RE: [scxml] Model Validation Framework

[Fasihullah Askiri <fa...@baypackets.com>]

<quote>

 * What markup is that?
 * Who consumes it? i.e. How does this integrate with what we have?
 * What benefit does it bring, as compared to writing a SCXMLUnit
(JUnit with helpers for Commons SCXML) test case to a similar effect?

Lets articulate that.
</quote>

I am suggesting a markup which can be used to validate a flow. The XML that I had pasted in my orig mail was what I thought it could look like. 

Now the question that I was thinking of myself, what benefit does it provide over the SCXMLUnit.
Whereas SCXML suite is very useful for testing SCXML internals, but to validate a document, it requires too much of internals [IMHO]. The XML based testing makes the testing available to the end user. i.e. an end-user like me, can write the transition that the FSM goes thru, the behaviour expected etc. As you can see in the example that I had pasted in the earlier mail, the user just says that this is what my state was in the beginning, this is the transition it goes thru. The SCXMLUnit work is done behind the screens. 

Thanx for the encouragement to think aloud. If you feel that this does not have a real benefit, we can drop this off. By the way, it would be great if the users can comment if they feel a need for such a thing in the first place or is it OK with the SCXMLUnit type of a validation framework?

+Fasih

Re: [scxml] Model Validation Framework

[Rahul Akolkar <ra...@gmail.com>]

On 4/27/06, Fasihullah Askiri <fa...@baypackets.com> wrote:
> <quote>
>
>  * What markup is that?
>  * Who consumes it? i.e. How does this integrate with what we have?
>  * What benefit does it bring, as compared to writing a SCXMLUnit
> (JUnit with helpers for Commons SCXML) test case to a similar effect?
>
> Lets articulate that.
> </quote>
>
> I am suggesting a markup which can be used to validate a flow. The XML that I had pasted in my orig mail was what I thought it could look like.
>
> Now the question that I was thinking of myself, what benefit does it provide over the SCXMLUnit.
> Whereas SCXML suite is very useful for testing SCXML internals, but to validate a document, it requires too much of internals [IMHO]. The XML based testing makes the testing available to the end user. i.e. an end-user like me, can write the transition that the FSM goes thru, the behaviour expected etc. As you can see in the example that I had pasted in the earlier mail, the user just says that this is what my state was in the beginning, this is the transition it goes thru. The SCXMLUnit work is done behind the screens.
>
> Thanx for the encouragement to think aloud. If you feel that this does not have a real benefit, we can drop this off.
<snip/>

As I mentioned in the last thread on this topic, you have a reasonable
suggestion. If you pursue it further, you should do it because you
find it useful (whereby my opinion isn't as relevant).

If you're asking whether this can be part of Commons SCXML, I would
first suggest not calling this effort the Model Validation Framework
(Commons usually does not entertain "frameworky" ideas, and in this
case IMO its a misnomer anyway). While I'm not thrilled that it
invents a new markup, if there is enough user demand, a really
lightweight piece of code that fits the test package in spirit, and
good documentation, you could propose it by attaching a patch to a
bugzilla request for enhancement.

-Rahul


> By the way, it would be great if the users can comment if they feel a need for such a thing in the first place or is it OK with the SCXMLUnit type of a validation framework?
>
> +Fasih
>

---------------------------------------------------------------------
To unsubscribe, e-mail: commons-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-user-help@jakarta.apache.org