3 ms·
There is a balance here. Good domain modeling can enable making illegal states unrepresentable. Sometimes, there needs to be more flexibility because of a lac
by ParetoOptimal 2y ago
There is a balance here.
Good domain modeling can enable making illegal states unrepresentable.
Sometimes, there needs to be more flexibility because of a lack of understanding.
However, other times you can put in a little effort and get the understanding needed for a good data model that simplifies future velocity and correctness a ton.
- temporarely 2y agoI think you miss the fine point here. Ontological truth =/= Good domain modeling. OP is not arguing against domain modeling, rather a likely common trap of creating an analog of 'the world' (however delimited) as foundation for the model. I know exactly what they mean because I am guilty as charged. I used to do this all the time in mid career.
- tilt_error 2y agoYes, you shouldn't "boil the ocean".
- kybernetikos 2y agoYou should usually also not "model the ocean"