Conversation

Think there might be some kind of traveling salesman/P vs NP thing going on here, though I'm a bit out of my depth here. Like, IIUC, NP problems are often easy to say, but hard to compute. Programming is often like that.
1
1
could be! although I think it's also a bit just a perceptual bias thing, as the easy/easy cases stick in my brain way less than the time where I need to very slightly change a workflow and it takes two weeks ...
1
Are the inputs/outputs and side effects of the "workflow" clearly specified? If not then I can very much empathise with the challenge of trying to wrap one's head around how it all works - it's nontrivial. Even if they are it can be a big challenge as stuff gets more complicated.
2