4 ms·

Checking how many language subtleties you can point out is no doubt a good interview if you're looking for people experienced in a specific language. It strike

by kubb 4y ago

Checking how many language subtleties you can point out is no doubt a good interview if you're looking for people experienced in a specific language. It strikes me as so different from what the FAANGs are doing: they ask you to write trivial algorithms on the whiteboard, which apparently filters 99% of engineers. Theoretically you can have a good candidate that doesn't know e.g. that appending to a slice passed in a function argument is a recipe for disaster. You'd show them the door over something that they could learn in a day.

Barrin92 4y ago
theoretically sure but chances are any good engineer knows that stuff and doesn't need to learn it, and FAANG can filter heavily because they get a lot of applicants. Basically if you don't know something you could have learned in a day in preparation for an interview at a top paying company that in itself is a sufficient signal.
- kergonath 4y ago
  There are many more “things you can learn in a day” than there are available days to do it, though. Nobody will ever know all the possible tiny simple things. I would think in general that it’s better to select people who adapt and learn quickly rather than people who seem to know everything but you’re right: we are not fishing in the same pond as the FAANGs.
kevincox 4y ago
You are making a lot of assumptions about the interview. Code review is an important part of the development process and seeing candidates who can suggest improvements to code is valuable. You seem to be assuming that the intended answers are language subtitles but I don't see evidence of that in the post.
nemo1618 4y ago
> Theoretically you can have a good candidate that doesn't know e.g. that appending to a slice passed in a function argument is a recipe for disaster. Huh? This is a common pattern. (hash.Hash).Sum, strconv.AppendInt, etc. I think you're referring to the fact that append can mutate existing memory (that is, it doesn't always allocate new memory), but in practice, this is rarely an issue, because, by convention, functions that append to a slice always return that slice.

by kubb 4y ago

Barrin92 4y ago

theoretically sure but chances are any good engineer knows that stuff and doesn't need to learn it, and FAANG can filter heavily because they get a lot of applicants. Basically if you don't know something you could have learned in a day in preparation for an interview at a top paying company that in itself is a sufficient signal.

kergonath 4y ago
There are many more “things you can learn in a day” than there are available days to do it, though. Nobody will ever know all the possible tiny simple things. I would think in general that it’s better to select people who adapt and learn quickly rather than people who seem to know everything but you’re right: we are not fishing in the same pond as the FAANGs.

kevincox 4y ago

You are making a lot of assumptions about the interview. Code review is an important part of the development process and seeing candidates who can suggest improvements to code is valuable. You seem to be assuming that the intended answers are language subtitles but I don't see evidence of that in the post.

nemo1618 4y ago

> Theoretically you can have a good candidate that doesn't know e.g. that appending to a slice passed in a function argument is a recipe for disaster. Huh? This is a common pattern. (hash.Hash).Sum, strconv.AppendInt, etc. I think you're referring to the fact that append can mutate existing memory (that is, it doesn't always allocate new memory), but in practice, this is rarely an issue, because, by convention, functions that append to a slice always return that slice.