Man, I'm currently in a project which started out with 2 major goals. Pretty early on, we got told that one of the goals is practically impossible, so we decided to ignore that. And we realized the other goal needs to be simplified significantly to be achievable in the slightest, although we still weren't sure, if it violates the laws of physics.

Now we're a year into development, we've only figured out that it might be physically possible in certain situations. And yesterday, we talked to a guy with domain knowledge, who told us like ten different bigger challenges we'd still have to solve.