be decorated with an Activity shape in the NorthEast corner, just to help you along.

Of course, the above recipe does nothing for you if you start with an Action. Nor does Set Feature Visibility work for Actions and Activities (and several other element types as well). In fact, turning off visibility with Set Feature Visibility seems to have no effect on the rectangle notation for Activities.

Please consider raising a feature request for visibility of tags, constraints, etc. on non-structural elements. You might also request that the Set Feature Visibility function apply to these as well. At Sparx' suggestion do this from within EA via Help / On-line Resources / Request-a-Feature. This gets the request straight into the Sparx hopper where it will be considered on its merits and adopted or explicitly rejected, rather than as a 'suggestion' that might or might not make it onto their radar.

David

[tim]

Let's see if I can get you started.
Try this:

• Create an Activity and add a tagged value; perhaps a constraint as well.
• On the diagram properties check Show Tags (and perhaps Show Requirements and Show Constraints - you know the drill).
• Right-click the element and choose Advanced Settings / Use Rectangle Notation. Take a moment at this point to wonder why you cannot do this from the Element main menu; just another of the imponderables of the universe.
• Voila, you can see your tagged values, and constraints if applicable. The class glyph will be decorated with an Activity shape in the NorthEast corner, just to help you along.

Of course, the above recipe does nothing for you if you start with an Action. Nor does Set Feature Visibility work for Actions and Activities (and several other element types as well). In fact, turning off visibility with Set Feature Visibility seems to have no effect on the rectangle notation for Activities.

Please consider raising a feature request for visibility of tags, constraints, etc. on non-structural elements. You might also request that the Set Feature Visibility function apply to these as well. At Sparx' suggestion do this from within EA via Help / On-line Resources / Request-a-Feature. This gets the request straight into the Sparx hopper where it will be considered on its merits and adopted or explicitly rejected, rather than as a 'suggestion' that might or might not make it onto their radar.

David

Thanks, works nicely for Activities. I think I will send them a feature request as you suggested, also because this is not the first time I needed it ...

Best regards,
Tim

[Paolo F Cantoni]

[size=13][SNIP][/size]

• Right-click the element and choose Advanced Settings / Use Rectangle Notation. Take a moment at this point to wonder why you cannot do this from the Element main menu; just another of the imponderables of the universe.

[size=13][SNIP][/size]

David,

im·pon·der·a·ble : That cannot undergo precise evaluation

Thus it is NOT an imponderable - it is precisely ponderable!  It is that way by design!  It could just as easily (in my view MORE easily) have been designed as you (and apparently the rest of us) expected it.

EA's UI [Unique Interface] is, in part, caused by fundamental separations between elements which are (to some of us) understandable, but wrong.

In my view, it is a case of an incorrect class hierarchy - which has emphasised the difference in classes, rather than their similarity:

Differences between Shapes and Connectors, between Stuctural Shapes and Behavioural Shapes etc etc.

When I give modelling classes, I ask students to use the "Show Cause" admonition when trying to decide if a new class is needed:

"Show cause" why you should be created.  Why can't your job be done by someone already created?  

As part of this process, you can often find you can refactor the hierarchy and discover similarities that may not have been so prominent before.

For my own part, I have mentioned elsewhere that I have been creating a single "point of interface" for UML type metadate (in XML form).  In that interface, EVERYTHING inherits from Element.  Thus, it is possible!  This design gives one great power as when some new functionality has to be implemented, one can test implement in the specific place needed and then promptly refactor it as "high up" the hierarchy as it is appropriate and it is thus available to all descendants of that node.

Hopefully vX.Y of EA will take the same view...  I suspect 6.5 won't, but we live in hope...

Paolo

[«Midnight»]

You're preaching my own sermon here Paolo, and I certainly agree.

Thus my suggestion to take a moment to "wonder" about things; it helps us to maintain some grip (however tenuous) on our sanity while we wait. We have a pretty good tool in EA, and the Sparx folks are clearly listening. I too would love to see a sudden across-the-board resolution to the underlying model, etc. But just where they'd find the time remains a mystery for now. I suppose it would give them something to fill the empty hours.

I was a (very) early adopter of CWM and other metamodels (some oriented around MOF as well as others). Thus the UML 2.0 metamodel was a breath of fresh air. Still work to be done of course, but one has to start someplace; we could have waited forever for a "better" metamodel, and still it would not be right.

Meanwhile, I think you are being too hard on our current community of developers  

Think about it. For some time we've seen persistent statistics along the line of 10% development versus 90% maintenance for software. This seems to persist no matter how much the rest of the business changes. Maintenance personnel have an ever-improving suite of tools and techniques available to assist them, and still application software remains full of bugs.

So here's my point. Strikes me that the poor developers have to create all those bugs, and somehow they manage to accomplish this herculian task in only 10% of the programming effort. Through the years they have continuously improved their capabilities, and thus have been able to continue to develop and insert enough errors to keep the maintenance folks busy for the remaining 90% of the effort, even with better tools etc.

Really Paolo, you have to respect that kind of innovation and dedication.

[Of course, since the maintenance folks are often referred to as debuggers because they take the bugs out, one can clearly see what we should call those folks who put the bugs in. I'll leave the reader to do the math.]

Like I say, this kind of mental chewing gum keeps me (ever so slightly) sane while I wait for something (anything!) to improve.

Life continues. Sigh...

David

[Paolo F Cantoni]

Your point is well taken, David.

I guess the difference between us is reflected in our taglines (signatures).  I like yours, and I think it's something to ponder on at the end of the day.  But, it seems more passive than mine.

Maybe it's the rapidly approaching question of my mortality that makes me want to be more action oriented!

Keep up the good work!

Paolo

[«Midnight»]

But I'm a happy cynic Paolo.

I understand what you mean though. I've just passed my 39^th aniversary of using computers (a bit less for professional reasons) and I find myself looking in the rearview mirror more often than I used to.