2,343 karma · joined October 8, 2016
This is just the most recent example I remember, unfortunately there have been enough at this point that if I'm doing anything important (such as filling and submitting some important form), I do it in Chrome because I don't want it to silently break in FF and cause other issues with invalid or corrupt data submitted to the service.
And they mentioned an Android update, but what about the millions of Android phones that aren't getting regular updates? Does that mean there's potentially millions of phones that can't dial 911?
I like Pixel and Android, but am seriously thinking of switching to iphone because I really need a phone I can trust will dial 911 when I need it.
So restaurants are getting squeezed on all sides. Supply costs up, wages are up, labor shortage, rent is up. Tips will go down as customers are upset with poor service, which means more pressure on wages. All of this means higher prices, which means fewer customers and more takeout orders. The old restaurant business model is simply not sustainable.
Buses have failed for a variety of reasons. Rail and BRT (bus with dedicated lane) are good and should be greatly expanded, but they only help with connecting major areas. We need to stop looking at Asian mega cities as an example. It just won't work here. Those super dense cities are designed for walking + public transit. The "last mile" problem in those cities are solved by walking because they're super dense enough that you can get to public transit easily by walking less than a mile. And also don't look at NYC or any old east coast city as an example. Those were designed before the car. Most people in the US live in places that were designed for the car.
Rail + BRT along major routes is part of the solution. But in the past, it hasn't worked well because of the "last mile" issue, which is more like the "last few miles" issue in the typical American suburban design. If someone has to still have a car to drive the few miles from their home to the park-and-ride, that doesn't do anything to fix the local road congestion issue. When they get on the train, it helps with freeway traffic, but does nothing for the local traffic. And the "last few miles" is too far to walk. And again, buses/shuttles have failed to fix this for a lot of reasons.
When ride-share was introduced, some thought that this was the solution. If someone depends on ride-share to get to places, that certainly helps reduce parking requirements. But, the traffic on the roads doesn't change, and may even increase due to idling ride-share cars waiting for rides. Uber Pool tries to help with this, but it's just not enough.
With the rise of electric micromobility (e-scooters and e-bikes), we finally have something that can fix this "last few miles" issue. But it will require a big step from the government and society as well. We have to be willing to sacrifice a lane in all the local roads. The current bike lanes are simply not extensive enough, not wide enough, and not safe enough. The only way to achieve true effectiveness for e-micromobility is to dedicate an entire lane of roads to them. It's a radical idea that would get a lot of opposition. People already upset with traffic don't want to lose a lane to make it worse. That can only be countered with a critical mass of people who own an e-microbility solution (e-scooter or e-bike).
Is it ever going to happen? Maybe if there's a city brave enough to try it and show it can work. But true support and infrastructure for e-bikes and e-scooters is the only way to get real density in the typical suburban sprawl layout without a disastrous increase in traffic and parking shortage.
- Government policy should require a regular hybrid as the base minimum in all new cars. It's 2021, there's no excuse now for a car to be pure ICE.
- New policy to favor PHEVs. The battery in one full EV can build four PHEVs. A 50 mile PHEV gets most of the benefits of CO2 reduction of a full EV. Also no new infrastructure is required, overnight level 1 charging is good enough. The suburbs will buy PHEVs (since they have a garage), the urban areas will buy hybrids (which will now be the minimum).
- All utility type vehicles (mail trucks, garbage trucks, delivery trucks), must be full EV. These vehicles are the perfect fit for the current battery technology (overnight centralized charging, frequent stops on short routes).
- Slowly raise the gas tax. Slowly is important otherwise there will be riots. But the impact on low income population should be minimal with the increasing supply of used hybrid and PHEVs.
The key feature of this plan is that it's a fast way to reduce CO2. Instead, look at what the infrastructure bill and the proposed BBB is doing. Billions on new charging stations, current version of BBB would reduce PHEV tax credit. Is this really the best way to get quick action on reducing CO2 from cars? Aren't those billions better spent getting hybrids and PHEVs (which have none of the charging or range issues of full EVs) to the public as quickly and widely as possible? Why is the goal suddenly to make all cars full EV, isn't the actual goal to reduce CO2 as fast as possible?
But we’re talking about young kids here. Pre-school (age 2 - 4) is an especially critical time and early elementary is important too. This time is a short window in brain development when the brain is super elastic and learns many things. Much of language development isn’t just hearing and speaking, it’s also seeing the mouth movements. And this is the age when the brain learns how to recognize and display facial emotions. Is it possible by masking so much these young kids, we’re doing major damage to this learning process?
Maybe the answer is yes, or maybe it’s no. But no one is even considering this question. It’s just endless stupid debates about masking that doesn’t at all consider the special circumstances of this critical time in life, this short window when so much happens.
It seems they care enough about wildfire data to make this big update for Maps, but can't do something this basic? Their headquarters is in one of the most affected regions in Caifornia by wildfires (they had an orange sky last year), and no one at Google headquarters thought that maybe showing AQI in the weather widget should be restored? (it's not even a new feature, just bring back the old AQI panel). What's the problem here, too simple for someone's promo-packet?
And I've never seen anything close to 5.9% for COL from any company. Most companies don't give that even for top performers. I don't think companies are prepared to give 5.9% COL, most companies are already so short sighted they would rather lose their entire engineering department instead of give raises to keep up with the market rate (even though they would be willing to pay that market rate when hiring a replacement). That kind of short sighted thinking simply does not allow a massive COL increase, no matter how much data is out there.
Reading and writing are two sides of the same coin. So why not test for the side that works much better for interview conditions and also allows more signal between a junior vs senior levels of experience?
- Car switching in/out of my lane, I manually take over
- Tight curve in the freeway, manually take over
- Very frequently check the dashboard indicator that shows if the sensors "sees" the car front or not
- Anything unusual like construction, cones, car on shoulder, manually take over
- Anything that looks difficult like weird merging lanes, manually take over
- Any bad weather or condition like sun directly in front, manual drive
- Frequently adjusting max speed setting on ACC. It's safer to not be too much above the prevailing speeds. Otherwise, if ACC suddenly becomes blind, it can accelerate dangerously as it tries to reach max set speed.
- I don't trust lane keep much, it's mostly a backup for my own steering and making my arms less tired turning the wheel
The key thing is to recognize just how dumb this technology is. It's not smart, it's not AI. It's just a bit above the old cruise control. With that mindset it can be used safely.
They also used shady practices like paying OEMs to not use AMD CPUs. They were heavily fined for this. If they try to do it again, the government will likely give a greater punishment this time since they would be repeating the same crime. Especially now with the anti-trust friendly sentiments within the government.
* The blog post complains that even experienced engineers they interviewed couldn't write some simple code on paper. Did anyone stop and think that maybe it's writing code on paper in an interview setting that's the problem? Stop and think about this. Imagine writing code like you usually do in a text editor. And then writing code with pen and paper. Backspace? Can't do that. Copy/paste some line to move code? Can't do that. Add some newlines? Can't do that. Run the code using the compiler? Can't do that. Of course most developers, even experienced ones, will do poorly under these conditions, this is not how code is normally written. On top of the already stressful situation of an interview, there's this suddenly completely foreign and limited way to write code.
* Even the blog post only mentioned fizzbuzz as a quick, easy question to filter out zero coding ability. Another way to say this, it was meant to filter out coding skill <= 0. Not skill <= 50 or skill <= 90. Zero. After that, the interview was supposed to move on to things that actually matter for software engineering. The present day interview has strayed extremely far from this. These days, you get asked two leetcode questions during an interview and even if you perfectly answer the first one, if you don't get the second one then you're an automatic fail. That was certainly not the intent of the original fizzbuzz, since answering any leetcode clearly shows coding ability. And the interview never has a chance to move on to the part that actually discusses software engineering skills. Or even if it does and they do very well on those interviews, the failure of only getting 1/2 leetcode questions correct overrules everything else since they must avoid "false positives".
Email addresses need the same regulation. The arguments that lead to phone number portability apply to email addresses as well. And I would argue that email addresses are even more important than phone numbers at this point (it's the single key to all online accounts, most bills, documents, statements are emailed as PDFs, a lot of government services expect a working email address).
Email addresses have become critical and portability needs to become a requirement for all email services. There are technical issues, for example if someone cancels Gmail service, how can the @gmail.com address be moved elsewhere? It's not as simple as phone number portability. Maybe a regulation that any email service must provide forwarding service to another email address even if the service is no longer active? Or maybe a trusted mapping that exists outside any single service, kind of like a DNS for email addresses.
So don't talk about the environment and say full EV is better than PHEV. Make other arguments if you want, but don't use climate change for your position, the math is clear, one Tesla = four PHEVs and 95% of trips are under 30 miles.
You can still hire in other time zones, but each time zone team must be an independent unit. So you can have a team that has members in Vancouver, Bay Area, and Los Angeles. And another team in New York, DC, and Atlanta. But these teams should rarely need to communicate and work with each other.
The Prius is one of the most reliable cars on the market. This whole EV = less complex = more reliable idea is way overrated by the EV community.
1. Government requires 50 mile range (about 20kwh battery) in all new cars. PHEV quickly becomes the default for all models (cars, SUV, trucks). Battery supply is less of an issue (A single 80kwh Tesla battery can be used in 4 PHEV cars).
2. DC fast charging (250kw) infrastructure is built in all cities. A small PHEV battery (20kwh) + DC fast charging means battery can be completely charged in minutes, as fast as filling up a gas tank. Maybe DC fast charging could even be built right next to the gas pumps (fill up gas and battery at the same time) This makes PHEV work for those without charging at home.
3. PHEV works fine in all weather. Battery range drops during winter? Just use some more gas, there's no range anxiety.
4. People get used to the feeling of driving in EV mode. And they notice how nice and quiet it is compared to when the ICE warms up. Also the environmental benefits.
5. Car manufacturers and battery companies get solid state batteries ready for mass production. Now is the time for 300 miles+ full EV cars. With solid state, no cold weather problems, much faster charge rate, no fire risk, no liquid cooling system, much higher durability. This is the battery tech that can fully replace ICE.
Such sites should be using minimal or no JS. These aren't meant to be pretty interactive sites, they need to be solid bulletproof sites so people can get critical services. And I haven't even mentioned how SPA sites often lack any accessibility features (which is so much easier to implement if sticking to standard HTML+CSS and no/minimal JS).
At this point, many teams (Waymo, Tesla, Uber, etc.) have been working on this for years. And there must have been times when the engineers thought, this would work so much better if only... Maybe it's time to gather these thoughts and sort them by low effort/high reward. For example, if all cars had a cheap sticker indicating forward vs back, would that improve self-driving performance by a lot? It wouldn't be that hard to get car manufacturers to add a sticker somewhere. Or what about traffic lights? Maybe adding a cheap light or sensor to it would vastly improve self-driving performance. Or those road reflectors, can they get some minor changes to help the self-driving software?
It doesn't seem likely that level 5 or even level 4 will happen when self-driving is literally defined as replacing what a human driver would do. But what if the computer had some help from both infrastructure and other vehicles? Why not start thinking about cheap changes that can be implemented to fix the blind spots and flaws in the self-driving software? If these changes are cheap and easy to implement and would vastly improve the software performance, isn't that worth it for true self-driving to work?
I need some help interpreting the location data. For example, if you click on Los Angeles, there's a bar chart with median and different percentiles. I assume this is across all engineer levels (I - V). But I'm interested in compensation for a level at a location.
Going by years of experience, my level is Staff (IV) and the description notes that this is 10% of all engineers. Looking at the LA bar chart, the 90th percentile is around $325k total compensation. So a Staff level engineer in the LA area should expect around this much. Does this analysis seem correct with the given data?
The interview should be all about READING code. The original reason for writing code during an interview was to filter those who can't code. And so problems like fizzbuzz were created and that expanded into leetcode. But does anyone really think that if you ask someone to read a moderately complex piece of code and they can fully explain what it's doing, they can't write a simple for loop? If you can read code, then you can write code. And reading code during an interview if far less stressful than writing. So you will get much higher signal from the interviews since the stress of writing code during an interview causes many good engineers to do poorly. Or never try at all since they don't have time to practice to the point where the stress doesn't matter because it's rote memory.
Reading code is also better at showing the difference between a junior and senior engineer. If a junior or senior crams leetcode and spits it out during an interview, the result will look the same and you really don't know if they are junior or senior. But reading code is where years of experience can really shine.
And I want to stress again, reading code is so important, I think more so than writing. If you're only good at writing, then you can churn out tech debt filled junk quickly. If you read well, you can see where to add a few lines to get the same result. If you only write well, you copy/paste stackoverflow answers without really knowing what's going on. If you read well, the stackoverflow is simply a starting point you modify or discard entirely because you can recognize junk. Or even that the top rated answer is actually not the best one.