Hacker News Reader

Comment by sokoloff | Hacker News Reader

I live in Cambridge (across the river from and on the same subway system as Boston). In theory Amtrak should be able to attract my business for trips to NYC. Prior to COVID, about twice as often I flew to NYC as took rail. The times I took rail, the train literally never arrived on time at either end of the journey. We’d sit for an hour waiting for some track or switch problem to be fixed. Or we’d sit for a few 15 minute stretches for no apparent or announced reason.

If I could count on the arrival time, I could use the train for business meetings near the train station. Since I can’t, I fly in and Uber over which, even if it takes longer per the published timetable, has less tail-risk of being late.

No one needs hourly rail service between city pairs (as the article implicitly calls for) while the variability is -10 to +150 minutes.

Looking for a Nov 15-19 trip to NYC, Amtrak is $306 on Acela and I could plan to start working around lunch on Monday and work a full day on Friday if I was willing to arrive home around midnight. On Delta, waking up around the same time, I could reliably attend a 9AM meeting on Monday and working the same full day Friday, get home in time to say good night to my kids. That would cost $157 plus two Ubers. That’s with spotting Amtrak the advantage of a meeting walking distance from Penn Station and riding on the crown jewel of the Amtrak network: Northeast corridor Acela.

View on HN

A few years back, I commuted almost every week between Boston and NYC on Amtrak (~80 trips total) and had a very different experience.

I took the NE Regional instead of Acela because it was $49 each way and only ~30 minutes longer than Acela. With the exception of blizzards (which wreaked havoc on everything), there were only a handful of times the train didn't arrive within 30 minutes of schedule (a delay of 10-20 minutes was pretty normal - but I just factored that into my schedule).

In that period, I flew the same trip maybe 5 or 6 times, and that was more unpredictable. Once, my flight was delayed by over an hour. Another time, I spent almost two hours in a Lyft getting from JFK into Manhattan due to construction.

Plus, there was the stress of trying to get to the airport on time on both ends. And worrying about whether there'd be a line at security, and trying to figure out what I'd do if I missed my flight. On the other hand getting to South Station and Penn Station was easy, I could walk right onto the train, and if I thought I was going to miss a train, I could just switch my ticket to the next one.

> No one needs hourly rail service between city pairs (as the article implicitly calls for) while the variability is -10 to +150 minutes.

While there's always a risk for delays (I've been stuck at Heathrow for 3 hours before thinking "If I'd have taken the train I'd be there by now"), you should be aiming for a network at 95%+ of services arriving within a few minutes of the arrival time. I understand that US passenger trains don't have priority, and thus this isn't workable.

On that assumption, given the distance and size of Boston/NY/Philly/Washington, I'd expect at least a 30 minute service, taking well under 4 hours (should really be under 3) from Boston to Washington (with stops in NY, and maybe Phillidelphia). It's only 400 miles, that's like Paris-Marseille, and far shorted than a 4 hour 560 mile Tokyo - Hiroshima trip)

>I understand that US passenger trains don't have priority, and thus this isn't workable.

This is largely true, but isn't on the NEC. It's one of the few corridors that Amtrak actually owns.

Mostly. They don’t own the stretch through Connecticut for example, and the train will come to full stops between stations as a result.

This stretch is owned by another rail network (ConnDot) that prioritizes other passenger trains (Metro North and Shoreline East)

There's a large advantage to passengers having a nationally organized network - same as with air traffic control (provided by government), or highways (provided by government). Of course track segregation between non stop services and local services would be better.

Comparing any rail service with France or Japan is futile. Nobody compares with them, least of all the US. For Tokyo to Hiroshima you have multiple departure times every day, and your choices of speed at each time include slow, reasonable, fast, really fast, and really, really fast. It's rather breathtaking, even if you're from a country with good train service.

The UK has 3 trains an hour between London and Manchester (160 miles, about the same as Boston-New York), takes 2 hours, running on victorian infrastructure and a mixed mode railway,

In Germany, Berlin to Hannover, 180 miles, multiple trains per hour, under 2 hours. Many similar services out of say Frankfurt or Cologne too.

Italy, Florence-Rome, about 170 miles, 2 trains an hour taking 90 minutes.

For longer trips -- 2 trains per hour from London to Edinburgh (390 miles) taking just over 4 hours, again based on 150 year old railway, reliable clockface services.

Madrid to Barcelona, 400 miles, hourly or better taking 2.5 hours

OK, France, Japan, Korea, China, Italy, Spain have the high speed services, but most of the western world - certainly areas as dense as the US's north east corridor - has this reliability, frequency and speed on the c. 200 mile trips like Boston-New York or New York - Washington.

(There's probably more argument for a 3 train per hour Boston-Washington multiple-stop service taking about 4 hours, rather than a dedicated high speed service which only stops in New York, but the point is that the US is so far detached from the western world it's unbelivable)

I am the opposite of you. I live in NYC and my parents in Boston, and I take the train all the time.

Flying is very unpleasant, as you still have to deal with getting in and out of airports, the whole stay in the TSA line, remove your shoes and belt thing. You are completely omitting the security crap you have to deal with air travel. Meanwhile the train trip starts and ends in downtown. (I live in manhattan and can just walk to the train station).

Also, just the taxi fare from JFK to Manhattan is about 70+ tips. Sure, you can take the subway ($8), but that will add another 1hr to your trip.

Doesn't really strike me as the opposite. If the GP post is talking about being on time for specific meetings and you are talking about visiting family, you're optimizing for different objectives.

Personally, I use the train vs fly to NY from BOS depending on my objective.

your air travel could be significantly optimized with TSA Precheck ($20year) and Clear ($120/year) to effectively eliminate security headaches

And into what little database are my deets being entered for eternity by using either of these options, and to whom does the visibility to this data go?

Presumably the same database your details go in anyway, just with less security theatre.

Yeah, we can’t make it look like we’re relaxing security, but if we charge you an application, sit on it, and make felons lives worse, that’s cool.

Looking over the felonies for which TSA PreCheck is unavailable, it doesn’t appear to a thoughtlessly compiled make-felons-suffer list: https://www.tsa.gov/disqualifying-offenses-factors

I can see it for Part C, but why A and B? If someone's been charged and done whatever was required to be free enough that they can fly, then that should be the end of it. Especially things on that list. If treason isn't enough for the government to execute them or lock them up and throw away the key, then it shouldn't matter to the TSA.

If you can’t see how felony conviction for having illegal explosives might reasonably preclude you from reduced security screening, I’m not sure anything I can say will convince you.

If you can't see how the government continuing to treat someone like a criminal after they've paid their debt to society is wrong, I'm not sure anything I can say will convince you.

Are they being treated like a criminal in this case? Or are they merely being treated as requiring standard (not reduced) security screening?

I’m far more concerned where felons are denied the right to vote or to bear arms than I am with them having to stand in the same security line as most people.

I'm against the TSA and its security theater on principle. Any gradient to the offense of its existence just adds insult to injury.

So you should have a situation where someone who hasn't been caught with illegal explosives can pay some money and take those explosives on a plane?

That sounds sensible.

Your deets are going in a little database every time you buy a train ticket.

Can you not buy anonymous tickets, for example at a counter in the station or from a machine?

There are still enough machines that take cash that this is a practical option in most of Europe. Buying a ticket in advance works, but requires a trip to the station.

Getting to/from the airport is the much larger headache than security in a lot of places, especially in NYC. Even if you're willing to pay $50-$100 each way for a car, it's not always easy to get a car and it's a long drive. It takes almost no time for Manhattanites to get to a train and costs virtually nothing.

Why would you pay extra and deal with the additional headache of being screened/having to enter your private data if you could just take the more convenient and largely anonymous train for under a $100? A train where you are not squeezed into the seat like a sardine?

I am personally also really uncomfortable with the notion that you have to pay an additional convenience fee just to have a mildly more pleasant travel experience. Coach class should not mean that you are cattle.

People in coach can still use PreCheck. Work only pays for economy tickets, so I can tell you PreCheck still works fine.

Also, FYI there are two price levels for TSA screening. One is the TSA PreCheck, the other is the 9/11 security fee everyone has to pay. A universal PreCheck would probably mean increasing the security fee.

Clear is one step too far to paying for freedom. Global entry money at least goes to one gang I already pay protection money to, the government. Clear is a non government gang that somehow got the ability to get paid in exchange for giving people more freedom.

That's how America has always worked, though: freedom is an asset you buy with money. They're just saying the quiet part out loud now.

When you buy it from the government, at least there is a veneer of it having come from democratically elected leaders and going back to a pot of resources that are democratically spent.

As in there was zero reason for the government to give Clear a cut of the money, other than outright corruption. They already have the department of homeland security.

I remember one time, before Uber existed, I was flying to LAX from SFO.

The cab from my apartment on Upper Market, in SF to SFO was $65 before 'tip'. 13 miles.

The FLIGHT from SFO to LAX was $99. https://i.imgur.com/uXCj3MR.png

After this, I had several Black Cars that I made contacts with the drivers - and would call them for rides to SFO (I flew out every sunday morning at 7am - and it was a nightmare getting cabs (I was banned by several taxi companies in SF because I would call three diff taxi services and 2 out of 3 would never show up... but apparently they shared dispatching info because the dispatchers knew that my number was calling multiple taxis... and the guy told me "your banned for calling multiple cabs!" to which I replied "because you never show up and I need to get to the airport (said in a George Costanza voice))

Then Uber was born...

I paid $65.28 (before tip) for a Lyft from downtown to SFO last night (14.27 miles), for a $120 flight to DEN.

I'd like to think that Uber/Lyft changed the industry for the better (and in many ways they did: app is easier, upfront pricing, easier/safer payment, more accountability through location tracking and ratings), but I don't think that the underlying economics have changed all that much (other than the period of time where Sand Hill Road was footing half the bill).

why not take bart, especially anywhere near market? it's roughly $10 to either sfo or oak and pretty reliable.

in LA, the flyaway is affordable and relatively reliable too, if your destination is near downtown, or on the red or orange lines.

my flights were early on sunday mornings.... bart is not an option at my timings.

People from the 1950's would be shocked at today's infrastructure progress in the US. "What!?? You guys are still traveling at the same or worse speeds 70 years later!? What happened?"

Also, it took 2x the cost and 8 years to build an access bridge to Golden Gate bridge than it did to build the bridge itself (3.5 years in 1930s) accounting for inflation.

Our government is so dysfunctional, it is hard to imagine where my taxes are going. I live in Bay Area - No solution to homelessness here, pot holes everywhere, trash on highways that sometimes make me wonder if I am in a developing nation, new construction is a forgotten concept, complete gridlock in the CA gov for issuing permits of any high rises or building new subway systems, security and crime through the roof, etc.

I am extremely against everything about CA government. It is rotten to the deepest core. Anything it does bites dust. The entrepreneur spirit and its natural beauty is what's keeping it relevant. There is no point in voting any of these people out, because the replacement is exactly the same kind of people. Besides the tech industry, people here have no optimism, they're more pro China than pro US, mostly busy destroying this great state.

> There is no point in voting any of these people out, because the replacement is exactly the same kind of people

So why don't you stand?

> So why don't you stand?

I am not an american, so I wonder whether it's realistic? I imagine the US as a 2 party system, so if you're unsatisfied with the performance with one party and unable to vote for the other because of their actual political stance...what are you're left with? In germany we have a spectrum of pasties and therefore i have second choice, a choice to build up pressure against the party I am unsatisfied with even if I feel like I belong to this party. What do you do? Vote for independents? Do they even stand a chance?

> In germany we have a spectrum of pasties and therefore i have second choice

That’s obviously a typo but, in this context, it’s hilarious.

With a two party system you join one of them and you stand for election to be that candidate.

I'm no expert on San Francisco politics, but looking at this

https://en.wikipedia.org/wiki/2018_San_Francisco_mayoral_spe...

suggests the mayor is elected

1) From non partisan choices (Although as it's SF I suspect most will have Democrat leanings)

2) Using an instant runnoff / ranked choice or similar situation

The recent election https://en.wikipedia.org/wiki/2019_San_Francisco_mayoral_ele...

Shows that people seem happy with the current mayor

https://en.wikipedia.org/wiki/2020_San_Francisco_Board_of_Su...

Shows that the "board of supervisors" is an instant runoff too, OK not brilliant, but far better than FPTP. People don't even bother running in many situations -- district 9 for example. 16,000 votes would have won that election.

Now OK if you aren't eligible to stand for election that can be a problem, but surely you find people with similar views than you and build a campaign (if you can't find them, then your complaints really are fringe).

Complaining about politicians all being the same is easy. Doing something about it isn't terribly hard though - not at a local level, if there's really a problem. Personal 100,000 SF residents to vote for you and you're in as mayor. Persuade 20,000 and you're a council member. Of course once you have pwoer you then need to fix the problems, and that's often far harder than posting a rant online.

The election that you picked lacked any credible candidates outside of London Breed: the rest had spent their shot during the special election after Ed Lee's death.

In the previous election, she controversially was given the role of acting mayor and held it through the election, giving her a huge boost of energy from people who were willing to just let her stay since she was already there.

> Persuade 20,000 and you're a council member.

Well, it's slightly trickier than that. Supervisor elections are district-based, so you need to capture a specific subset of San Francisco's voters. (And we call them Supervisors rather than council members as they serve as both city and county representatives for the combined city/county structure of San Francisco.)

Sure, the point is that if you aren't standing and your view isn't represented closely enough, you can't really complain there's no choice. I'm sure there are plenty of YIMBY software engineers who could find someone charismatic enough to get into local politics and push for those policies - certainly in the UK it's not exactly hard or expensive to run for local council, print some flyers, run some facbook ads, etc.

But it's easier to just say "they're all useless", or maybe there just isn't an appetite for change

Again, you're really undercutting yourself by trying to make San Francisco your example. The average Supervisor campaign spend is in the high five digits, and in some races over $100,000. The races are contentious, because SF Supervisor is often a stepping stone to a larger state office.

But San Francisco is a very unique system, so if you live literally anywhere else the barrier to credibly run is much lower.

To be fair, 11 people died in the construction of the Golden Gate bridge, and many bridges of the era fell down due to insufficient safety standards. It's not like there are no benefits to the extra time of modern construction

This is the exact defeatism that I am refering to. We had 90 years to invent safe and fast measures to build infrastructure. We have failed and its not because it is technically infeasable.

Out of safe, fast, cheap you can only choose two.

That’s basically saying we made zero progress at all in nearly a century, we just have different values.

This is incomprehensible. Building bridges should be 100x as safe, cost 1/10th, and be done in 1/2 the time, by now.

> Building bridges should be 100x as safe, cost 1/10th, and be done in 1/2 the time, by now.

This seems like the sort of thing that only a software engineer would say. So please, tell me that's not your job.

Mate, can we start by honestly and openly criticizing the current situation? If we're not even able to criticize it, reform feels like a distant fantasy. 90 years a very long time. We invented transistors to deep learning. Computer-tech has seen tremendous innovation. Physical world is completely in a death grip of regulations. I am actually fine with regulations where they make sense.

But, it all starts by not throwing rocks at each other and name calling. It starts with deep introspection of our policies. Try asking your fellow neighbor if they can criticize their own party - you'll be met with blank stares and profound confusion.

Why would bridges EVER be 100x safer? Twice as safe would be a huge transformation for any mechanical engineering project. 100x is the sort of process that happens over centuries, and is probably asymptotic (there's a limit to how safe any physical object in the world can ever be).

Cost 1/10th? Why? The materials haven't become massively more available (or changed significantly). The technology and therefore the labor involved in building large scale civil infrastructure hasn't changed much in at least a century, which you seem to view with alarm rather than the idea that maybe we're pretty good at it. Costs due to regulations represent the growing importance of values other than "get it done as cheaply and quickly as possible".

Take 1/2 the time? Why? The physical world isn't amenable to the same sort of reconceptualization that symbolic computation allows. Although the numbers and materials have changed over time (e.g. the arrival of the iron bridge, or suspension systems), the changes generally involve trading off mass (e.g. compared to stone bridges) versus complexity and precision. Why would anyone expect this to reduce construction time by 50%?

And yes, software has seen incredible strides forward. At the same time, sending someone a file is still an enormous challenge for most computer users, most software has failed to incorporate at least 50% of what we've learned about UX over the last 40 years, and most of it is still riddled with bugs that are not that hard to find. So we have to be a little careful throwing stones at those who build material objects.

Going back to gp's question, though...

Can you openly and honestly criticize the current situation?

Not defend, not provide explanations for, but openly address and describe problems that exist?

I think there's huge room for criticism. But my starting point would not "it should be 100x safer, 1/10th the price, 1/2 the time".

There have been some excellent links here on HN over the last year or two (at least) about difficulties in the USA in particular with civil engineering projects. I would defer to their much-more informed authors (and the often great comment threads here) regarding these matters. Clearly, things are "not right" with the time and cost of those projects in the USA relative to elsewhere in the world.

If we were discussing "world best" rather than US-specific, then I'm not so sure if there's the same need for criticism.

If you see the multiple orders of magnitude reduction in cost to launch satellites we've seen recently, as a result of reusing landable rocket stages, how can you been so bearish on physical engineering improvements?

Even computing advancements are technically a physical engineering triumph, taking the form of increased precision in chip creation.

we haven't been putting things in space for very long.

we've been building bridges, roads, buildings and more for thousands of years. from time to time, there will be "breakthrough" developments that transform the processes involved in these activities, but at this point they will be rare, certainly compared to novel human activity like near-atomic circuit fabrication.

To clarify I meant 100x as safe for the workers.

I usually took LIRR from JFK when I visited NYC, which is faster and more comfortable than the subway and often faster than Uber/taxi. Penn Station is convenient for most of Manhattan, and Atlantic Terminal is good enough for the parts of Brooklyn I typically stayed at.

A bit more expensive than the subway, especially during peak hours, but worth it IMO.

> $70 plus tips

LIRR into Penn. $16 air train + peak direction ticket (you can buy the latter on your phone) and much faster (20 minutes after you catch it at Jamaica). Less annoying going into Penn than going out, at least.

Your experience in flying between NYC and Boston on a Friday afternoon/evening flight must be really different from mine. I’ve spent hours in one or the other airport at times waiting for god knows what “equipment” to become available.

The other side of the coin is that I find I can get a lot of work (or book reading, or movie watching) done with a beer in my hand on the train.

Obviously different people have different preferences, but I wonder sometimes if our preconceived preferences affect our perception of the real events we’ve experienced - that includes my own to be sure.

> riding on the crown jewel of the Amtrak network: Northeast corridor Acela.

It is worth pointing out that it's unanimously agreed that the worst part of the NEC is the Connecticut segment, especially New Haven-New Rochelle, as Metro North is responsible for the dispatching there and they do not do Amtrak any favors. Every time I'm riding Amtrak on that segment, it feels like the train is stuck behind some pokey local commuter rail rather than expressing around it, and this is not something I ever felt on the MARC, SEPTA, NJT, or MBTA portions of the line.

It'd be great if Amtrack could improve its only profitable line instead of using it subsidize all the ones that lose money.

Completely agreed. The only thing I will add is that it’s not only it’s only profitable line, but equally importantly, it’s only useful line.

Amtrak, as the original article points out, is simply not useful in much of the rest of the country.

>That would cost $157 plus two Ubers

Depending on market, those two Ubers could nearly double the ticket price. I just checked my app, and a ride to the airport is $53. It won't be any cheaper coming home. Just a reminder that you can't really just write off the cost of the Ubers. nevermind the hassle of just dealing with airports

GP does have $149 worth of wiggle room to fit those two Ubers in and have their point still work, so I think not trying to exactly estimate their prices was fine.

Yes, I can do the arithmetic. $300 for a train ticket or $150 + $50+ + $50+ is only ~$50 off. Plus, the hassle of TSA x2 airports twice. My point is that, all things considered, the price difference really isn't that great and can be even smaller depending on the flex pricing of Uber. I wasn't trying to quibble over a couple of dollars.

But their argument also priced in such massive time savings on the flying route that even an equal price is clearly in the flight's favor.

Interesting, my experience for NYC-DC is completely different. It's pretty much always on time, at most 15-20 minutes late, at least if you buy NE Regional or Acela. I do avoid buying northbound tickets on long-distance routes that originate way off (e.g. the Crescent from New Orleans, or Palmetto from Savannah), since they can get delayed way before they get to DC. I've been taking it every other week or so since August. I didn't realize NYC-Boston was that much worse. I also buy my tickets way in advance (my schedule is predictable, and Amtrak is offering free changes currently anyway) so usually pay around $30 for the NE Regional or $70 for the Acela each way, which is an amazing deal compared to flying (or even driving).

My experience was similar to yours. It's been a few years but I used to commute weekly from DC to NY. Sometimes I'd take the Acela and sometimes the regional, depending on price. I never had an issue either way. My biggest inconvenience was the shitty waiting area and mob of people at Penn Station, but even that was better than the TSA lines at the airport (or the dreaded Dulles people mover shudder).

I just looked for an approximately similar trip in terms of distance and dates in Germany. On Monday morning Nov 15, you could take a train from Munich to Heidelberg starting at 23.90 euros (one way), while the first-class ticket (one way) appears to be 33.90 euros.

Why are Amtrak tickets so expensive?

I suspect largely because the ridership is so low. There were 3.5 million total Acela passengers in 2019 (chosen to avoid COVID effects on volume).

Acela is only part of the Amtrak service offering, but it’s the part they tout as being the best Amtrak has to offer. Total Amtrak volume was around 32.5 million trips in 2019 (unclear if a round trip is 1 or 2 trips, but I assume 2).

In any case, that’s slightly under 1 train trip (however they are counting it) per 10 US residents per year.

From quick googling, the German comparables seem to be 151 million long-distance train trips for around 83 million people, or around 5x the total usage and 20x the per-capita usage.

Amtrak charges the market clearing price. There is quite frankly a fuckton of demand in the Northeast corridor with its 52 million residents along the one line, and not a whole lot of rail capacity due to the fact that Amtrak barely has money to keep the lights on and the NEC is 100+ years old and in need of significant capacity upgrades. (There are only two tracks connecting New York to the entire south and west of the country, for example.)

Even at the very high Amtrak prices the trains run full.

If you converted Euros to USD based on today's exchange rate, that's

23.90 EUR = 27.61 USD

33.90 EUR = 39.16 USD

As an american that's the lowest I've seen any kind of train ticket regardless of distance.

Chicago to Milwaukee is ~$25, though that's only 90 miles.

20 bucks from suburbs to boston round trip. ~18 miles one way.

I regularly travel DC-NYC for $29 on the Northeast Regional Amtrak train (225 miles, about 3.5 hrs). You do have to buy about a month in advance to get those prices though.

That's slightly over the distance of PDX-SEA on the west coast, which is currently available for $28-47 depending on capacity. The equivalent bus trip (Greyhound) costs $22.

The vast majority of people just drive instead. People doing anything except driving is unusual.

I found European train tickets between popular/useful destinations to be just as expensive as airline tickets. And, generally, airline tickets were cheaper if you could put some advance purchase days into your schedule.

When we were touring Europe (pre-Covid), we never had a city-to-city transition where the trains were significantly cheaper than air travel.

I suspect this is a function of the popularity of the cities being connected.

The solution seems quite obvious: a big tax on flights, reinvested in rail infrastructure. At some point we have to help the market find lower carbon solutions, what better way than replacing short haul flights with trains?

If you want to tax net carbon emissions, I’m all ears and, unless you work very hard to come up with an intentionally terrible plan, I’m very supportive.

If you instead want to tax specific conveniences to help inconvenient services survive despite their inconvenience and lack of improvement, well, let’s just say that I’m significantly less receptive.

Most of the major issues with Amtrak in the US come down to not actually having control over the track that's in use, and that's never going to be fixed without a real big money infusion.

I want to tax something that is very bad for the environment and fund something that is very good.

The free market is great for many things, but not for core infrastructure. What’s missing in the US is a basic passenger rail infrastructure. Once you’ve built that, fine, let operators work in a free market environment

If you’re still talking about the environment at the end of your first sentence, rail is not “very good”. At best it’s “less bad” for the environment; to that end, it seems like a broad carbon emission tax provides a better balance (and is easier to ramp up on a schedule needed to effect change) than a quilt work of individually targeted taxes against things you don’t value.

Everyone emitting ancient carbon is a problem; tax it all.

Rail is amazing for the environment. A Marseille to Paris on the TGV emits less than 2kg of CO2 per passenger. That’s 80 times less than a plane, and 20 times less than a car!

And how is this CO2 miracle achieved? Public spending on core infrastructure, namely nuclear power plants and high speed rail

We will need to tax CO2 to oblivion, but before we do so we need to make sure that people can still move around, that’s key to a healthy economy. Public transport is the only viable option for cheap mass low-CO2 transport

surprisingly flying is not that much worse per passenger mile, really depends on occupancy. A full flight can be just as efficient as a half empty train. (EDIT: consider how many passenger miles you get from one flight from NYC to LA, vs running two 1000hp diesel engines for 4 days for a similar number of passengers)

pollution is a different story, the new amtrak fleet is very clean burning, and doesn’t dump jet fuel into the ocean, so that’s a plus.

Where are your figures from? For electrified lines like the NEC emissions will depend a lot on the grid’s generation mix. For various routes from London the train emits 10-15% per passenger of an equivalent flight: https://www.seat61.com/CO2flights.htm

You're right, electrified lines are very good. The majority of Amtrak's routes are diesel.

I don't keep notes on where I read what, but this BBC article is fairly thorough, and puts electric trains at 40 grams carbon per passenger mile, diesel trains at 90g, long haul flights at 102g, and short flights at 133g. It mentions that occupancy is a factor but doesn't say if those numbers assume full occupancy or are averaged out over actual ridership.

https://www.bbc.com/news/science-environment-49349566

That's a typical BBC mess of statistics. "Some" diesel trains as as high as 90g, but you haven't quoted the "some" electric trains being as low as 6g.

Routes but not passenger trips. The NEC has 260M passenger rail trips per year. Amtrak has 32M on its entire system 17M of which are on the electrified NEC. The diesel routes compete with cruise ships. They’re lovely but they’re not a transportation solution.

I guess it depends where the power comes from. The French TGV is close to 0 Co2

I do not know if this was also true in USA, but in many European countries, more than one hundred years ago, but still also when I was a child, in the rural areas people used to set their clocks at the correct hour when they heard the train passing.

For some reason, unlike in the old times and despite the advanced technology that we have now, wide variations in the train arrival times have become common in many countries, even if not so extreme as in USA. (Even in Germany, which in the past had a reputation for punctuality.)

Perhaps the train companies have become too greedy and they no longer allocate the appropriate funds for maintenance, like in the old times.

In europe the infrastructure is run far more intensively than in the past, a 60 second delay at a station while someone holds the doors has a knock on at a busy junction which leads to 20 minute delays 100 miles away for the next few hours, just like someone cutting in on an interstate can cause a brake wave causing delays an hour later 20 miles away.

This intensiveness is to transport the largest number of people possible.

I like the idea of the BOS-NYC Amtrak. When there aren't delays it's fantastic. More pleasant trip than flying, no security theater, and I'm right in the middle of the city when I arrive.

But to your point, more than half the time I take that trip there's some delay on the rails, usually in the CT corridor.

The idea is fantastic. The reality is not.

The reality is so bad that people would rather ride the Fung Wa bus.

Oh man! I remember riding fung wa over a decade ago. Wow, time flies.

I have some "fond" memories of the Fung Wah circa 2002-2005. I use to take it roundtrip about once a week during that time period. There was the time that we broke down on the Triborough Bridge and were stuck for hours.

Or the time that a woman managed to lock herself in the bathroom without anyone else knowing. And there were no lights on in there. And it was night. So she spent a few hours screaming for help in a stinky bus bathroom, with no help to be found.

they’re also speed limited in that CT area because it’s a residential area iirc, so it’s not even full speed service

Well the commuter trains don't run at the Acela speeds, and even if they did they stop and go more often. If you were to run the Acelas even faster that would mostly have the effect of bumping into the commuter trains in front more often. The only way to do this would be to run less commuter trains, but those trains are fairly busy, and also run by the actual owner of the tracks.

Amtrak's various unfunded fantasies for upgrading the corridor have usually included a dedicated pair of tracks for Amtrak for a reason. (I say fantasy because the proposals are always in the $115B+ range and every time I blink the number is probably going up, and that is an order of magnitude more money than Amtrak has ever had in its life.)

"Not having a system where trains can pass each other usefully" is, from a european perspective, exceedingly strange.

There are four tracks, but they are shared by local commuter services, rapid commuter services, intercity regional services, and the actual high speed trains.

To put this in context, by the time suburban development started in earnest in the US, the private railroads were in free-fall after the war. It took until the '70s for Amtrak to put the (passenger) divisions of said railroads out of their misery, and Nixon did not anticipate Amtrak surviving the decade; Amtrak was supposed to decline out of existence. It never quite did that, but it also never quite got back on its feet, and so now the passenger rails are quite lacking.

In London on the main lines out of the main termini it's usually 4 tracks abreast with commuter services + stopping services taking one pair of tracks and express + intercity services taking the other pair for in and outbound services. High speed railways (greater than 250 km/h (155 mph)) are a different story usually running on bespoke track.

Is that not the case in the Northeast Corridor?

There are two tracks leading from Manhattan to the west and south of the country. These tracks are shared between both commuter trains from the New Jersey suburbs (population 4.5 million) and Amtrak. Elsewhere, there are still two tracks in places.

The particular problem segment being discussed here does have four tracks, but it has commuter, regional, and intercity services. This is causing congestion problems; IIRC UK HS2 is being built because of a similar problem where they need to get the intercity trains out of the way of fast regional services.

The New Haven-New Rochelle stretch of track is a 4 track corridor. You'd run the intercity trains on the inner tracks and the regional trains on the outer tracks. However, I don't think Metro North has ever successfully had all 4 tracks in operation at the same time...

part of it is also that MetroNorth runs a very complicated schedule of zoned stopping (one train A, B, C then express to Manhattan, another train D, E, F then express to Manhattan).

With the amount of padding in the schedule required to run this effectively, it's possible that running an all stops service might actually be faster because it wouldn't require as much coordination.

Yeah they usually have to wait for Metro North CT trains to move on from commuter stations.

I think the Acela is comically more expensive than conventional Amtrak though. They only offer the premium class seats with comical levels of legroom.

I took it once in 2016 (Washington-Baltimore) and it was $45 one way, and the conventional, ACS-64 and Amfleet coaches return trip was $15.

Consider the fact that selling an Acela ticket from DC to Baltimore means Amtrak can’t sell that seat to someone who wants to travel on the popular DC-NYC segment. They price highly to discourage this usage. $15 on the Regional trains is a steal: the MARC commuter train is around $7.

Amtrak milks the North East Corridor to subsidize its woefully expensive and useless service in most of the rest of the country.

There is absolutely no reason anyone should be driving to and from Boston/NYC other than Amtrak being a shit show.

Boston to NYC should be a little more than an hour on real high speed rail. Even assuming you can’t do real high speed rail it should not be more than about 2.5hrs.

And it should be running every hour at most.

Are these shared rails with freight priority? I think that is a key issue with Amtrak service in most of the country. It is a halfway investment that deters further investment due to its poor quality of service but with dedicated infrastructure it would be very different in performance.

And Acela is the useful Amtrak train. Outside the NE corridor, Amtrak is often many hours late. It's really only used by people afraid of flying, railfans (like me), and Amish people. That's just not a big enough customer base to support a train service.

Their service has degraded so much in the last 10 years that even I am considering never riding Amtrak again outside the NE corridor; it's just not fun any more.

Reply on news.ycombinator.com