|
qhat posted:What are people's opinions of non-compsci grads with little experience but with a "diploma" from some random 10 week programming boot camp? I usually just throw the resume directly in the bin. We actually got a really solid non-cs guy who went to a bootcamp, and a friend of mine has had a similar experience. Obviously they're inexperienced, but they tend to be really motivated and interested in the industry. My wife is doing one of those data science bootcamps and I can see why: depending on the program they demand insanely long hours (she's easily putting in 90 hour weeks). So the class is essentially acting like a filter for people who give enough of a poo poo to put up with that for a few months, not to mention teaching them some really practical stuff. So I would consider them if you're looking for a junior dev that can jump on any new technology and will probably enjoy it. Our guy has become our domain expert in the new CI/CD stuff we're doing... he had never used this particular build tool before but then again neither had we. The downside is they're gonna be light on fundamentals like code organization, testing, stuff like that... then again, new CS grads could easily be light on this as well. Also I would check out the program's reputation too, since all this poo poo is so new I don't know which programs to trust and which are garbage. edit: my background is traditional CS and it took me awhile to come around to the idea of bootcamps, but I've gotta admit... it just doesn't take 4-6 years to learn how to touch computers and write the kind of lovely code that most companies actually need. ADINSX fucked around with this message at 14:53 on Aug 28, 2018 |
# ? Aug 28, 2018 14:49 |
|
|
# ? Jan 25, 2025 19:44 |
|
about to go in and interview the dude. I wonder how far I should go in coaching them through the knights tour problem or if that’s verboten
|
# ? Aug 28, 2018 14:59 |
|
ADINSX posted:We actually got a really solid non-cs guy who went to a bootcamp, and a friend of mine has had a similar experience. I guess my sticking point with this role in particular is it demands a couple years experience, mostly because this company really does not need any more junior devs or people with little understanding of basic things like testing and best practices. Even then, I disagree that these bootcamps are all you need to be an effective software engineer long term, there's just so much poo poo that is taught in a cs program that you will never get from a 10 week boot camp. Networking, distributed systems, advanced databases, security, cloud engineering to name a few big ones that I really wish people at this company had even a cursory understanding of.
|
# ? Aug 28, 2018 15:22 |
|
Pollyanna posted:about to go in and interview the dude. I wonder how far I should go in coaching them through the knights tour problem or if that’s verboten dont ask that question, its bad
|
# ? Aug 28, 2018 15:25 |
|
qhat posted:I guess my sticking point with this role in particular is it demands a couple years experience, mostly because this company really does not need any more junior devs or people with little understanding of basic things like testing and best practices. none of those are core classes, though? other than networking fundamentals and some security stuff I'm learning all of that poo poo on the job
|
# ? Aug 28, 2018 15:31 |
|
qhat posted:Even then, I disagree that these bootcamps are all you need to be an effective software engineer long term, there's just so much poo poo that is taught in a cs program that you will never get from a 10 week boot camp. Networking, distributed systems, advanced databases, security, cloud engineering to name a few big ones that I really wish people at this company had even a cursory understanding of. i think your opinion of the typical 4-year cs degree program is overly rosy also i’m surprised you expect your fresh-out-of-school hires to be proficient in distributed systems and advanced databases
|
# ? Aug 28, 2018 15:32 |
|
Lol I just looked up the knights tour. That is indeed is rly bad question to ask in interview conditions.
|
# ? Aug 28, 2018 15:33 |
|
raminasi posted:i think your opinion of the typical 4-year cs degree program is overly rosy Maybe it's just me, but I learned a lot of stuff like that that I'm still drawing on many years later despite not doing it for a long time? The stuff I find most time consuming to learn is the stuff that I had 0 exposure to in college or work.
|
# ? Aug 28, 2018 15:40 |
|
those would all be electives in a CS program
|
# ? Aug 28, 2018 15:44 |
|
qhat posted:I guess my sticking point with this role in particular is it demands a couple years experience, mostly because this company really does not need any more junior devs or people with little understanding of basic things like testing and best practices. qhat posted:Networking, distributed systems, advanced databases, security, cloud engineering to name a few big ones that I really wish people at this company had even a cursory understanding of. my degree contained none of them
|
# ? Aug 28, 2018 15:48 |
|
qhat posted:I guess my sticking point with this role in particular is it demands a couple years experience, mostly because this company really does not need any more junior devs or people with little understanding of basic things like testing and best practices. I agree that bootcamp students (and other non CS graduates) are gonna play catchup for the first few years of their career... but after that? There doesn't seem to be much of a difference in practice. I think the most lasting thing my degree taught me is a method of problem solving that non-CS grads will eventually pick up (if they're successful at their jobs). As people have pointed out a lot of those things you listed aren't part of a core curriculum, and many of them are quickly dated... Distributed Systems for example, when I was in school almost 10 years ago we covered CORBA and looked at the MPI spec... now at best I vaguely remember those oh and also we just use REST interfaces that trade JSON back and forth now... The more universal concepts like election algorithms, logical clocks... I remember these exist and would probably pick them up more quickly if I started reading about them. But 1.) I still haven't found the need to 2.) it wouldn't be THAT much quicker than someone who had never heard of that stuff.
|
# ? Aug 28, 2018 15:49 |
|
nobody wants to hire juniors
|
# ? Aug 28, 2018 15:50 |
|
hobbesmaster posted:nobody wants to hire juniors eh its a great way to get a dev who won't figure out you're severely underpaying them for a year or two
|
# ? Aug 28, 2018 15:54 |
|
ADINSX posted:As people have pointed out a lot of those things you listed aren't part of a core curriculum, and many of them are quickly dated... Distributed Systems for example, when I was in school almost 10 years ago we covered CORBA and looked at the MPI spec... now at best I vaguely remember those oh and also we just use REST interfaces that trade JSON back and forth now... The more universal concepts like election algorithms, logical clocks... I remember these exist and would probably pick them up more quickly if I started reading about them. But 1.) I still haven't found the need to 2.) it wouldn't be THAT much quicker than someone who had never heard of that stuff. stuff like byzantine fault tolerance, synchronization, analyzing parallel/distributed algorithms etc is technology agnostic. as you say thats what you're supposed to remember out of the class
|
# ? Aug 28, 2018 15:55 |
|
ADINSX posted:eh its a great way to get a dev who won't figure out you're severely underpaying them for a year or two but you can underpay an intern even more!
|
# ? Aug 28, 2018 15:56 |
|
interview over. am I wasting my time trying to understand and improve our interviewing process and criteria because I’m not confident in it at allllllllll
|
# ? Aug 28, 2018 17:17 |
|
I also can’t justify why we shouldn’t ask the knights tour problem with anything other than “I hate it”
|
# ? Aug 28, 2018 17:19 |
|
i love hiring juniors and would only hire juniors if given the opportunity to build a team
|
# ? Aug 28, 2018 17:20 |
|
Think to yourself which of these things take up more time: - the coding - understanding an obtuse problem not relevant to the role If it's the latter, it's a bad question
|
# ? Aug 28, 2018 17:22 |
|
Pollyanna posted:I also can’t justify why we shouldn’t ask the knights tour problem with anything other than “I hate it” do you actually arrange knights on a chess board in your day to day? if you're hoping the candidate wrangles a graph problem out of it, why not just ask a graph problem that's relevant to your work
|
# ? Aug 28, 2018 17:22 |
|
In fact if you think your question even slightly resembles a brain teaser, it's no different to the nonsense trapped in a giant blender questions that Google gave up on a decade ago
|
# ? Aug 28, 2018 17:25 |
|
google still asks brain teaser coding questions despite that policy though
|
# ? Aug 28, 2018 17:31 |
|
idk at least in the part of goog i interviewed in a couple years back they didnt ask me any bullshit questions
|
# ? Aug 28, 2018 17:33 |
|
JawnV6 posted:do you actually arrange knights on a chess board in your day to day? if you're hoping the candidate wrangles a graph problem out of it, why not just ask a graph problem that's relevant to your work in my opinion this is better than a graph problem because it is at least based on rules that both parties know going into it i suck the most at interview questions where its just "build a thing that keeps track of the max of a stream of numbers and updates the max when i pop or unqueue" because its harder to implement something thats just abstract and not serving any kind of realistic purpose
|
# ? Aug 28, 2018 17:40 |
|
I got a pretty good phone screen question the other day if someone wants to steal it: "imagine we have a function that takes a string and we want to have it detect the language the string is in". I came up with a solution that would download a corpus of each language we care about (the interviewer actually had some stock code for this she pasted in so I didn't need to waste time on it). Then tokenize those source files and make a set of the words. Each word in our source string that is in one of the sets, add a point to that language. The language with the highest point value wins. There are better ways to do it of course, we could try and gauge how common certain words are in some languages, especially if they are NEVER used in any other language, thats really good proof its a specific language... so I said that'd be my next steps. And of course if this was real production code, google has APIs to do this, python has nlp stuff that does this, etc etc etc.
|
# ? Aug 28, 2018 17:42 |
|
i would spin up a million by million RNN, train it on every language's wikipedia, and then feed it the string
|
# ? Aug 28, 2018 17:52 |
|
Corla Plankun posted:in my opinion this is better than a graph problem because it is at least based on rules that both parties know going into it RUNNER: MARATHON :: A) envoy: embassy B) martyr: massacre C) oarsman: regatta D) referee: tournament E) horse: stable
|
# ? Aug 28, 2018 17:59 |
|
hobbesmaster posted:RUNNER: MARATHON :: thanks for the suggestion, will use this when phone screening applicants.
|
# ? Aug 28, 2018 19:12 |
|
I still need to collect my thoughts over today’s interview but I still don’t understand what exactly we want, and I’m not convinced that our approach is appropriate for what I do know we want aside from setting down ground rules for proctoring, and the fact that knight’s tour is not a good fit for a 45 minute timespan, it’s mostly a matter of “what the gently caress do we even want and how are we gonna get it” yes, it’s hard, but the answer isn’t to give up and do stupid poo poo
|
# ? Aug 28, 2018 20:07 |
|
One thing I like to do is think about what pissed me off about a previous new hire and test that thing specifically. Like our last experienced new hire outright refused to write unit tests until his manager smacked him down, and as it turned out he actually had no idea how to. From then on, I made sure to include that as a staple in my interviews.
|
# ? Aug 28, 2018 20:58 |
|
Pollyanna posted:I still need to collect my thoughts over today’s interview but I still don’t understand what exactly we want, and I’m not convinced that our approach is appropriate for what I do know we want This is a good perspective from which to approach the process, because if the interview panel can't agree on what the ideal candidate is, how are you all ever supposed to pick somebody? For the knight's tour problem, I'd probably mind it less if the interviewer straight-up started with, "We'd like to get some insight into what your problem-solving process is, as far as how you go about understanding a problem and starting to break it down. Picture a chessboard…" Because I've had a few interviews where I answer a question and everybody kind of just stares at me like I didn't say what they wanted to hear, but nobody will give me further direction, even when I ask.
|
# ? Aug 28, 2018 21:43 |
|
so I accepted that Boston position with a projected start date of Oct 1 and I’m still interviewing in Manhattan. 1 screen and an in-person scheduled. also I haven’t notified current employer yet. not sure if I’m doing this right. I’d probably feel bad if I changed my mind but here we are.
|
# ? Aug 28, 2018 23:08 |
|
qhat posted:One thing I like to do is think about what pissed me off about a previous new hire and test that thing specifically. Like our last experienced new hire outright refused to write unit tests until his manager smacked him down, and as it turned out he actually had no idea how to. From then on, I made sure to include that as a staple in my interviews. i can’t believe i never realized this but having a candidate write tests for a piece of your already existing code would be real good interview material
|
# ? Aug 28, 2018 23:41 |
|
Rex-Goliath posted:i can’t believe i never realized this but having a candidate write tests for a piece of your already existing code would be real good interview material Ask them how to unit test a function that calls to a remote database when you can't guarantee a network connection at the time of test execution. That can sometimes be a pretty good way to sift out peasants.
|
# ? Aug 28, 2018 23:59 |
|
Rex-Goliath posted:i can’t believe i never realized this but having a candidate write tests for a piece of your already existing code would be real good interview material sounds like an unpaid internship
|
# ? Aug 29, 2018 00:01 |
|
Pollyanna posted:I still need to collect my thoughts over today’s interview but I still don’t understand what exactly we want, and I’m not convinced that our approach is appropriate for what I do know we want i just looked up knights tour and lol. what does your company do again?
|
# ? Aug 29, 2018 03:53 |
|
Rex-Goliath posted:i can’t believe i never realized this but having a candidate write tests for a piece of your already existing code would be real good interview material no it wouldn't. tests are a thesis that you then attempt to prove. they don't mean poo poo if you start with the conclusion and derive a thesis
|
# ? Aug 29, 2018 04:51 |
|
I would never refer a fresh boot camp grad for an interview but if one sat down in front of me I’d def ask what tech theyve worked with outside the camp. God help u if you hire someone who signed up for a boot camp because they couldnt learn anything on their own
Gazpacho fucked around with this message at 05:36 on Aug 29, 2018 |
# ? Aug 29, 2018 05:03 |
|
the talent deficit posted:no it wouldn't. tests are a thesis that you then attempt to prove. they don't mean poo poo if you start with the conclusion and derive a thesis part of the interview could be "refactor this code so its easier to test". I know you're supposed to write tests first but if you're saying its literally impossible to write tests for code that exists thats probably a red flag on your part
|
# ? Aug 29, 2018 06:19 |
|
|
# ? Jan 25, 2025 19:44 |
|
ADINSX posted:I know you're supposed to write tests first nah. gently caress that.
|
# ? Aug 29, 2018 06:29 |