The diversion through phase changes is awkward, here? I would use that metaphor to point out that many of the properties and requirements to maintain one phase are not relevant when in another phase. More, some properties become harder with different phases. Sure, holding water is generally easier than holding steam. Neither is easy at large scale, of course.
I think you could stick with construction metaphors for a lot of learning. Scaffolding key stones is particularly instructive. You will build a large structure that is intended to be torn down in order to build another structure. And there is basically no avoiding it.
I'm not clear how to move those ideas to formal methods. Typically you do that by keeping different layers that you formalize to different degrees? Along with processes to confirm compliance at the different layers.
Software seems to get tough as we often wind up trying to expand solution and construction models to be fully inclusive of each other. We don't necessarily do this for any other construction. Do we?
I think you could stick with construction metaphors for a lot of learning. Scaffolding key stones is particularly instructive. You will build a large structure that is intended to be torn down in order to build another structure. And there is basically no avoiding it.
I'm not clear how to move those ideas to formal methods. Typically you do that by keeping different layers that you formalize to different degrees? Along with processes to confirm compliance at the different layers.
Software seems to get tough as we often wind up trying to expand solution and construction models to be fully inclusive of each other. We don't necessarily do this for any other construction. Do we?