21 ms·
>Firstly, I'm really glad to say that the interviews I did were not the stereotypical tech interviews. As an industry, it's great to see we're getting better, a
by DeathArrow 5y ago
>Firstly, I'm really glad to say that the interviews I did were not the stereotypical tech interviews. As an industry, it's great to see we're getting better, and moving away from whiteboard coding, dealing with super abstract problems like inverting a binary tree, or doing things that we'd generally just use a Search Engine for in our real jobs.
I would expect a junior developer out of college knows data structures and algorithms. That includes knowing what a binary tree is. And if you know what a binary tree is, it isn't terribly difficult to invert it.
Why shouldn't a "senior" developer know what a binary tree is?
While I don't have much regard for leetcode and hackerrank, I wouldn't want to hire someone who doesn't know about algorithms and data structures and thinks he can solve anything with Google.
Sure, many thinks are solvable using Google, but many require knowledge, experience and practice.
>Unfortunately one company - who outsourced the first part of their interview process - had something similar to this, which I wasn't impressed with. There were some arbitrary questions around several topics, and then in the coding exercises you were asked what the O(n) complexity of time/space were.
The author is even outraged he was asked about time complexity, something even junior programmers shouldn't have a with.
- zeroonetwothree 5y agoI don’t like the “trick” questions where you have to think of some very specific approach or else you fail. But I agree that basic knowledge of things like trees and runtime complexity is reasonable. That kind of stuff does come up in actual work.
- avereveard 5y agoMemorizing stuff that any reference book can tell you is a waste of time and skill. Knowing when to use a binary tree is much more important. Knowing not to write a binary tree and using a library is a delivery saving skill. At least abstract problems off a whiteboard show candidate problem solving approaches. But dry coding koans are absolutely a waste of time for anyone involved and tend to produce teams that care about knowledge tidbits hoarding instead of, idk, serving the business that pays them producing value. And rest assured, there is no value in writing a binary tree for the business. As a matter of fact if one of my dev doesn't use off the shelf components for these kind of things they'd get a stern talking about time efficiency, security and the likes. Why would I go and stress that at a interview?
- DeathArrow 5y ago>Memorizing stuff that any reference book can tell you is a waste of time and skill. If you are ill, would you prefer a surgeon that will search on Google doing surgery, ore or one that knows how to do it? Would you be happy to fly with a pilot that will search on Google how to land? >And rest assured, there is no value in writing a binary tree for the business. As a matter of fact if one of my dev doesn't use off the shelf components for these kind of things they'd get a stern talking about time efficiency, security and the likes. Why would I go and stress that at a interview? Of course, for trivial stuff you won't need binary trees. But you're wrong in assuming all software is trivial and all problems are solvable with library calls. And even for trivial stuff it's important to know and understand the underlying principles and how software and infrastructure works. Otherwise you will take bad decisions like searching lists instead of hash maps. And bad decisions will cost your employer money. Sometimes a lot of money. So yes, knowledge matters. And experience matters. And problem solving skills matter.
- avereveard 5y agoFalse equivalency. And broken metaphor. First of, you don't have the same urgency as a surgeon. Management might want you to think you do, but it's all smoke and mirrors. If you don't have equity and stakes aren't people lives, it's not even close. Second, surgeon do spend time before surgery to freshen up procedures from references, as they see a lot of different things, and while the gist of everything are in memory, the specifics are reviewed with the team before operating. Are you doing equivalencies out of TV show knowledge by any chance?
- sdevonoes 5y ago> Why shouldn't a "senior" developer know what a binary tree is? Because it's not about binary trees. You could as well ask why a "senior" developer shouldn't know what: - a TLB is - a Trie is - UDP is - consistent hashing is - the CAP theorem is about - NFS is - BFS is - and a large etc. As you can see, a question like "what a binary tree is" could easily be as well "what BFS is about". So, it's not about "senior" developers not knowing their stuff, but that there is so much to know about, that certain developers may know better about certain topics than others. For instance, I don't know by heart the algorithm for inverting a tree; sure, I can give it a try and may come up with a solution... but I'm sure a fresh grad student who has studied in the last semester how to inverted a binary tree, will come up with a solid algorithm faster than I can come up with my weak algorithm. > I wouldn't want to hire someone who doesn't know about algorithms and data structures and thinks he can solve anything with Google. Same can be said about not hiring someone who doesn't know the basics of Operating Systems (e.g., TLBs), or the basics of networking (UDP vs TCP), or the basics of distributing systems (e.g., CAP theorem, distributed transactions)... I'm not saying "There is too much to learn about! Please don't ask me anything!", I'm saying that at a given point in time, any experienced developer can provide a poor algorithm (and so fail the interview) to invert a binary tree.
- charcircuit 5y agoEven an entry level developer should know those things at least at a surface level since they are taught at university. Regardless in interviews you are not required to know what a binary tree is. Your interviewer will be able to define it for you.
- ipaddr 5y agoWhy assume developers all come from cs programs and filter for only those candidates? When I went to college I got to work on a mainframe. I'm not going to assume everyone did and start asking JCL questions.
- charcircuit 5y agoNothing is stopping you from learning these topics on your own. In fact the vast majority of these topics I learned about on my own before encountering them in university. The problem with asking mainframe questions is that they aren't as relevant in today's landscape. Meanwhile TLB is relevant for performance, UDP is a very common network protocol, consistent hashing is relevant for designing distributed systems, etc. Especially if you are just self teaching yourself you aren't going run into a chance to work on a mainframe unless maybe you use a mainframe emulator, but why?
- karaterobot 5y ago> Why shouldn't a "senior" developer know what a binary tree is? Well, wait a minute. He didn't say it's unreasonable to ask what a binary tree is. He said he was glad he wasn't asked to invert a binary tree on a whiteboard. There's a big difference between those things. I think your reading is uncharitable.
- scarface74 5y agoOn a day to day basis, most jobs are simply CRUD using a framework. The day to day need to know algorithms is nil.
- laichzeit0 5y agoThese specific questions don’t help much in my opinion. I prefer asking questions that I know someone can’t game through studying some leetcode crap. - What was the most difficult problem you’ve had to solve as a developer? They talk about it. You get some data, and you start asking more detailed questions about it for them to elaborate on. Soon you discover if they’re bullshitting, or at least what’s their “technical ceiling” based on what they considered the most difficult thing they’ve ever done. - Have you ever had a performance issue that you’ve had to debug. How did you go about finding the root cause? Etc. Basically nothing with a textbook answer. Just get them to talk and not be autistic about it. Surely if you know your craft you’d be able to suss out if this person can solve technical problems and get an idea about how they go about doing it. I really don’t care if you can invert a tree or have memorised The Art of Computer Programming… Will: Wood drastically -- Wood 'drastically underestimates the impact of social distinctions predicated upon wealth, especially inherited wealth.' You got that from Vickers, 'Work in Essex County,' page 98, right? Yeah, I read that too. Were you gonna plagiarize the whole thing for us? Do you have any thoughts of your own on this matter? Or do you...is that your thing? You come into a bar. You read some obscure passage and then pretend...you pawn it off as your own idea just to impress some girls and embarrass my friend? See the sad thing about a guy like you is in 50 years you're gonna start doin' some thinkin' on your own and you're gonna come up with the fact that there are two certainties in life. One: don't do that. And two: You dropped a hundred and fifty grand on a f----n' education you coulda' got for a dollar fifty in late charges at the public library.
- lostdog 5y ago> What was the most difficult problem you’ve had to solve as a developer? They talk about it. ...and then you get accused of trying to steal secrets from their previous company. In seriousness, we're no longer allowed to ask questions like this in our current interview process, since it could expose us to competitors' internal secrets. It sucks, because it was one of the best ways of determining what a person was capable of. But there's no winning in designing an interview process--every one has flaws. Now I come up with a hypothetical problem in a related industry, but it's hard to tell whether the person is just failing to think through the problem, or was too unfamiliar with the domain to have useful ideas.