re: #312 dog philosopher

fashions in software engineer interviews change

in the past few years most interviews seem to involve at a certain point being asked to stand up at a white board and code a routine on the spot

they give you a description of the problem and you have to show that you are going through all reasonable solution stages just like at work:

1. ask questions until you are clear exactly what is required
2. do high level planning until you are ready to do pseudo-code
3. pseudo-code
4. real code
5. discuss the ins and outs of the big-O runtime of your solution and how you might improve it
6. discuss error handling
7. talk about how to unit test and system test it

generally if the problem seems too simple it’s because what they really want is a recursive solution - although in practice pgmrs hardly ever write recursive routines

i am hoping this fashion runs its course and we get back to normal interviews

I would not want to work for anyone who thinks it is possible to solve a complex problem in 20 minutes. For some reason assholes like this end up in management (because they are good at bullshit?)