It also pisses off the 0.5% that know how to program
Best programming test I've seen was one that was done online, and automatically ran your test through unit tests.
Doing a test is fine, but do it, don't waste my time and move on accordingly to the result. Also, make me do it first, so the technical ability is off the table at an interview, don't make me 1) do it "on paper", (or like a quiz show) and in your location and 2) rate on objective measures (like passing/failing some test) and be transparent about scores (like they would be in a code review)
It doesn't piss off anyone I know that knows how to code because it's simple and it only takes a few seconds to complete. Everyone I know that can code is done the exercise before its even a second thought.
I agree with being transparent about the discussion around the result of any programming done during an interview but having the person do it before hand defeats the purpose of using it to catch out people who can't actually code because they will just Google the result.
I just had a make this app in 2 hours. Usually I like to think about a design for at least half a day before coding. So I did this, evolved my design on the way got it 80% finished, then was criticised for not including tests. Seriously, who even has to produce something other than a two line bug fix with a 2 hour deadline? As I get more experienced, I realise thoroughness is way more important than raw speed. Over the long term it saves a lot of time.
I had to implement a spreadsheet (with formula) in Perl with a 2 hour deadline. Completed the task (apparently no-one else had) but didn't get the job because the CTO had a hard-on for Scala and I said it was stupid to rewrite existing money-making systems into a brand new language for no good reason.
You're not the target of Fizzbuzz though clearly. But with that in mind you should be able to finish it so quickly that it would take less time than a standard question about your last job.
FizzBuzz irritates me because they haven't gone to the trouble of thinking something else.
Because they don't have to. It's simple and easy to use filter that will remove a lot of noise. It's not perfect, some people will memorize a solution to it but simple changes to the question solve that. But it does really work.
I think having to sit on the other side of the table for a while gave me a lot more respect for the pain that is being an interviewer.
Another thing I was given was "make me something interesting in 30min" or "try to solve this problem in 2 hours"
Which are totally different and often useless beasts though done right really helpful but honestly I was never willing to invest 2 hours into a single interview nor make someone sit in a room for 2 hours while I went back to work.
Comments
It also pisses off the 0.5% that know how to program
Best programming test I've seen was one that was done online, and automatically ran your test through unit tests.
Doing a test is fine, but do it, don't waste my time and move on accordingly to the result. Also, make me do it first, so the technical ability is off the table at an interview, don't make me 1) do it "on paper", (or like a quiz show) and in your location and 2) rate on objective measures (like passing/failing some test) and be transparent about scores (like they would be in a code review)
It doesn't piss off anyone I know that knows how to code because it's simple and it only takes a few seconds to complete. Everyone I know that can code is done the exercise before its even a second thought.
I agree with being transparent about the discussion around the result of any programming done during an interview but having the person do it before hand defeats the purpose of using it to catch out people who can't actually code because they will just Google the result.
"defeats the purpose of using it to catch out people who can't actually code because they will just Google the result."
Then don't use FizzBuzz. It's that simple
FizzBuzz irritates me because they haven't gone to the trouble of thinking something else.
Another thing I was given was "make me something interesting in 30min" or "try to solve this problem in 2 hours"
I just had a make this app in 2 hours. Usually I like to think about a design for at least half a day before coding. So I did this, evolved my design on the way got it 80% finished, then was criticised for not including tests. Seriously, who even has to produce something other than a two line bug fix with a 2 hour deadline? As I get more experienced, I realise thoroughness is way more important than raw speed. Over the long term it saves a lot of time.
I had to implement a spreadsheet (with formula) in Perl with a 2 hour deadline. Completed the task (apparently no-one else had) but didn't get the job because the CTO had a hard-on for Scala and I said it was stupid to rewrite existing money-making systems into a brand new language for no good reason.
Yeah, I'd rather deliver it 90% without tests than 50% with tests
You're not the target of Fizzbuzz though clearly. But with that in mind you should be able to finish it so quickly that it would take less time than a standard question about your last job.
Because they don't have to. It's simple and easy to use filter that will remove a lot of noise. It's not perfect, some people will memorize a solution to it but simple changes to the question solve that. But it does really work.
I think having to sit on the other side of the table for a while gave me a lot more respect for the pain that is being an interviewer.
Which are totally different and often useless beasts though done right really helpful but honestly I was never willing to invest 2 hours into a single interview nor make someone sit in a room for 2 hours while I went back to work.
Well, I've sat on the other side as well, and I've managed to gauge the technical knowledge of people with simple, but directed questions.
But I understand that FizzBuzz gives you a high degree of confidence that the candidate can work in the given language.
"I was never willing to invest 2 hours into a single interview"
Well, lucky for you, at my last unemployment stint this was a given (but not with one single interviewer). However, the 2h test was "offline"