This is really cool. Is the Rails server production ready? I was always under the impression you had to run it with Uvicorn or similar, although I haven't been following Rails development recently.
Puma is certainly fine as the app server - but normally you'd still have a proxy/load balancer/tls terminator/"ingress server" in front. Something like traefik, nginx, haproxy or caddy.
If (one of the) front-facing servers do regular http caching (a good idea anyway to play nice with rails caching[1])", you can probably "serve" the static assets straight from rails and let your proxy serve them from cache (if you don't have/need a full cdn).
This has been my experience too, and I've never even used Rails at scale. Puma will struggle to serve "lots" (tens on a single page) of (for example) static images (though the actual user experience varies across browsers - safari does okay but firefox chokes and chromium is in the middle). This is with puma on a decent intel prosumer CPU (an i7 iirc) using x-sendfiles through nginx. So puma isn't even actually pushing the image bytes here.
I replaced it with nginx intercepting the image URLs to serve them without even letting puma know and it was instantly zippy. I still use sendfile to support when I'm doing dev and not running nginx in front of it, and I'm not happy with the kind of leaky abstraction there, but damn are the benefits in prod too difficult to ignore.
It feels like this has become common for any older industries that are moving deeper into tech. Even the companies that are bleeding edge seem to have trouble running software teams.
It'll be interesting to see the effects of this since it both impacts very highly specialized labor as well as jobs like contract nannies. I think overall it's a positive direction for worker mobility. The other thing still making it difficult is healthcare.
I can't wait to see this technology improve so I can continue reading my current book of choice while I'm driving. I hate having to repurchase the book in Audible or otherwise and keeping the two in sync.
I dunno, "we've made things worse but someone will come along and make it better" is not exactly confidence inspiring. Especially given the amount of wasted money/resources you will generate along the way.
Not great framing because nobody made electric vehicles worse, they started off at a certain level of performance and then they'll just improve from there.
Now you could say well we made snow plowing worse, and of course in this specific example you'd be right, but you'd be ignoring longer time horizons and not really comparing or accounting for the impact of negative externalities.
An easy way to think about it is if you uninstalled an existing window and had tarp on it for a few days while you installed a more energy efficient window. You wouldn't say "things are worse now!", you'd recognize that you're making a change which is better over the long term.
> An easy way to think about it is if you uninstalled an existing window and had tarp on it for a few days while you installed a more energy efficient window.
A more apt analogy would be if the new window was not even invented yet.
Yeah, I don't understand this pattern of comment-making on the internet. For whatever reason, your comment was downvoted slightly so then you got targeted by people over and over again, even in subsequent replies, with the least generous interpretation of every comment made. I suspect your initial comment also triggered a reaction in saying "it's not a big deal", to which a bunch of people thought, "Oh yeah, well it's a real big deal actually!" despite this being a single article about a very specific situation in NYC, something HN readers will forget about in less than two weeks.
I do not see the point in downvotes on this site. It seems like any slightly political article/discussion results in this sort of behavior.
HN's guidelines are to downvote if something doesn't add to the discussion.
I think in this case it's fine to argue something "is not a big deal" but I think you have got to add more to the discussion than "technology always gets better".
In that case, I'm not sure your comment added much to the discussion either. But the OP generically asked what the solution was, and the generic solution as far as I can tell is the long march of progressively better technology. It's very clear that the tech will continue to improve and is only in its infancy. Had the OP asked something like "what specific improvements can we expect to see in battery weight reduction and charging speeds so that this won't be an issue in the future" I probably wouldn't have replied because I don't know the specifics.
It's fine if you think my comment is low-quality or something, but cherry-picking mine to disagree with and then to go off on an argumentative tangent with poorly formed discussion points isn't really the answer either. It's not a big deal, this is all for fun and everything, but I'm not sure what exactly you were trying to get out of my original comment.
That's not a good analogy because in this case the snow plows still plow so some snow is still removed. To use your analogy then they would have sold existing plow-capable trucks and not had any to plow, but that wasn't the case.
The concern here is that the stopgap measure (the tarp/electric trucks) are implemented too early. The superior replacement window does not yet exist. There is no need to remove the older window and deal with a hole in the side of your house for half a year while you wait for new windows to hit the market.
As an EV owner. I can use my car in about 95% cases: that’s plenty! Way cheaper than gas, no maintenance required, super comfortable, etc. The remaining 5% is when I need to drive somewhere rough or when it’s very cold outside and I have concerns about getting stuck.
There is no way I will give up the convenience of EV for those 5%.
All of the above. Must Run trucks should probably be gas turbine hybrids, that can burn anything for fuel (diesel, biodiesel, Jet A, etc), versus full electric.
>ust Run trucks should probably be gas turbine hybrids, that can burn anything for fuel (diesel, biodiesel, Jet A, etc), versus full electric.
That's just high tech engineer fantasy though.
Turbine is great for steady loads and terrible for cyclical loads which is what a garbage truck does all day. Electrical is great for cyclical but batteries just don't support the energy density to do it all day. You can theoretically bridge the gap real well with a hybrid system but it's only theoretical because in the real world other people's money is not actually an unlimited resource and you're not getting a turbine into anything cheaply. There's a reason you only see them in vehicles that are already fantastically expensive (tanks) and benefit greatly from some of the specific performance attributes. It would be really cool though...
Right now the trucks can do 1/3 of what they need with the batteries they have. Commercial vehicles like this are very much constrained by weight. The "nearly free" and shovel ready solution is to just raise the weight limit for the vehicles in question so they can pack on the other 2/3 of the batteries they need and let them roll around at 120k+ all day like concrete trucks. Sure you'll get a little more wear and tear on stuff but this solution doesn't require an unforeseen technology (battery) or economic (turbines) breakthrough.
> Turbine is great for steady loads and terrible for cyclical loads which is what a garbage truck does all day.
Right. This would certainly apply to the mechanically coupled turbine truck prototypes from the 70's.
But if its driving a generator then you can let batteries or capacitors handle the cyclic loads of acceleration and have a computer throttle the turbine according to the overall demand.
This technology has been talked about lately in the context of Abrams tanks which could potentially be used in Ukraine. Doesn't it burn a lot more fuel?
Yes! But that might be a penalty you’re willing to make versus long turnaround times, especially if you’re in NYC vs a harsh military theater and fueling infra is solved.
The truck doesn’t care if the battery slab is pulled and replaced with another battery or a hybrid powertrain, it just cares it has enough power to accelerate and decelerate.
Swappable batteries is how some countries are migrating to ev for motorcycle/scooter taxis, just swap batteries and keep moving passengers. The cost savings seems to be worth it, read that its about 75% savings against fuel costs.
The easy solutions will go first, working its way up with better tech and design.
I think car/fleet size batteries isn't really feasible with swapping batteries. Small scooter batteries seems like a no brainer.
I don't know about this. I don't think battery swapping scales up very well.
Firstly, I know someone who works on Daimler's electric trucks and he assures me that their electric powertrains are EXTREMELY dangerous just from the amount of power represented. Union operators have expressed negative interest in connecting/disconnecting the batteries outside of the factory.
Secondly, the economics - the battery pack on a Tesla already represents over half of the materials and manufacturing cost. And it's the only scarce thing about EVs. If you are a fleet operator and have to maintain a bank of batteries to charge, you may as well just buy the bodies to go with.
So I would guess that in an EV fleet future where the cost of batteries has dropped significantly, you will probably see the size of the fleets increase as operating costs drop.
> Union operators have expressed negative interest in connecting/disconnecting the batteries outside of the factory.
To be fair, union operators express negative interest in anybody outside their union doing anything that might be construed as otherwise the union's work.
Well, in this case, someone was really badly electrocuted when trying to connect the power leads between the truck battery and motor. So customers apparently "noped" out of wanting them self-serviceable.
> you will probably see the size of the fleets increase
This is already happening with municipal bus and school bus fleets. The bus count is over-provisioned to allow for the longer electric 'refueling' times.
As fleet managers understand, when a vehicle is being refueled it is out of service.
Citibike wants to connect to the city grid so they don't need to send technicians to every station to swap dead batteries. It's not too much of a stretch to think that, perhaps, the stations could recharge truck batteries, too, at some point. Many fewer residents bike during the winter, let alone during a snow storm, so you could, conceivably, sacrifice a bike space or two or five in December-February (or just when you know snow is coming). Stations are ubiquitous by now: especially in the denser areas, there's one every few blocks.
That's assuming you can deal with safety, theft and other related issues.
The other point I forgot to make is that, during a storm, you could ground a lot of other non-essential electric vehicles, e.g. buses, and place their batteries strategically throughout the city (at a bike station, or empty parking spots). Once the emergency is over, you're not stuck with a lot of excess inventory. That, again, assumes a lot: common batteries, etc.
Its already being used at the scooter level, scales fine there for small applications where batteries are as easy to replace as filling the tank of gas.
Solid state batteries show promise. But honestly for a set or requirements like this, just using traditional fuel + paying for carbon sequestration would probably be a cheaper option.
It's ultimately heating the atmosphere above them. I heard this planed had some trouble with thermal balance recently, things were getting too hot.
In seriousness, heating streets with abundant solar energy harvested nearby would be reasonable. But sunshine in winter months is not as intense, and you'd have to build, as usual, a huge battery to keep the energy for the night. (When this is solved, more problems would get solved along the way than just de-snowing streets.)
From what I have heard from my (admittedly few) friends there, it hasn't changed much outside of the pure tech sector. The culture of being in the office hasn't changed as much as it has in the US.
Within the tech sector lots of companies went permanent, partial work from home (need to be in the office a couple of days a week or month depending on the company). Some examples:
- Office downsized to only hold 30% of staff, and you can work from home as little or as much as you want.
- Need to come a few consecutive days per month, travel expenses from anywhere in Japan (within reason) are reimbursed.
I also know a handful of people in non-tech (but not what I would call salaryman/desk jockey jobs) that went to 75% or more work from home.
Some non-pure-tech traditional big corp introduces remote work regardless of covid ending (but usually don't allow living far away). They noticed that having big office in Tokyo is expensive.
How can you proof that you are present after hours, if you are only remote employed? Remote work sounds thoroughly uncompatible with Japanese working culture.
Given the prevalence in the real world of scams and fake ads - is there a danger that hiding them from people just means that they are more likely to fall for them when the inevitable happens.
Any time I hear of a new scam, I tell people about the method so they can look out for it. In the heat of the moment the "IRS tax problem call" or "its me your child I have been in an accident can you send me some money" can be convincing.
> Given the prevalence in the real world of scams and fake ads - is there a danger that hiding them from people just means that they are more likely to fall for them when the inevitable happens.
Strong disagree. Once you've avoided ads for a while, the ads that do slip through seem bizarre and flagrantly manipulative. Ads work best on people who are accustomed to ads, to whom ads are normalized. Once you denormalize ads, they become less effective because their manipulative nature comes across plain as day.
My general advice is "If you didn't initiate the conversation with the party with the intent to take an action or exchange money and they want you to do those things then don't take action or exchange money until you reach back to that party via their official lines of communication or otherwise verify the situation makes sense". In the end it's not the method you use to scam someone that matters it's that you trick them into trusting you really represent something/someone you do not. This also helps with not-exactly-a-scam scams like special offers you didn't seek out but are being sold to you because it gives more time to think if it's really a good deal or think what the downsides would be.
You can give 1,000 examples of how fake urgency is created to make you skip that verification step but past the first one or two to explain the general concept of needing to verify it doesn't matter it just means you'll need 1,001 examples to give the next day.
This isn't bulletproof either but I haven't found anything generally applicable that solves the problem better without making more of a problem than it prevents on average.
The times that one is most exposed is when the circumstances all line up - through luck and the law of large numbers ... you submitted your tax return 2 weeks ago, you did some things you are sure are right on the line, you get the scam tax man phone call - you are primed then to fall for it.
The trust and fake urgency really is the method of the scam (or ad!) - always verify!
> is there a danger that hiding them from people just means that they are more likely to fall for them when the inevitable happens.
Probably, but there's so much variance in human behavior and circumstances that your vulnerability to scams never drops to 0. A call telling you your child is in the hospital and you need to pay them because their check has bounced sounds a lot more convincing during times when you know your child is actually in the hospital.
Exactly, our system for ads and scams works in the same way as the immune system.
Personally I think I have one of those hyperactive immune systems that starts to eat its host alive, having told my actual bank to f off and stop ringing me to try scam me when a payment really didnt go through.
I've found that even paid ads can be challenging -- I imagine these would be best suited for product that target end users and need to sell large quantities.
I'd like to experiment with using both browsers for all my searching for a bit. Has anyone cut over entirely from Google (or others) and remained happy?
I used brave for quite some time and there have been no real downsides, only major difference I really noticed was less google products being sent at me. For example, if I searched for “Translate” I wouldn’t just immediately get google translate.
