2024 Presidential Election
Yesterday
1099
Donald Trump has become America’s first felon president.
2024 Presidential Election
Yesterday
928
🤡
Ask Blinders
Yesterday
282
Who can explain me what and how exactly Trump falsified business entries
2024 Presidential Election
1h
361
Trump is still leading to win US presidential election on PredictIt
Tech Industry
Yesterday
2237
Google silently firing a lot of people this week
I hear this shit all the time when people are talking about Leetcode and I'm really sick of this. Every CS sophomore learns how to invert a binary tree. Even if they didn't, it doesn't take a rocket science red-black avl splay tree theory to figure out how. The fact that you built a successful application does not mean that you're qualified to develop Google scale application. It's more about timing and luck. I don't want my application dependency to be managed by a program which is written by who don't understand basic tree structure anyway.
Yet you probably still use Homebrew anyway.
Sorry you were rejected.
Depends on what you’re going to work on...but really, I think doing leetcode problems on a whiteboard for an interview is one of the worst ways to waste the hour I get that could tell my about a candidate and if they’re going to succeed on my team or not.
Yeah, has near zero correlation with job performance other than it's become a sort of industry imposed IQ test. This guy is a better developer than likely 90% of Google engineers but he didn't see the point in cramming leetcode so they said no. That's really a failure on Google's part.
I think I once saw a follow-up comment from this guy somewhere else; he admitted that he had tweeted this when his emotions were running high. He wasn't feeling so caustic once he looked back at the situation with a cooler head. Gayle McDowell (CTCI) also commented on that post. Maybe in Quora.
You’re right, and he admitted that Homebrew is not well written and full of bugs. He also says that his personality makes him not work well with people. https://www.quora.com/Whats-the-logic-behind-Google-rejecting-Max-Howell-the-author-of-Homebrew-for-not-being-able-to-invert-a-binary-tree
That hubris, mothafuckas
I used to think leetcode was total bs before i joined. I have changed my opinion a bit since. Google has hands down the highest concentration of engineers (and even PMs because a lot of them used to be engs) with rock solid CS fundamentals. Couple that with design review process that feels likely academic-like paper writing, you end up systems that are rock solid and well thought out. Obviously this doesn't work well with product dev where you need to "move fast abd break things". But as an engineer I freaking love it here. Fellow engineers use a vocabulary that is precise and well defined (in CS and industry). This means that emails, designs and texts in general are precise, and free from fluff. Google's hiring must be doing _something_ right to create this atmosphere. So I have become less sceptical of LC style. Maybe it is like SAT that measures the baseline intelligence and more importantly grit, both of which are crucial for day to day job functioning. My 2¢.
I agree as well, as long as the problems aren't too cookie cutter. I like introducing another layer on the problem to make it more interesting and less abstract. As close as I can to a real word problem, that would require a LC like solution
Coming from Amazon, I have a different perspective. I think moving fast and launching early with a few bugs is much better than creating a perfect product/rock solid systems. Compromising a bit to simplify the system, is completely ok. Look at reinvent and see the number of things they launched. Google is very slow to launch things. After a while, it's not about how much code did I write. It's about what impact did my single line of code create. And Amazon, although has a bit lower hiring bar, offers more such opportunities. Here, quarters are spent building things which might not have any impact.
Leetcode questions just test it you are a quick learner. If you truly are, you can spend 3 months and solve any of them (of course there is still some bullshit such as "you were too slow, you needed 35 minutes to solve 2 problems instead of 30 minutes, try again after 1 year", but it isn't this case). And system design questions just test common sense. So HoneBrew inventor had common sense, ok, but he wasn't a quick learner. Google needs both qualities, case closed.
You are missing one other quality that LC questions test, which is dedication. You are assuming that Homebrew guy is not a quick learner because he failed the interviews, but in reality he probably never even tried to learn D&A in the first place. He thought that his experience alone would be enough for G, but G makes it clear that is not the case, and you also need to be willing to be dedicated/committed enough to play their game. I know many great engineers who would never get hired at G because they refuse to play the game.
Google needs 1000s if people to do grunt work. Keep on leetcoding.
Then google lobbied to have more H1b because they can’t find qualified engineers in usa. Bunch of hypocrites
Switch to Windows