krschultz

Is another way of framing this:

"flat road" => assume you can't continue until you prove it's safe

"obstacle detection" => assume you can continue unless there is something blocking the current path

The bit of autonomous vehicle work that I did started from a set of potential trajectories that were then pruned by obstacle detection. So we always assumed you could continue moving forward unless every potential trajectory was pruned. We ignored the cliff problem as beyond our scope.

I see what you're saying, and I think that's why more cargo is moving by rail in Asia. If it was sold as a premium service it might work, but I have no idea how big that market really is when you are talking about crossing whole oceans.

Note that they're not talking about two separate props on separate shafts, they're talking about 2 props rotating in opposite directions on the same shaft. Many ships don't have a 1:1 ratio between propellors and engines anyway, if 2 separate shafts & props were faster they could do that easily. But counterrotating on the same shaft adds more critical moving parts in an already sensitive area.

Someone had to keep us honest! Fun work, I would have stayed longer but I moved to another area for family reasons.

Hydrofoils are generally higher drag at low speed, they only make sense when you want to go faster than displacement speed. That doesn't make sense for most cargo, anything time sensitive is going by air.

As someone that has designed (military) ships, here are my thoughts.

- LNG is definitely happening, and that alone will make a massive difference in emissions.

- No shit smoother bottoms and bulbous bows are better for efficiency, that's been standard practice for decades and large cargo ships are generally not that old.

- Counter rotating props are interesting, I could see that being a big win but the mechanical complexity might not be worth it. Cargo ships are generally kept pretty simple for reliability.

- I doubt we'll see widespread wind or solar power any time soon. The dynamics of wind turbines on top of a pitching platform get pretty wild quick. The energy density of solar doesn't seem high enough.

- Lol, get the fuck out of here with the idea of nuclear power for cargo ships. The US Navy doesn't even bother with that complexity for anything other than subs & carriers where it's the only viable option.

The endpoint that resolves GraphQL requests scales horizontally, so the answer is likely a high number.

I think what you're really getting at it is how many different GraphQL schemas are there? In general I would expect the answer to be one for the whole company, unless the services are truly not related at all. The power of GraphQL comes from being able to compose random sets of data that are not obvious up front.

Aren't they contractually obligated to comply with CTS / VTS in order to get the Google Play stuff? The Chinese OEMs and Amazon Fire line opt-out, but pretty much the rest of the Android ecosystem opts in.

Just rhetorically, if it's the response to your concern every time you write a post, then you need to incorporate why it's not a solution in your original post. I assumed you hadn't heard of Treble and was just about to point it out to you.

"One possibility is that we have a shortage of programmers. If that’s the case, you’d expect more programmers to enter the field, bringing down compensation."

This is true. I was a mechanical engineer professionally. I moved into software engineering, and the increase in compensation was a dominant factor. I know several other non-CS engineers that have made the same move. While it's not driving down compensation per-say, the difference in demand between disciplines is startling.

There's a massive amount of "driving through" Manhattan. People in Queens / Brooklyn / Long Island heading anywhere west of Manhattan can be a cause of through traffic. The entire point of congestion pricing is to setup incentives so that people drive north or south of midtown rather than through it.

It's foolish to count driving to the train station in CT. Any time the traffic is minimal into NYC, the parking at train stations is also empty. Really everything you cited is fairly naive for getting into and out of the city. Parking is ~$40 a trip unless you can get a street spot (which can be really hard to count on in most neighborhoods).

The optimal path is to drive to a train station close to NYC that has frequent service, park and take the train for the last bit. I used to use the Greenwich train station and it worked well. The same is true from New Jersey, you can park at Metropark or get dropped off in Seacaucus.

Source: I lived in CT for 3 years, lived in Manhattan for the last 5 years, bought a car and keep it in Manhattan for the last 6 months.

Both of those are bids, not projects for which the Boring Company has been selected.

Why do you think cost per mile would be cheaper than the subway? With CBTC (existing technology, though not rolled out to most of the subway) and intrusion detection, we don't need any labor on a subway. What's less expensive to maintain, a 10 subway car set that carries 1500 people or 1500 pods?

Maybe paved roads are cheaper to maintain than subway track, but that's hard to really compare apples to apples.

Honestly, I find it easier to imagine a NYC where there are no private vehicles allowed at all during rush hour than I find it possible to imagine one w/o subways.

This fails all 3 points on Scott Galloway's checklist of how to respond to a crisis. https://www.youtube.com/watch?v=PB-AyvgE8Ns

Plenty of people that had no intention to kill someone end up in jail for homicide after they do something that they didn't believe was going to kill the other party. SWATing wouldn't be a threat if there wasn't some (small) chance that someone gets hurt. Kind of like how pushing someone at a bar doesn't always end in death, but when the guy you push hits his head on a barstool on the way down and cracks his skull open, it's homicide.

Hacker News, a place where someone puts together a comment that has more info than your average blog post, and the only reply is a hater.

bogomipz thanks for the recommendations. I've lived in NYC for 5 years a few of those were new to me.

Leaving aside the actual project at hand, this is a great example of a well thought out project plan. There is a clear rationale, clear end state, a bunch of known problems to tackle, a front loading of risk, and it delivers incremental value along the way.

That's a complete misreading of Amazon. Their entire competitive advantage is having really cheap capital which enables unbelievable capital investment. First it was in distribution centers (3PL), then servers (AWS), and next likely be parcel delivery (look out FedEx). The e-commerce business is the first and best customer of each of these services, but they're basically a conglomerate of capital intensive low margin infrastructure services.

Your entire comment is wrong, and toomanybeersies is right.

Source: https://www.defensetech.org/2013/09/27/ohio-class-subs-to-sh...

Reading the raw accident report, I draw the opposite conclusion. The sequence of events that led to the accident started with one person controlling both steering & throttle. The reason for splitting responsibility is that the original helmsman was overwhelmed by the number of actions required to control both.

This reminds me of the Air France crash that was largely due to split control. The UX of transferring control of a system is difficult, and I don't know the details of what it looks like in an Arleigh-Burke.

Though the 2nd helmsman getting confused by having the throttles split points to training. That's ship control 101. I do think the loss of practical training is a root cause here. How many hours does the average helmsman have under their belt? It's just not that much.

What city in the US has a subway system worth bragging about? Philly's subway has 3% of the ridership of NYC's. It's not even in the same ballpark.

I actually think this is the wrong approach. You don't get to reap any benefits until you start eliminating the old system, and with the parallel approach you defer that cleanup period until the transition is complete. It'd be far better if adding CBTC yielded the cost savings of eliminating the old system now as it rolled out. That requires more complex inter-op systems and more expensive upgrades to trains, but both of those things are not the constraint. The core constraint is the time you can shut down the subways for in order to do track & signal work in place. When you have a critical constraint, you should do lots of extra work in other places to maximize the productivity of the constrained operation.

Think through the 3% bit a little bit. Do you think that the banks charge 3% because their settlement software is that much less efficient than a blockchain?

Of course not. You're paying for customer service, marketing, fraud protection & settlement, and a host of other things with that 3%. It's not 3% because that's what it costs to run a computer to figure out payment processing, it's 3% because that's what the banks can charge to run their business. In effect most people pay 2% or less because of credit card points, so really it's just a forced discount on retailers. Again, because the banks can.

Good luck building the "economy of the future" without the US banking system. There is a reason banking sanctions are the US's most powerful non-military diplomatic tool.

Redfin does more to move in that direction than other listing websites such as Trulia. Redfin is the actual broker on your side of the sale, and they take a smaller cut than most other brokers.

The reason it's not cheaper is that few brokers compete on price.

In addition to the obvious political angle, it's nice to see an engineer talking about their duty to the populace. Software engineers hand waive this off because we generally aren't building things that can kill people. We are however building things that have tremendous impact on people's lives, and many in our profession do not seem to take that seriously.

I've seen flights from NY->SFO for less than the cost of a night in a Silicon Valley hotel. I don't think it's optimal though.

The Air Force and Navy definitely crank down their service numbers during "peace" time to invest in weapons systems instead. The lead time on ships & planes is so much longer than sailors & airmen through bootcamp that it makes sense. Plus they stick with a pretty low deployment cadence that could easily be cranked up if necessary.

The big problem is that whole categories of the military budget are waste. They are job programs in the shape of a weapon, but are overkill in our current asymmetric wars and likely useless against an equal adversary. The LCS & F-35 are trillions of dollars in waste.

Even when you are talking about a decent system, the internal level of waste is insane. I was an engineer for nuclear submarines. We were cited as an example for all other programs because we were on time and under budget. There was still a waste in every direction I could see.

I think things would get better with a switch to fixed priced contracts instead of cost plus. It requires the service to nail down their requirements up front, but at least you get the cost incentives right. Until then, the saying "first time half time, second time overtime" was pretty apt.

The northeast corridor is faster than flying. The trains stop in the downtown of each city which cuts out the travel time to and from the airport, and you don't have to arrive at the train station as early before boarding.

The ticket prices are high because Amtrak is subsidizing the non-profitable routes with the North East Corridor. If it was run as a business they would cut all the trains in the middle of the country and re-invest the profit in the NEC, or lower fares. If it was run as a public good, then there would be more of a subsidy. Right now Amtrak in this place where it's restricted by being beholden to politics but not enjoying the benefit of full throated government support.