Topics

1

Suggestions and Requests / Vertical activity diagrams (decision nodes, etc.)

« on: January 28, 2013, 09:35:28 pm »

By default, it seems (to me) that EA prefers a horizontal layout for activity diagrams:

• Partitions are drawn horizontally, by default.
• Decision nodes are drawn as taller-than-wide diamonds.

It's somewhat annoying to rotate decision nodes and partitions every time I want to draw them as I like (from top to down).

A project-wide or tool-wide option to set this preference would be ideal for me.

Doesn't anybody else miss this feature?

2

General Board / Namespace in sequence diagrams. How to show it?

« on: February 13, 2013, 02:50:05 am »

Hello.

I am trying to show the namespace for the "Sequence instances" of a sequence diagram, to no avail.

I have checked Diagram -> Show Namespace and Diagram -> Fully Qualified Namespace. The box Diagram -> Disable fully scoped object names is un-checked.

It does not matter if the lifeline is created from the Toolbox and later assigned a class with Ctr + L, or if I drag a class from the browser: no namespace appers.

Am I doing something wrong?

3

General Board / Changing [default] directions in associations

« on: February 08, 2013, 11:47:24 pm »

I'm drawing a big deployment diagram and I'm tired to change every communication path's direction from the default (Source -> Destination) to what I usually need (Bi-Directional) or what I could stand (Unspecified).

1) Is there a simple way to change the default direction of communication paths (or associations in general)?

2) If not, is there a simple way to "batch-change" all the associations / communication-paths to a specific direction?

Thanks.

4

General Board / Active classes

« on: February 12, 2013, 03:56:11 am »

I suspect that EA supports neither the standard notation for active classes ("box with an additional vertical bar on either side"), nor the property isActive, nor a stereotype like <<active>>.

Is that true?

5

General Board / Provided / Required Interfaces and their Types

« on: January 10, 2013, 04:22:23 am »

I have a model I made some time ago, with EA 9.x. In a certain part of it, I defined several interfaces, which I later used in a component diagram, like this:

1) Add new component.
2) Add new port to component.
3) Add exposed interfaces to component, selecting the required and / or provided interfaces from the previously defined interfaces.

I noticed the following behaviour, which I can still see if I open that old  model with EA 10:

a) In the project browser, the provided and required interfaces appear with a name and a type, both equal to the selected (previously defined) interface (such as MyInterface:MyInterface). Fine.

b) If I click with the right button on a provided / required interface (in the diagram), I can select Find --> Locate Classifier in Project Browser.  Very useful indeed.

c) If I click with the right button on an interface (in the browser or a diagram), I can select Find --> Find in All Diagrams, and it finds every use of the interface as a provided or required interface.  Very useful too.

d) Odd enough, there are a couple of provided / required interfaces that do NOT appear to have a type (in the browser), their classifiers  canNOT be located (there is no Locate Classifier option in the Find sub-menu), and they are NOT found when searching the diagrams where their interface types are used.

Moreover, I cannot change those provided / required interfaces in order to reproduce the 'good' behaviour described in a), b), and c), and I cannot create new models with that behaviour (from scratch) with EA 10.

Even now I am not sure if I created the provided / required interfaces in the old model (with EA 9.x) just as I described in steps 1), 2), and 3), or if I did something else or something slightly different.

If I right-click on a port in the old model, and select Structural Elements..., I can see a dialogue like this:

Name       Element           Type      Stereotype Visible Owner
ILocalApp  RequiredInterface ILocalApp            True    app
IRemoteApp ProvidedInterface                      True    app

The second row is what I see for the 'good-behaviour' interfaces; the third row is what I see for the 'bad-behaviour' interfaces: no Type!. And I cannot modify the bad ones, even with the dialogue's 'Modify' button.

How can I set the right Type for these provided / required interfaces?

Is this something that has changed in EA 10?

6

General Board / Find transitions by name

« on: November 21, 2012, 03:19:22 am »

Transitions seem not to be first-order citizens in EA. I cannot find transitions by name with a model search. Yes, I can search triggers by name, but this is not the same.

This is my use case: I have plenty of state diagrams, inherited from an ancient tool, that gave each transition a name (such as TMAS2S32). The trigger, guards and effect of each transition was given in text form with tables.

The name gives us a unique identifier for each transition, which has two benefits: it is easily searched, and can be used for tracing purposes (in documentation, code, etc.).

But I have just found that search by name, apparently, does not work for transitions. Is this true? Is there a workaround?

Thanks.

7

General Board / Abstract operation modification not propagated

« on: November 20, 2012, 02:22:13 am »

I have an abstract class Parent with and abstract operation doSomething (parameter :ThisType).

I also have a concrete class Child that specializes Parent. When I draw the generalization relationship, EA lets me specify the operations from Parent that Child will implement. I check doSomething and Child now has a doSomething operation with the same signature as Parent. That's good.

Now, if I modify doSomething in Parent (for instance, I change the type of parameter to ThatType), the change does not propagate to Child.

Is there a way to perform this propagation, or do I have to manually enter the change in each child?

I could break the generalization relationship and make it again, but I wanted to avoid it, if possible.

Thanks.

8

General Board / Internal transitions in statemachine diagrams

« on: October 29, 2012, 06:19:14 pm »

How can I set an internal transition in a state?

9

General Board / Find and replace a word in a model

« on: September 06, 2010, 05:46:09 pm »

Oh, well. I have a model with lots of requirements. Some of them were called "mandatory" requirements, but we have made up our minds, and now they are called "basic" requirements.

So I have to update the model in order to reflect this change. Changes to be made:

1) Names of some packages (change Mandatory / mandatory to Basic / basic)
2) Names of some diagrams (idem)
3) Content of some notes (idem)
4) Content of some elements' descriptions (idem)
5) Names of some requirements (change of MA01, MA02... to BA01, BA02, ...)

I suppose this should be easy for someone knowing SQL or the like, but I don't know a single word of it. Can anybody offer me a hint, other than the brute-force approach of using an hex editor on the .eap file ?

10

Uml Process / Multiplicity in ports

« on: February 11, 2013, 10:35:40 pm »

I am a little confounded by the relationship between multiplicity in ports and multiplicity in connectors, inside a composite structure diagram.

Let's suppose we have a device like this:



And its composite structure diagram:



So far, so good. Now, suppose I want to take advantage form the fact that ports pact1 and pact2 are of the same type; ditto for pcpu1, pcpu2. I think I can model itlike this:



And the composite structure diagram:



Now, how can I be sure that the latter  diagram means the same as the first composite structure diagram, and not this one?:



Or is this not equivalent because of some restrictions derived from the multiplicity (1) of the end associations?

In other words, I think I know how multiplicity works with associations between classes (and parts) without ports, but I am not sure how it works with ports. Can someone point to the correct rules?

I've seen complicated examples in Internal Block Diagrams (A Practical Guide to SysML, by Friedenthal and others) that I don't fully understand, because no clear (to me) rule is formulated.

The above examples are the simplest ones I cannot fully elucidate. I have more complicated examples for the adventurer who dares to offer me a good response.  

11

Uml Process / Merge node missing?

« on: July 27, 2012, 03:27:06 am »

I have always thought that, when you have a branch (decision node) in an activity diagram, and the flows reunite in a future action, you must use a merge node before that future action. Otherwise, that future action will wait (by default) for a token on each of the arriving edges, an event that will never happen. In other words, two edges arriving at one action are equivalent to an implicit join, not an implicit merge.

But in the "Activity Pattern" named "Basic Activity" provided in EA 9.3, you can find this very error! Who is mistaken, me or "EA 9.3"?

You can see the same (presumed) error in http://commons.wikimedia.org/wiki/File:Activity_Diagram_1.jpg (in "Attivita 5") or in http://upload.wikimedia.org/wikipedia/commons/e/e7/Activity_conducting.svg (in "Explain problem").

12

Automation Interface, Add-Ins and Tools / UML profile for safety analysis

« on: September 15, 2010, 10:33:00 pm »

Has anyone developed a profile for performing safety analysis, similar to that developed by Bruce Douglass for Rhapsody? (see http://www.ibm.com/common/ssi/fcgi-bin/ssialias?infotype=SA&subtype=WH&appname=SWGE_RA_SY_USEN&htmlfid=RAW14124USEN&attachment=RAW14124USEN.PDF)

I'm specially interested in FTA.

And is (s)he willing to share it?