[IGSTK-Developers] Re: State machine transition diagram (Visio file)

[Luis Ibanez]

Hi Kevin,

mmm,
If the numeric identifiers are appearing the transitions you may not
be running the most recent version of the code. I fixed that on Sunday.

You may want to CVS update your checkout and try again. The current
version should use the textual description of the States in the
ellipsoids and the Inputs in the transitions arrows.

Note that the look of the graph can be customized a lot. It will be a
matter of looking at the multiple options in graphviz and chosing what
we want to use.


BTW,
The state transitions are also changing a lot (right now) in the classes

      Tracker
      TrackerTool
      SpatialObject




because I'm reworking those clases trying to get
Iteration 3 running this week...



   Luis



-------------------------
Kevin Gary wrote:

> Luis,
> 
> Very nice! We were running the export function here before and got
> the numeric identifiers on transitions. With the labels this looks very
> nice. We are looking at how to export an XMI document.
> 
> This looks to have a different set of states and transitions than the one
> Luis sent me?
> 
> Kevin
> 
> 
> Luis Ibanez wrote:
> 
>>
>> Sohan, Kevin,
>>
>> On a lateral note in this topic,
>>
>> A new test was added yesterday to IGSTK in order to generate
>> the StateMachine diagrams for every class. This is done by
>> invoking the Export method of each class.
>>
>>
>>
>> The test is called:
>>
>>        igstkStateMachineExportTest
>>
>>
>>
>> and should be invoked with a directory as argument, like:
>>
>>        igstkStateMachineExportTest   C:/myDiagrams/
>>
>>
>>
>>
>> The diagrams are exported in "dot" format (files .dot) and can be
>> used for generating JPG, XFIG, PS, and several other image formats.
>>
>> The "dot" tool can be obtained as part of graphviz:
>>
>>     http://www.research.att.com/sw/tools/graphviz/
>>
>>
>>
>> The typical dot command will be
>>
>>     dot -Tgif -o Diagram.gif  igstkTracker.dot
>>
>>
>> (the igstkTracker.dot is the file generated by the new test above).
>>
>>
>>
>> As an example, attached is the diagram exported by the igstkTracker
>> class in its current form.
>>
>>
>>
>>    Luis
>>
>>
>>
>> -------------------------
>>
>> Kevin Gary wrote:
>>
>>> Sohan,
>>>
>>> It had been a while since I had looked at this, and now I am not
>>> sure what modifications need to be made. My notes simply state
>>> to put it in UML syntax, not that there were corrections to make.
>>> So I am not sure why this held you up. I can't really put it in
>>> UML syntax either since there are no targets for the failure events.
>>>
>>> The one area that could be cleaner is the listing of the failure
>>> events. These events have no targets, so it is unclear what
>>> happens when one of these events occurs. It is also unclear
>>> whether these would be synchronous or asynchronous events.
>>> If they are synchronous, it implies a transition could not be
>>> completed until whoever handles the failure event completes,
>>> which would probably not be accurate.
>>>
>>> In any event, if this still troubles you we can discuss next week.
>>> But it looks OK.
>>>
>>> K2
>>>
>>>
>>> ranjan wrote:
>>>
>>>> Hello Kevin.
>>>>
>>>> I would appreciate if you could make necessary modifications/
>>>> corrections in the attached Visio file for the state machine
>>>> transitions, and send me back.
>>>>
>>>> Thanks and regards,
>>>> Sohan.
>>>>  
>>>>
>>>
>>>
>>
>>
>> ------------------------------------------------------------------------
>>
> 
>