That is a good question I believe, it exposes how embedded knowledge software engineers actually have and that wouldn't be something to measure very quick. Every interviewer is likely just trying to confirm if engineer is who he/she claims to be during the application.
Did you know you are exposing your identity for each "permission denied" error? SSH repeatedly tries all of your public keys to authenticate. Someone was able to harvest that and figure multiple github profiles of the client in past.
This is a great book I have read in past. It primarily focuses on underlying math for interactive graphics but you will notice it will also touch photo-realism later. Indeed, it describes how raytracing works at chapter 10 "The Lighting Equation: Putting It All Together" and covers wide span of graphics knowledge.
You might got lucky or not, but your comment is very dangerous for the people who have no idea and haven't been in the field. FAANG compensation is not peasant to billions of people, so please don't do that.
I'll tell you a mini story about a coffee shop I visited few days ago. That place was hidden in yelp search when I looked for 'coffee & tea' in my area (their yelp page existed). While I don't know the actual reason why this happened, I immediately discovered that coffee shop using google (as a double check). It gave me a charm because it reminded me a fact that if you have the 'right service', people will find out. Given this flow, I started to believe gatekeepers might begin losing their odds.
I deeply regret clicking this. I've never worked with FAANG however I've been through 5-6 interviews. The way I see it, it's material that can be studied/recalled but you can make much more difference in your life with that amount of energy, unless you're a new graduate OR your job is interviewing other people. I understand the side effects that study can bring on to help ease the pain, though it won't work if you only rely on knowledge alone. That's where experience/context helps. Just be aware that fangs are not a measure of what a best software engineering can be, they're the ones make use of their branding well as a high entry wall. So please do yourself a favor and don't take the harsh to your heart, keep enjoying what you do.
I've tried thinking hard on the second video. I guess most confusion comes from as it doesn't explain why traction situation (where wheel speeds up relatively) is different than the actual turn (where wheel slows down relatively). Here, as speeding up wheel gear can not spin the worm gear, power is captured "back" to the other wheel through the outer locking cage. However, first comment at youtube and an another article (https://www.kmpdrivetrain.com/differentials/torsen-vs-plated...) states that it wouldn't work with 0 traction and can get only the amount of slipping traction torque to the other wheel.
I have to disagree.. Not everyone has enough time to learn "ready" systems either, they can get pretty overwhelming. When I decide to build instead of buy, it's mostly because I don't believe in the design of the product that is favored by thousand people with reason being "it just sounds cool".
This reminds me of nand2tetris course, which I think really helps to understand practical implementation of today's computers (https://www.nand2tetris.org)
This aligns with the idea of running what users have really downloaded at first. Apple removed apps who wanted to run ad-hoc code before. I'm curious why it waited years for realizing in-app app stores might occur similar issues.
Isn't that why gunicorn(+gevent) was implemented and does the switching behind the scenes w/o waiting that api call to finish? Is there a good reason I should manually "await" network calls from now on?
I thought this device would act as a keyboard but it requires another process to listen. Maybe sending 3 times esc + :q + return could quit in most cases..