|
Doom Mathematic posted:Teachers should be paid as much as tech workers. as much as politicians would be a good start
|
#
?
Mar 12, 2022 14:06
|
|
|
|
| # ? May 26, 2024 21:42 |
|
|
toiletbrush posted:
not really, no it’s best to have a short, sweet and positive story for any role.
|
|
#
?
Mar 12, 2022 14:18
|
|
|
toiletbrush posted:Whats a good approach to talking about previous roles that you left basically because your job satisfaction comes from delivering, and nothing got delivered, without sounding negative or like you're making excuses? Is it ok to say 'I really didn't enjoy that role, can I talk about my other jobs?' You can't use that language, but can you phrase it as something like: "That place required a lot of consensus-building before advancing any initiatives, so it was much slower paced than my experience at X (other job you're wanting to switch convo to). I appreciated the opportunity to develop those consensus-building skills, but preferred the $delivery and $speed and $impact I saw where we could move fast at X. Your role sounds exciting to me because I would be able to contribute $skillset to projects that are $more buzzword impact / delivery / language." e: and yeah, you should have a short way to tell a positive story about every role. you can try to spin a win from getting everyone on board to an initiative, even if it was so excruciatingly slow you didn't stick around for execution. or whatever similar thing you can come up with. jemand fucked around with this message at 14:43 on Mar 12, 2022 |
|
#
?
Mar 12, 2022 14:40
|
|
|
toiletbrush posted:I'm currently stagnating like hell in my current contact, my umbrella company just doxxed me and a few hundred other contractors by CC'ing rather than BCC'ing everyone on an email to tell us that they've not been calculating our tax correctly and we all probably owe a bunch of money to HMRC, and a good mate from a couple previous roles just hooked me up with his HR as they're hiring...deffo time to hit the interviewing thread for some advice. quote:So in my current role, I've achieved very little worth talking about (endless meetings and red tape, terrible processes and a codebase utterly strangled by tech debt), and writing about it on my CV was really hard. I'm going to have to talk about it in my interviews, too. Whats a good approach to talking about previous roles that you left basically because your job satisfaction comes from delivering, and nothing got delivered, without sounding negative or like you're making excuses? Is it ok to say 'I really didn't enjoy that role, can I talk about my other jobs?' You should ask your friend how to talk about it. I wouldn’t assume that I your interviewers will be aware of your friend’s history or, if they are, connect the dots. “I’m working on this project, this is what it’s going to do, this is where we’ve gotten so far, this is what I have contributed. We are facing [obstacle]. I think I’m learning that I’m most productive in an environment where [thing new company would provide].”
|
|
#
?
Mar 12, 2022 15:46
|
|
|
Well that "chat" turned out to be more of an interview, but I guess it made sense. probably should have been a bit more prepared, but it seemed to go well. they are gonna send me an "automated testing" test where i have to write tests for an api they send me so Its off to back up all that bullshit about python to make sure i don't fall on my face there.
|
|
#
?
Mar 14, 2022 22:51
|
|
|
tk posted:loving hell Just got to see if they liked my cv enough to get me an interview.
|
|
#
?
Mar 15, 2022 14:07
|
|
|
I am having trouble trying to figure out what, in general, I should be learning from doing some of these leetcode exercises. Sorting, searching, and graph traversal are all useful. In fact, I have used what I have learned solving these problems at work. But things like palindromes, what is the point of that? I spent hours banging my head against a problem to find the longest palindromic substring and my solutions were always too slow. When I look at the solution, I find that the reason it works is that the "fast" solution works from the middle of the tested palindrome out, whereas I was working from the outside in. It was easy for me to come up with an example where my solution was faster, though. So the best I can come up with as the lesson here is that palindromes of the kind tested on leetcode are "broken" more quickly from the middle. Is that generally true? Moreover, when I get to a question, I can usually (most of the time?) come up with a functional but slow answer, because that is how I usually work when there isn't an obvious set of test cases I need a way to generate them, so I make a way I know will work in order to make some test cases. Then I start looking for ways to make it faster. But the point of these trivia questions seem to be just crank out the optimal solution in one go. I feel like I'm doing something wrong here, but I don't know what I'm supposed to be doing, so I don't know how to course correct.
|
|
#
?
Mar 16, 2022 18:04
|
|
|
stuff like palindromes are to see if you can apply anything better than o(n^2) (or worse) to a “real” problem. top down vs bottom up isn’t really the point, just so long as it doesn’t have a terrible runtime. typically you at least talk through the “straightforward “ approach first, maybe actually code it so something is on paper.
|
|
#
?
Mar 16, 2022 18:08
|
|
|
if you're not doing it either because it is fun, or because you need it to get a job by some weird mechanism, it is probably indeed not useful. but then very little programming is.
|
|
#
?
Mar 16, 2022 18:09
|
|
|
I guess it counts as “fun” but I find it’s good to go do some of those problems every once in a while to remember that everything isn’t just gluing CRUD apps together…
|
|
#
?
Mar 16, 2022 18:12
|
|
|
MickeyFinn posted:I am having trouble trying to figure out what, in general, I should be learning from doing some of these leetcode exercises. Sorting, searching, and graph traversal are all useful. In fact, I have used what I have learned solving these problems at work. But things like palindromes, what is the point of that? I spent hours banging my head against a problem to find the longest palindromic substring and my solutions were always too slow. When I look at the solution, I find that the reason it works is that the "fast" solution works from the middle of the tested palindrome out, whereas I was working from the outside in. It was easy for me to come up with an example where my solution was faster, though. So the best I can come up with as the lesson here is that palindromes of the kind tested on leetcode are "broken" more quickly from the middle. Is that generally true? as you get more arcane the risk of algo knowledge being useless goes up. but this isn't like playing on train tracks where there's risk without reward, there's a lotta product quality and professional satisfaction from pulling out a truly weenie solution that actually solves problem. longest subsequence poo poo specifically is omnipresent in computational genetics poo poo. longest palindrome is basically an introductory tractable problem that you sidle at subsequence problems with to learn em i've shipped compressed sensing poo poo in prod. ive shipped neural nets and not the crack-em-open-out-of-a-can poo poo in prod. ive shipped fourier basis poo poo in prod. but i've learned like 30 mathematically sophisticated things that entail their own study that im never gonna ship on prod or just haven't shipped on prod (never shipped mixed integer programming poo poo tho i learned it, never shipped percolation deanonymization, never shipped markov net altho i did ship bayes net once, etc etc). so like dozens of rolls on the dice, a few hits. the rolls of the dice have to cost you something more like 0.5-6 months than years for it to be worth it but there's no recognition of that risk reward thing in leetcode poo poo because thats just not how interviewing works. in real situations where peeps hire phd's they recognize and remain conscious of the underlying risk of specialization in addition to the rewards but interviews are also a power thing so they can't admit risks and doubt and poo poo bob dobbs is dead fucked around with this message at 18:18 on Mar 16, 2022 |
|
#
?
Mar 16, 2022 18:14
|
|
|
My research involves designing and implementing new algorithms and I find leetcode to be tedious bullshit. It's a hoop you have to jump through as part of the interview process, and we'd all be better off if that weren't the case, but it's not going anywhere any time soon. Suck it up and do what you have to do in order to get the jobs.
|
|
#
?
Mar 16, 2022 18:17
|
|
|
hobbesmaster posted:I guess it counts as “fun” but I find it’s good to go do some of those problems every once in a while to remember that everything isn’t just gluing CRUD apps together… for sure. currently preparing a new summer course for this kind of problem solving, so practicing my realtalk "no, this will not change the world and/or make you good at ai"
|
|
#
?
Mar 16, 2022 18:18
|
|
|
LeetCode is stupid bullshit that is almost never worth your time unless you don't currently have a job or the job you're applying for is going to pay you serious figgies. If you pay the premium fee then they outright tell you the questions asked at the big companies. Other than that it's completely and utterly worthless.
|
|
#
?
Mar 16, 2022 18:50
|
|
|
Remember when companies used to ask dumb gently caress brainteasers in interviews and expect you to translate a problem of gnomes stuck in the ground to some kind of binary checksum problem? Be glad we don't still exist in that world.
|
|
#
?
Mar 16, 2022 18:53
|
|
|
MickeyFinn posted:I am having trouble trying to figure out what, in general, I should be learning from doing some of these leetcode exercises. you should be learning the specific leetcode questions so that when an interviewer asks you them, or questions very similar to them, you can quickly regurgitate the answers without having to think too hard. (make sure to pretend to think, though. breaking the kayfabe conveys arrogance.)
|
|
#
?
Mar 16, 2022 18:57
|
|
|
raminasi posted:you should be learning the specific leetcode questions so that when an interviewer asks you them, or questions very similar to them, you can quickly regurgitate the answers without having to think too hard. (make sure to pretend to think, though. breaking the kayfabe conveys arrogance.) This is maybe too cynical for me (and that is saying something). Are you being serious? ultrafilter posted:My research involves designing and implementing new algorithms and I find leetcode to be tedious bullshit. It's a hoop you have to jump through as part of the interview process, and we'd all be better off if that weren't the case, but it's not going anywhere any time soon. Suck it up and do what you have to do in order to get the jobs. This is basically where I was at when I started, but I guess I got confused when I got something useful out of it.
|
|
#
?
Mar 16, 2022 19:17
|
|
|
MickeyFinn posted:I am having trouble trying to figure out what, in general, I should be learning from doing some of these leetcode exercises. You should be getting comfortable doing those sorts of exercises in an interview environment. If I'm asking you a leetcode type question I'm going to pretend like I just asked you to do some extremely basic coding task, and I want to know that you can do that without me holding your hand. Take the question, break it down, ask clarifying questions, walk through the approach you're going to take, work through some test cases to find edge and error cases, write the solution, walk through the test cases to make sure they actually work, then tell me the direction you would want to take this to make it The Best It Can Be.
|
|
#
?
Mar 16, 2022 19:18
|
|
|
Oh and make sure if there's an easy brute force approach and you can write it out in a few minutes, do that so that you have a working solution at the very least. If you're lucky can get away with not finishing the real solution bug-free if you've already done something which works and you've discussed your approach for the optimal solution in detail. I hate to say this because I absolutely loving despise these questions, but if you're applying to these big companies then you should be buying and reading CTCI and doing these questions at least a month in advance. These companies actively invest in LeetCode and in all likelihood pay people like Gayle Laakman McDowell to perpetuate this notion of a tech interview, and they expect you to behoove yourself by playing ball. There's also a flawed notion that if you interview "generally" for programmers you're going to avoid creating silos, which is almost never the case in reality, and generally it's difficult to find and train people willing to interview. A lot of the interviewers often have no clue at all what domain they are interviewing a person for, and also have no experience in the domain of hiring, which is why these questions end up being asked. It's loving lovely, but that's just the environment at the big companies. If a small company with a small team asks you these questions, my opinion is don't bother, there is no way they are so detached from the domain of hiring that they have to resort to hand waving like LeetCode.
|
|
#
?
Mar 16, 2022 19:34
|
|
|
MickeyFinn posted:This is maybe too cynical for me (and that is saying something). Are you being serious? 100% serious. interview processes based around this type of question are ostensibly looking for your ability to break down questions, organize your thoughts, and approach problem solving and communication in a disciplined, effective way, but it’s much easier to do all of that when you know the solution out of the gate. a less cynical framing is that you should be practicing not getting nervous when presented with this type of riddle in a high-stakes environment, but since the best way to do that is to just memorize the answer key, i don’t think that’s materially different from my interpretation.
|
|
#
?
Mar 16, 2022 19:35
|
|
|
Yeah you aren't learning that much from solving the palindrome really, you're learning to recognize the types of algo questions that can be reasonably asked of you in a one hour time span. Eventually you'll do enough of them that there aren't many rocks left to turn over in a 60 min session with a stranger. Your problem turns into "how is this like something I've already done" instead of "solve this new issue with no reference materials under pressure"
|
|
#
?
Mar 16, 2022 19:42
|
|
|
if you're interested in knowing why the middle-out solution to the palindrome puzzle is better than the outside-in version, consider how many candidate palindromes you need to consider. With the middle-out solution, after finding out that (Ai, Ai+1) is not a palindrome, you have also ruled out every other candidate centered at that midpoint without needing to explicitly consider them. With the outside-in approach, after finding out that (Ai-100, Ai+101) is not a palindrome ... you have learned nothing at all about any potential shorter palindromes, you still need to check them all. This is a common trick in algo puzzles, solving them in one direction will allow you to just skip doing a lot of the work. Identifying when that's the case is pretty essential.
|
|
#
?
Mar 17, 2022 00:09
|
|
|
Yeah there's a lot of those problems that become extremely easy if you just look at the problems from just the right angle. They do sort of border on brainteaser imo, if you know the trick then the problem is trivial, else you'll spend hours on it. Thankfully it's possible to recognize the issues in advance, like if you know your result array already has enough space for the result set, that means build your results backwards. If your recursive solution is not tail recursive, that means it does not translate easily to an iterative solution and you have to use an explicit heap allocated stack at the very least. If the problem seems to be of exponential complexity then you may have to do a DST or BST. DST means use a stack, BST means use a queue. If you read CTCI it does actually explain a lot of this, given that these jobs can often in the 6.5 figgies range, it's well worth the hundred bucks or whatever (as dirty as it might feel).
|
|
#
?
Mar 17, 2022 01:55
|
|
|
qhat posted:it's well worth the hundred bucks or whatever (as dirty as it might feel). poo poo, i had no idea typing "cracking the coding interview pdf" into google cost $100
|
|
#
?
Mar 17, 2022 02:02
|
|
|
i think someone in yospos or twitter had a google recruiter send a pirated copy to them lol
|
|
#
?
Mar 17, 2022 02:33
|
|
|
I'm sure the author doesn't give a poo poo. She probably gets mad bank just to mention those companies in the book.
|
|
#
?
Mar 17, 2022 02:39
|
|
|
the first hit for that google search is a GitHub 
|
|
#
?
Mar 17, 2022 02:47
|
|
|
PCjr sidecar posted:i think someone in yospos or twitter had a google recruiter send a pirated copy to them lol This definitely happened to me. It was some old copy. I didn't find CTCI particularly helpful, except as a recipe book, which I can just google. I think the trivia interview has moved on to being the endless list of leetcode problems, which the bootleg copy I was given didn't cover. And the advice about how the interviews are conducted was quaint. Things like "you aren't expected to finish the solution in its entirety, bug free, and optimal on the first go," has not been my admittedly limited experience. Jabor posted:if you're interested in knowing why the middle-out solution to the palindrome puzzle is better than the outside-in version, consider how many candidate palindromes you need to consider. I had been circling this point today, thanks for making it explicit.
|
|
#
?
Mar 17, 2022 04:31
|
|
|
you dont have to get the leetcodes right to work here but it would help
|
|
#
?
Mar 17, 2022 05:29
|
|
|
MickeyFinn posted:This definitely happened to me. It was some old copy. I didn't find CTCI particularly helpful, except as a recipe book, which I can just google. I think the trivia interview has moved on to being the endless list of leetcode problems, which the bootleg copy I was given didn't cover. And the advice about how the interviews are conducted was quaint. Things like "you aren't expected to finish the solution in its entirety, bug free, and optimal on the first go," has not been my admittedly limited experience. I’ve been on both sides of this. I’ve gotten through tech interviews just fine without actually completing the problems in the most optimal way, and I’ve also put people through when they didn’t get a working solution in the end but the approach was very good and given a bit more time they would’ve got it. And this was for senior positions. Yes it depends on the interviewer, and yes you do have to actually have to undeniably be on the right track. It also largely depends on how you interact with your interviewer, like talking a good game won’t save you if you’re just floundering around, but they might not judge you as harshly than if you just sat there in relative silence. So yes it all depends and the only thing you should take from that is that interviewing is garbage and literally nobody likes doing it because it’s a crapshoot.
|
|
#
?
Mar 17, 2022 10:20
|
|
|
Hey YOSPOS, I'm about to go on an interviewing run which I haven't done in ~5 years. I think I'm paid well but the only way to be sure is to get some offers in and see. Since I'm still (happily?) employed I can be a little tactical about this. Usually I sling 7 resumes a day x 5 days = 49 a week, I usually get 7 - 10 phone screens out of that but I think I want to shoot for a bigger household name company this time around.qhat posted:I hate to say this because I absolutely loving despise these questions, but if you're applying to these big companies then you should be buying and reading CTCI and doing these questions at least a month in advance. ^^^ Should I: A) Start hitting CTCI for a month and some change, then sling interviews to the bigger companies B) Sling interviews to companies I don't necessarily care about + CTCI C) Sling like crazy now but potentially be unprepared for a bigger companies interviewing process What's the best way to get that interview muscle back?
|
|
#
?
Mar 18, 2022 18:27
|
|
|
Rudest Buddhist posted:Hey YOSPOS, I'm about to go on an interviewing run which I haven't done in ~5 years. I think I'm paid well but the only way to be sure is to get some offers in and see. Since I'm still (happily?) employed I can be a little tactical about this. Usually I sling 7 resumes a day x 5 days = 49 a week, I usually get 7 - 10 phone screens out of that but I think I want to shoot for a bigger household name company this time around. How about sling interviews to the companies you want + study for the interviews? It ain't like they're gonna get your resume and go holy gently caress Rudest Buddhist applied?! Call them NOW NOW NOW. Figure out what is asked in the interviews at those companies. If they're household names there are tons of people writing about their interviews. Then study the things that are in the job description in a way that will help with the interview questions. Apply to multiple reqs at big companies. Rant: Hiring managers often don't talk to each other and the procedures by which a req get posted online can be loving dumb. E.g. they post for Systems Engineer 3 online because they want to promote an internal 2->3 but have some policy that says they need competing candidates. So that job won't get filled ever. Then at the same time theres a near identical job req for a different team that says Systems Engineer 3 that is a real req for an outside candidate. The real one doesnt even know about the fake one. Yes that is mind bogglingly stupid. I have personally seen it at a company that is #3 in it's industry and a household name with thousands of engineers. CarForumPoster fucked around with this message at 18:41 on Mar 18, 2022 |
|
#
?
Mar 18, 2022 18:39
|
|
|
CarForumPoster posted:Apply to multiple reqs at big companies. Oh hell yeah, I didn't realize you can just keep taking multiple cracks at them. Thanks!
|
|
#
?
Mar 18, 2022 18:46
|
|
|
CarForumPoster posted:How about sling interviews to the companies you want + study for the interviews? It ain't like they're gonna get your resume and go holy gently caress Rudest Buddhist applied?! Call them NOW NOW NOW. I mean these days they can get back to you pretty quickly and get a tech screen set up within a week. I would honestly study diligently for a couple weeks at least just so you can take your time to absorb the information, and then start the application process. Go on to glassdoor and read about how the company likes to interview, you can send out applications immediately to companies that don't look like they lean heavily on leetcode, but really it's up to you. Mostly the coding training is just about recognizing general patterns in the questions and getting good enough at that that you don't need to memorize everything. Once you've seen enough of the questions and you've been over them multiple times, you'll just kind of know the correct direction to take without really having to think about it.
|
|
#
?
Mar 18, 2022 19:46
|
|
|
Rudest Buddhist posted:7 resumes a day x 5 days = 49 a week Yeah, I'm guessing you may want to brush up a bit before your interviews  But seriously, you may want to do a little prep beforehand to research what the itnerview process is like at the big boys and start your interview prep beforehand so you can hit the ground running. When I first applied at a bigtechco I realized how over my head I was when the recruiter actually phone screened me and I asked them to delay my interview for a month so I could cram, and they were fine with that. These big companies are basically always hiring so you can ask to delay interviews to give you more time if you want.
|
|
#
?
Mar 19, 2022 02:53
|
|
|
qhat posted:Oh and make sure if there's an easy brute force approach and you can write it out in a few minutes, do that so that you have a working solution at the very least. If you're lucky can get away with not finishing the real solution bug-free if you've already done something which works and you've discussed your approach for the optimal solution in detail. The bruteforcing + optimizing your code solution has been my go to solution for leetcode when I have no idea how to proceed. 
|
|
#
?
Mar 21, 2022 17:26
|
|
|
Code test seemed to go well. A few little mistakes in the code (types, forgetting i changed something from a string to a list) but felt like i explained the concepts strongly enough. Questions at the end went well and the interviewer went over time answering one of my questions so he didn't want to bail immediately!
|
|
#
?
Mar 23, 2022 19:43
|
|
|
a hiring manager for a role casually mentioned to me that two team members had recently transferred internally to other teams. it’s not taboo to ask why and what their tenure was, is it?
|
|
#
?
Mar 27, 2022 15:01
|
|
|
my homie dhall posted:a hiring manager for a role casually mentioned to me that two team members had recently transferred internally to other teams. it’s not taboo to ask why and what their tenure was, is it? sounds like self preservation rather than taboo
|
|
#
?
Mar 27, 2022 15:04
|
|
|
|
| # ? May 26, 2024 21:42 |
|
|
speaking of transfer internally, my job hunt has taken an interesting turn. my boss, who I hate, who knows I’m trying to leave because I’m not happy with what we are working on, offered me to take over as lead dev on our semi-internal tool we use to build stuff. our lead dev left about 6 months ago and they haven’t filled the role, and they would bump my pay significantly. this will let me finally fix all those loving bugs I have to work around and when we sell poo poo it will be less “ok so the docs say this but it actually behaves this way and you must remember that or you will waste lots of time” and my team will get a better tool (hopefully) some downsides: I haven’t written c++ in ages still would have same boss there’s some tech debt. this hasn’t been actively developed on in almost 2 years (why the original dev left, he was shunted off to help other random projects) upsides: pay me now bitch they will pay for a bunch of training at my request if it doesn’t work out I can revert to my current role (with pay decrease but still, small safety net) the tool already exists and I know how it works from user level so less time “getting to know the product” I can set product release and feature schedule so I can prioritize what I know needs to get done va whatever the gently caress was happening before. the level of documentation left for me is staggeringly complete. I worked on this in like… 2016 so I have a vague familiarity with the code base already. I’m pretty sold on it for the pay bump, change in title, etc, and I will probably keep looking for a job anyway. this also helps me bring my code skills back to the forefront and I could pivot to some pure code role later down the line for making my exit I’m going to do it but I need to make sure to get a lot of poo poo in writing.
|
|
#
?
Mar 27, 2022 16:13
|
|
