Test-case reduction slippage (moving from one bug to another due to an insufficiently precise test): Impossible to avoid when you don't want it, impossible to trigger when you want to demonstrate it.
More generally, my impression is that there’s startlingly little explicitly known about bugs, empirically or theoretically? Just folk-understanding. Almost nothing in writing when I was a grad student, and I haven’t heard of much since? Seems obviously extremely important…
-
-
I think we're in a slightly annoying state where we know quite a lot about finding bugs, but there's a lack of unifying theory. Part of the problem is the issue I described in https://www.drmaciver.com/2018/04/some-of-my-problems-with-correctness-research/ … - it's very hard to properly study this without industry/research collaboration.
-
That being said there's definitely something more than that going on. There's quite a lot of research in bug finding in compilers, because software testing researchers understand compilers for the most part, but I still don't think we understand the shape of the problem.
End of conversation
New conversation -
-
-
There was Kurt van Lehn: https://onlinelibrary.wiley.com/doi/pdf/10.1207/s15516709cog0404_3 … and a lot of Schankian stuff is sort of in the ballpark. Certainly important and there could be more.
-
This looks super interesting, thanks, and I'm going to give it a more detailed read, but I think it's focusing on a bit of a different of the problem - the relevant question here is more about the shape of inputs that trigger a given bug than where bugs come from per se.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.