6 ms·

What do you think about OP's comments that the algorithm questions were unrelated to the front-end positions?

mkozlows 10y ago
Of course they're unrelated. The interviewer asking this doesn't expect you to write these algorithms all day, they want to know that you have the basic mental dexterity to figure things like this out. They are, ideally, abstracted problems of the complexity that you'd be expected to be able to solve on your own. There are some positions where you don't have to care if people can really problem-solve; as long as they can tweak CSS until it works right-ish, or know the particular arcane implementation details of installing some WordPress module, they can do the job. But there are a lot more positions where you want someone who can solve the problems that need solving, pick up new technologies as needed, and just generally be a flexible and productive contributor in a way that doesn't relate to the fiddly details of one particular technology.
- falcolas 10y ago
  > they want to know that you have the basic mental dexterity to figure things like this out If you have the "mental dexterity" to coming up with your own solution which is identical to Dijkstra's pathfinding algorithms in terms of complexity within an hour interview, you are a god among mortals. And that's what they expect - a solution to a problem that is equal to the best minds of our field. I say this from experience: I was expected to do exactly that, and criticized when I came up with a O(n^3) instead of O(n^2).
  - SamReidHughes 10y ago
    Solving a maze isn't identical to Dijkstra's algorithm. And Dijkstra's algorithm was specifically chosen to be an easy problem with an understandable answer, to show off a new computer. It was never some grand unsolved problem. Yes, it's reasonable to expect somebody with a modern data structures education who mysteriously hadn't heard of it to be able to invent it. Or to reinvent it, having forgotten it, because it's your basic breadth first search with care taken not to get "lost in the woods."
- cwilkes 10y ago
  They aren't asking those questions to see if you have mental dexterity. They are asking them as they asked hundreds of people that same question before and kid themselves to thinking it is an objective measure as, well, everyone working at the company passed it and everyone here is a 10xer and so if you don't pass it you are a loser.
  - _lce0 10y ago
    I find the 10x funny.. If everyone is an 10x, then everyone is an 1x .. Isn't?
- crazy5sheep 10y ago
  Passing such interview questions could only proof: 1. you are prepared 2. you are eager to get a job 3. you are not dumb
kllrnohj 10y ago
I think that's an unfounded and problematic stance. Front-ends almost always involve working with data. It is increasingly common for front-ends to have filters, searching, sorting, etc... of that data as well. Doing that efficiently is key to ensuring a fast, responsive UI. Wrangling with divs and CSS may take a disproportionately large amount of your time but that doesn't mean it's the most important skill to look for, nor does it mean the CS questions are somehow irrelevant to the work.
jameshart 10y ago
Complaining that breadth-first-search is irrelevant to frontend dev seems concerning to me. The DOM is a tree; so are JSON structures; so are website directory structures; so are half the data structures you will encounter in most businesses (departments, categories, etc.). You WILL need to traverse a tree in a frontend position at some point. I absolutely want to know whether a frontend developer has a variety of tools in their toolbox for handling tree structured data - and ideally the ability to refactor between recursion (using the call stack), explicit stack descent (which will be depth first), and queue based descent (which will be breadth first).
- cc81 10y ago
  That would be incredibly rare to implement a breadth first algorithm in front end work. I mean you could pick any question about the most obscure browser quirks would probably be more relevant.

Of course they're unrelated. The interviewer asking this doesn't expect you to write these algorithms all day, they want to know that you have the basic mental dexterity to figure things like this out. They are, ideally, abstracted problems of the complexity that you'd be expected to be able to solve on your own. There are some positions where you don't have to care if people can really problem-solve; as long as they can tweak CSS until it works right-ish, or know the particular arcane implementation details of installing some WordPress module, they can do the job. But there are a lot more positions where you want someone who can solve the problems that need solving, pick up new technologies as needed, and just generally be a flexible and productive contributor in a way that doesn't relate to the fiddly details of one particular technology.

falcolas 10y ago
> they want to know that you have the basic mental dexterity to figure things like this out If you have the "mental dexterity" to coming up with your own solution which is identical to Dijkstra's pathfinding algorithms in terms of complexity within an hour interview, you are a god among mortals. And that's what they expect - a solution to a problem that is equal to the best minds of our field. I say this from experience: I was expected to do exactly that, and criticized when I came up with a O(n^3) instead of O(n^2).
- SamReidHughes 10y ago
  Solving a maze isn't identical to Dijkstra's algorithm. And Dijkstra's algorithm was specifically chosen to be an easy problem with an understandable answer, to show off a new computer. It was never some grand unsolved problem. Yes, it's reasonable to expect somebody with a modern data structures education who mysteriously hadn't heard of it to be able to invent it. Or to reinvent it, having forgotten it, because it's your basic breadth first search with care taken not to get "lost in the woods."
cwilkes 10y ago
They aren't asking those questions to see if you have mental dexterity. They are asking them as they asked hundreds of people that same question before and kid themselves to thinking it is an objective measure as, well, everyone working at the company passed it and everyone here is a 10xer and so if you don't pass it you are a loser.
- _lce0 10y ago
  I find the 10x funny.. If everyone is an 10x, then everyone is an 1x .. Isn't?
crazy5sheep 10y ago
Passing such interview questions could only proof: 1. you are prepared 2. you are eager to get a job 3. you are not dumb

kllrnohj 10y ago

I think that's an unfounded and problematic stance. Front-ends almost always involve working with data. It is increasingly common for front-ends to have filters, searching, sorting, etc... of that data as well. Doing that efficiently is key to ensuring a fast, responsive UI. Wrangling with divs and CSS may take a disproportionately large amount of your time but that doesn't mean it's the most important skill to look for, nor does it mean the CS questions are somehow irrelevant to the work.

jameshart 10y ago

Complaining that breadth-first-search is irrelevant to frontend dev seems concerning to me. The DOM is a tree; so are JSON structures; so are website directory structures; so are half the data structures you will encounter in most businesses (departments, categories, etc.). You WILL need to traverse a tree in a frontend position at some point. I absolutely want to know whether a frontend developer has a variety of tools in their toolbox for handling tree structured data - and ideally the ability to refactor between recursion (using the call stack), explicit stack descent (which will be depth first), and queue based descent (which will be breadth first).

cc81 10y ago
That would be incredibly rare to implement a breadth first algorithm in front end work. I mean you could pick any question about the most obscure browser quirks would probably be more relevant.