Leetcode problems are more like university-level exercises and don’t always reflect real-world job requirements. Instead of focusing on whether a candidate can solve Leetcode-style problems, check if they have a Bachelor's or Master's degree in Computer Science. The real challenge in software engineering is understanding core principles such as dependency injection, immutability, clean code, testing, and system design. When conducting interviews, consider having an open discussion about what the candidate values in software engineering and why. Anyone who has studied can relearn Leetcode problems, but can they explain why dependency injection is valuable in a real-world system?
My new favorite coding channel, will continue to follow throughout my computer science journey
leetcode don't assess your technical skill, it assess your capabilities to be an human LLM yourself by learning and then redo, if you think this is the job of an engineers, then you are the reason why engineers have it bad right now
Agree with this take. It seems like the contracting world is entirely programming jeopardy instead of leetcode, since its so easy to fire contractors.
I am unsubscribing we want kick leetcode out of the interview and u are taking of combining.like why ?
Leetcode is mostly algorithms, not math as much. It comes into play for sure, but the majority of problems aren’t discrete math. I’ve done over 700 Leetcode questions for reference
You'll have a way easier time getting hired by one of these FAANG companies if you have a masters or PHD in CS.
Yeah, I don't care for the interview questions for leet code. But if that is what is required. I'll take a couple hours out of my work schedule every week to learn them. So, an hour every day to learn React + 2 hours to learn leet, makes 7 hours a week just to keep up with all this stuff. So, the business prefers that I use 20% of their time learning stuff that is rarely used or is bad for the customer (React) then I'll do that. Seems like a lot of wasted time. But, businesses seem like to like us devs to waste a lot of time rather than helping the customers with good code, then I'll do that for the business, because the business is my core customer even though I would prefer that the actual customer would be my customer.
I have two main problems with Leetcode interviews. 1) Given the niche nature of many Leetcode questions, how can I as an interviewee be sure the interviewer has thought about the question before giving it? If an interviewer came up with the question or it's a common scenario, the interviewer should have a general idea how long it takes to solve. But if it's straight up "rotate a linked list by x and tell me the value of the nth node", then my question after doing it is going to be "what challenges did you face when solving this". And if you can't answer, my trust in being on a team with you plummets. 2) Folks online used to make fun of script kiddies by calling them "L337" becuz they didn't actually understand the code they were running. And in the sweetest ironies, Leetcode has enabled this mindset due to folks just memorizing answers to pass coding interviews and not understanding what they were doing. We've come to a reality feared by the programmers of the 90s/00s: rule by the "l337" lol
The problem with leet code problems is when developers then incorporate them into their code. It leads to an unmaintainable mess. The code may look cool much like the code I learned when I was a kid, but it's difficult to debug when you nest 5 or six ternary operators because you think it will create the most optimal code. If you want to see whether your code is optimal, you have to profile it and look at the assembly code the compiler generates. But, that doesn't take knowing leet code problems. There maybe companies out there whose software require devs to know leet code, but that's not the norm.
As someone whos self-taught, I think you need both practical and leetcode. Practical showcases how you can adapt and implement, leetcode showcases understanding of optimizations with big O with time and the why. That being said a lot of leetcode problems fall into a find 30 different edge cases (kill me) or reading the question is like talking to someone who doesn't speak your language, these are bad for interviews imo because your not testing knowledge but patience with an unrealistic problem.
I don't think you need leetcode specifically but you should always have a technical evaluation that tests your ability to problem solve. In a project I was on, we didn't do that. The leadership only asked if they can answer questions about terms and concepts. We did not do a coding exercise or anything that tests your ability to problem solve. We got a bunch of people who claim to be mid level developer or senior, none were. I can tell on week 1 something was up and their experience and skills are deficient.
What kind of Leetcode questions do you ask in your interviews? Is it a mix of all difficulties or mostly popular Medium questions? The one thing I find difficult to justify with Leetcode are questions that are exceptionally difficult to do without previously seeing the problem. I realize it's extremely difficult to define which problems fall under this category, but some problems are essentially RNG in that if you have not seen it before, finding an optimal solve or a solve at all is nearly impossible. For context, I'm about 100 questions deep into my grind as I look to transition from Defense software engineering to a tech SWE role. I can do most Mediums now that I know core patterns and tricks but there are still many problems that I have simply not seen the trick to before and cause me to get stuck. That said, I'm not fully certain how to overcome that barrier when interviewing. Getting the interview is already difficult enough, and at that point I hope it's a question I have seen before. Do you have any recommendations to effectively prepare for the questions you might encounter, and do you have any thoughts on vetting questions that you think have a reasonable probability to be completable by a worthy candidate?
If I want to go into just web development and I’m not that mathematical am I just screwed?
Im not wasting my time on BS LC and dont tell me what to do if you're a recruiter, instead of wasting my time on LC i built couple projand now i work for myself and travel the world and soon ill be a millionaire based on my project growth, i dont trust companies anymore they make you do all that work to tell you sorry this job offer is canceled or fire you after few years its better to build your own project and win on long term
I would say yes and no. If you gain an understanding what leetcode trying to teach you which is optimisation strategies like for example prefix sums and what they are for what particular domain they solve yes. If your copypaste no.
What would you need these for?
College is definitely not a requirement
@dranon0o