Re: [papyrus4edu] TR on Papyrus in education

[Bran Selic <selic@xxxxxxx>]

On Mon, Oct 19, 2015 at 11:13 AM, Marc Pantel <Marc.Pantel@xxxxxxxxxxx> wrote:

Dear all,

This subject is very interesting and in line with my 25 years of teaching experience both in programming and modeling.

I have one feeling related to the availability of tools that allow engineers to get confidence in the correctness of their proposals.

It is very easy in programming to test early versions of programs and get confidence that we are on the right direction.

​[bvs] Indeed. Which is why I have always advocated for executable modeling -- even for high-level designs. For those who have doubts that high-level design models can be executed, I can confirm that this is perfectly feasible in most cases; I've worked with tools that do precisely that. In such situations, the high-level behavior is typically specified in terms of simple interactions​

 

​ (e.g., sending and receiving of high-level messages)​. In some cases, the tool may reach a point where it will require input from the modeler in order to proceed. But this too is relatively straightforward to implement. 

[bvs] This not only increases confidence as you point out and which is crucial, but it also helps designers gain an intuition for the design and the system that is difficult to achieve purely through "mental" (i.e., inside the head) execution. Last but not least, it would overcome much of the resistance to modeling and modeling tools. 

Cheers... Bran