Modeling Feedback

From Event-B
Jump to navigationJump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Hereafter is a gathering of some modeling feedbacks, which may be useful to the whole community. They come from:

What is the goal of a model

Real goal is to obtain a proof, not a model.

The model is only our mean of choice to obtain a proof.

Highly iterative task between proof and model.

What is a safe system?

TODO: Safety preservation

What is a good model?

TODO: model validation via animation

How to take into account degraded cases

TODO: model must be totally closed

WD PO may bear essential model semantic

Safety at some point, unexpectedly relied upon a WD PO:

{next}({head}_{train}) is well defined.

Abstraction is heavily needed

Concepts provided in informal specification are insufficient to express safety property.