[TeMPOraL]

The more I have to deal with corporate project management patterns, the more I think the whole approach behind specifications is wrong. Trying to focus on what the "domain objects" are and what is their flow is pointless, especially if you're dealing with real-world objects (like the fridge from TFA). You can model something this way, but it will be usually the imaginary perfect path, a narrow vision of how things should be done, that will break on first contact with the real world - a world that is messy and requires humans to improvise on the spot.

I'm starting to believe that one should focus on the workflow instead. After all, the reason for writing software as a tool is to simplify or automate some existing process. So focus on what people are already doing and what they're trying to achieve and find a way to remove repetitions, unnecessary delays, etc. Just look on what people are doing and find a way to do some parts of the process faster and better.

[josh64]

That's where user stories can help:

"As a user, I want to store my food and drinks in the refrigerator so they don't spoil"

"As a user, I want to store drinks in the refrigerator so I can drink them cold"

"As a user, I want to store my medicine in the refrigerator so it can stay below 25 C"

[epochwolf]

> Just look on what people are doing and find a way to do some parts of the process faster and better.

This requires buy-in from different parts of the company that are going to have their own opinions about how to do things. They can be surprising resistant to change.