3 ms·

Even better, if you can prove that the variables are independent, it suffices to test any permutation in place of all permutations... (I think this is why the

Even better, if you can prove that the variables are independent, it suffices to test any permutation in place of all permutations... (I think this is why the medical devices had almost all errors show up at shallow interaction levels: they were probably designed with ruthless simplicity in mind) > While searching I came across "combinatorial testing", seems relevant but still fail to see how it all ties together? the tl;dr for the combinatorial testing deck is that many bugs will surface with interactions between small numbers of factors: so covering all-pairs, all-triples, all-quads, etc. goes a long way towards surfacing almost all bugs at a fraction of the cost of covering all-interactions.

_5hxt 3y ago
Thanks for taking the time to reply, very valuable. Little tangent but in regards to the coupling/cohesion part; makes me wonder why some developers seem to have no problem with accepting strict boundaries when dealing with third party (web) APIs they have no control over, they use parts of the interface they desire, they cherrypick/reduce the response content and create a separate model appropriate for their usecase/context and they tailor the behavior to that. This appears to go out the window to some extent when they do have control over the existing model and behavior. Far more appealing to use the model directly and modify or extend the behavior than to "needlessly" introduce some duplication in order to establish clear boundaries.