Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: coding challenges as part of interviews: yes or no?

news.ycombinator.com

Post reply on HN
Print view View original

51–58 of 58 posts

Re: Ask HN: coding challenges as part of interviews: yes or no?

#51

Post by etal »

hugh wrote: post #50
raganwald wrote: post #24

Earlier quoted context omitted.

I never use Ruby's case or switch or whatever-it's-called. I could make up some mumbo-jumbo about case being a code smell and that it should actually be refactored to a method dispatch over polymorphic objects, but the truth is that I have no idea why I don't write case statements, I just don't.

Add me to the "never uses switch statements" list. I've never seen the point in learning the syntax in any of the languages I know, since a bunch of else ifs (or elseifs, or elsifs, or elifs, depending on the language) does the same thing. If I waste a few keystrokes, but save a few brain cells, I think it's a worthwhile bargain. Of course if I had to write that kind of construction often I'd probably learn the synta…

It's really just a performance optimization for C, in situations where the switching value can be hashed. In languages that make method dispatch easy, and the overhead of a function call is not something you're terribly concerned about, it makes sense to just roll your own as needed -- that's why Python never bothered with a switch statement.

Re: Ask HN: coding challenges as part of interviews: yes or no?

#52

Post by andrewcooke »

icky wrote: post #44
sfpeter wrote: post #39

I like to believe that I can get a pretty good understanding of someone's technical abilities by asking thorough questions about his previous experience, and maybe ask them to explain a couple of basic patterns or programming concepts to me. I remember interviewing a senior java developer who claimed 5+ years of experience and had no idea what a Singleton was. What made it worse was that even after explaining it to h…

> Would you rule out a candidate who doesn't know 0xFE ? If they didn't know it after a moment's thought, I'd cut them some slack; If they couldn't calculate it, with pencil and paper, in 10 minutes (and that's giving them plenty of extra time for interview-nervousness), then I'd be seriously concerned.

i read through some answers here and i nodded in agreement with raganwald saying that this only worked as a filter on failure.

but then i read this and i thought "would i really want to give a job to someone who needs a pen and paper to subtract 1 from 255?"

i'm not sure what the lesson is here - perhaps that programming tests are dangerous because it is very easy to over-interpret the results, even when you start with the best of intentions?

[edited for grammar]

Re: Ask HN: coding challenges as part of interviews: yes or no?

#53

Post by thwarted »

jzell wrote: post #14

If you have the sense from whiteboard discussions that the candidate understands computer science (like gets recursion) but you want to make sure they can code, why not choose a subset of a problem that you (your company) needs to solve?

I do this, and provide a subset of REAL input data and perhaps a sample of the expected output. This ends up showing capabilities for more than just minor singleton operations, and shows that someone can use the atomic units of programming (loops, recursion, functions, etc) to actually produce a result. The chosen programming task is one someone in the office recently had to tackle themselves, so there is a working benchmark to compare against. I give it to them ahead of time, before an in-person interview, and say tell them I'm not necessarily expecting a working program (however, that would be nice), but come to the interview having considered the task and be prepared to talk about it, or talk about their specific solution/implementation. I find this also helps communicate the kinds of tasks the company deals with, and maybe a little of the company culture.

Re: Ask HN: coding challenges as part of interviews: yes or no?

#54

Post by comatose_kid »

sfpeter wrote: post #39

I like to believe that I can get a pretty good understanding of someone's technical abilities by asking thorough questions about his previous experience, and maybe ask them to explain a couple of basic patterns or programming concepts to me. I remember interviewing a senior java developer who claimed 5+ years of experience and had no idea what a Singleton was. What made it worse was that even after explaining it to h…

Depends on the context. If this was a question posed for an embedded s/w eng position, I would not be inclined to hire.

Re: Ask HN: coding challenges as part of interviews: yes or no?

#55

Post by eru »

lsb wrote: post #37
eru wrote: post #33

Earlier quoted context omitted.

Neither Python nor Haskell have it - as far as I know. Perhaps you can create one out of monads though.

Actually, case is built into Haskell: it's the evaluation primitive in Haskell Core; see Don Stewart's blog post http://cgi.cse.unsw.edu.au/~dons/blog/2008/05/16 .

Yes, there it is. Sorry, I did not research my claim.

Actually pattern matching can also be described as something like a case-switch.

Re: Ask HN: coding challenges as part of interviews: yes or no?

#56

Post by TFrancis »

sanswork wrote: post #23
TFrancis wrote: post #22

Earlier quoted context omitted.

I don't think that I would ever hire a insert language programmer. When looking for someone to program, I'm just looking for a programmer; language isn't important.

A lot of people get hired in for emergencies. In such cases waiting for person to learn your offices language of choice isn't ideal.

Though they aren't always possible to avoid, I consider "emergencies" examples of poor business decisions and try to avoid them.

Re: Ask HN: coding challenges as part of interviews: yes or no?

#57

Post by icky »

andrewcooke wrote: post #52
icky wrote: post #44

Earlier quoted context omitted.

> Would you rule out a candidate who doesn't know 0xFE ? If they didn't know it after a moment's thought, I'd cut them some slack; If they couldn't calculate it, with pencil and paper, in 10 minutes (and that's giving them plenty of extra time for interview-nervousness), then I'd be seriously concerned.

i read through some answers here and i nodded in agreement with raganwald saying that this only worked as a filter on failure. but then i read this and i thought "would i really want to give a job to someone who needs a pen and paper to subtract 1 from 255?" i'm not sure what the lesson is here - perhaps that programming tests are dangerous because it is very easy to over-interpret the results, even when you start wi…

No, I would be very liberal in accounting for interview-nervousness.

Now, if it appeared that the candidate had no idea how to solve the problem, that would be a different story.

Re: Ask HN: coding challenges as part of interviews: yes or no?

#58

Post by Xichekolas »

sanswork wrote: post #23
TFrancis wrote: post #22

Earlier quoted context omitted.

I don't think that I would ever hire a insert language programmer. When looking for someone to program, I'm just looking for a programmer; language isn't important.

A lot of people get hired in for emergencies. In such cases waiting for person to learn your offices language of choice isn't ideal.

While I can see some merit in this argument, I'd also point out that I've never met a decent programmer that can't pick up a new language at least passingly in a day or two.
Post reply on HN