I think you are kind of wanting you cake and eating it too.
Earlier, you criticized and said that questions should be asked to see how a candidate thinks about problems.
Fair enough, and it’s why I have these tests.
But when I say that I look at a candidate in a negative light because they blinding initialize a variable without considering the implications… you don’t like that either?
Yes, I hope this would get picked up in testing.
But a candidate who just shotguns code without being mindful about what they are doing… can I trust them to write the test cases?
This kind of thing is a “canary in the coal mine”, or a “brown M&M” for those who are Van Halen fans.
It doesn’t mean the candidate is terrible, but it means they are much more likely to be careless.
And honestly, I can say after administering this test to many dozens of candidates, that there is zero correlation between writing these handfuls of lines of code 100% bug free and number of years of on-the-job experience. But there is a strong correlation between getting this simple little thing right, and doing well down the road after we hired people. The more LOC someone can write without a bug, the more use cases they can consider and account for without having to wait for tests to pick it up (or worse, not even write the test for it), the more effective they are and the better the code is.
You are free to disagree all you want, but I think it is very clear that if someone averages 1 bug for every 5 LOC, or 1 bug for every 10 LOC, that the second is going to have much better outcomes.
Personally, I have spent *days* writing code without even trying to compile/deploy (never mind actually test) and when it was done, I had 5, 6 small bugs to fix. While I certainly think that’s an extreme example and hardly the standard I hold anyone else to, I also don’t consider myself a “code hero”, and in fact my opinion of myself is that I am a B+ developer. I’m met true A and A+ developers, and I am not anywhere near their skillset. So by that reasoning, someone who can’t get through 3, 4 LOC without an obvious bug… are they going to be a net positive for the team, or is everyone else going to be dragged down trying to get them to keep up?
J.Ja
