a1371

Maps tools actually work extraordinarily well in denser cities. They find you the best bus and transit times. They allow you to use mixed mode transportation like bicycles in between while your enjoy the city. Arguably this is a richer experience than a route only accessible with car, for which a paper map would have been just fine. The maps business also thrives where businesses are accessible to people, not where cars are whooshing past them. So I don't think they're enthusiastic about cars either.

There is no heavenly mandate for the US to have this population density. The reason Americans live like this is not that it's the most effective use of land or energy. When we talk about car dependency, it doesn't mean go ride a bus in your suburb, that doesn't work. It means let's abolish out of date and discriminatory zoning laws. Let's introduce density to the city and bypass interstate traffic.

The US issue is deeper than an app. It's the extreme car dependency.

One of the truly amazing things about HN is stuff like this comment. Someone who's actually "been there".

Could you say a little about the motivation behind your LikeBetter?

I respectfully disagree. I found it nice, engaging, and easy to follow.

I'm looking for other studies considering this idea of "life years". Generally around where people's time is going, what is wasting it, or taking it away. Does anyone have any suggestions?

Well, blender itself is a whole world. But you can export to threejs and those are the only two tools you'll need at the very least.

What are energy conserving materials?

My best (frankly uneducated) guess would be that they trained the model with famous movie posters as the objective, and still frames of the movie as the training data. Then they gave frames of these movies to the ai to get a poster out.

Funny how your purchase is fueling their squatting.

I think the context is missed here. It's a useful tool, but should it be an active-by-default feature shipped with the browser? It doesn't add security and doesn't enhance speed. Wasn't it more acceptable if this was an add-on?

This is one of the nice things about this game: that it challenges your sense of time value. If you stay in your base and keep expanding without attacking, the AI exponentially grows and takes over easily. But if you sacrifice resources to destroy more resources from the opponent, it will almost feel easy to play.

As a whole https makes internet better for sure, but what's wrong with the classmate's argument in particular?

My guess would be that the pencils of the same property are on each shelf, and the older stuff are buried under newer ones.

For example, one of the labels says "oily pencils".

So perhaps when you go to him and ask for a purple pencil he pulls out a number of purple ones from the bunch that's relevant to you. Then he explains which one is better for you.

Same heat transfer dynamics apply to cooling as well. Earth is a really good insulator, that's why it has a steady temperature in the first place. If you dump too much heat in it, it will warm up. That's seen as an advantage when you can harvest the same heat in the colder months using geothermal heat pumps.

Some buildings that have access to the ocean or a lake can use that for heat rejection because those take the heat away.

None of these applications require going too deep in the ground. This article is all about going deep.

Iceland has access to very hot water in shallow ground. They are unique that way.

Yes it will matter a lot. Apple is one of the giants standing in the way of right to repair. Being able to say "Apple cofounder is with us" will be a huge convincing point for the general public. The plan is to put the right to repair on the ballot, so public support levels do matter.

The larger lesson might be that most people see a product's status and not its momentum?

How did you learn to read the claims?

It's a pretty big logical jump to say fusion has proliferation problem because it creates a neutron stream. It will be as if they build the Hadron collider to warm up soup. It's unrelated to the plant operation, unlike fission.

I think this highlights the bias in the article you shared. Moreover, it solely focuses on ITER which is not a great example any more.

Fusion doesn't really have the problems of fission. Because fusion has been extremely underfunded, the money is being spent extra carefully on ITER. But MIT SPARC is using new super conductors to get much smaller reactors.

Reactors will produce some low level waste, but once we get a handle of the confinement, that issue can be eliminated with hydrogen isotope mixes. Also, Tokamak is only one fusion design, there may be better ways to capture the neutron stream.

Good comment on testnet. From what I understand, I don't think that's all the plan for the these clusters. Otherwise, why do they specify this?

> When two clusters share a common genesis block, they attempt to converge.

More importantly, the cluster is only 150 validators:

> A Solana cluster is capable of subsecond confirmation for up to 150 nodes with plans to scale up to hundreds of thousands of nodes.

This is not comparable to Mempool where it is in your best interest to broadcast a transaction across the network. Yes, everyone sees the mempool but it doesn't allow you to manipulate transaction order as long as the original sender has paid a high enough fee.

Grab a beverage HN! This one might get a bit long:

There is a saying that "Engineering is finding the best compromises". You know about the Crypto Trilemma: Secure (S), Fast (F), Decentralized (D).

Think of it as you have 15 eggs (I show them with 0) and 3 baskets. For Bitcoin/ETH, it may look something like this:

S: 00000

F: 00000

D: 00000

When we say Fast, think of it as "fast for the transaction fee you spend". So a cheaper network is somewhat considered a faster one too. Bitcoin Cash was not happy with the speed of BTC, so they went with:

S: 00000

F: 0000000

D: 000

But not all these 3 are "perceived" equally. Usually people only see speed, so for a new blockchain to "look" impressive, it can do something like:

S: 0

F: 0000000000000

D: 0

EOS and XRP are examples of this approach. Basically, AWS with extra steps (not very decentralized). For EOS, everyone votes for iirc 12 nodes to do all the transaction processing until the next election.

Let's look at Solana's claim on "Proof Of History". Wondering why other networks did not do an update if it's a great idea. Looking at their dev docs, they introduce this concept of "Cluster", initiated by "Leaders" where "Validators" join them and and "Clients" send transactions to.

> [clusters] simply ignore the existence of the other. Transactions sent to the wrong one are quietly rejected.

So they are breaking down the blockchain into small Validator groups. You have to know which Cluster you want to contact. That's... a compromise. It's not even comparable with ETH or the internet where you have this massive pool of validators that all participate over the same network.

But ok, let's go with the compromise. Then the issues start to emerge. When there are clusters, what happens to the network hash rate? is it divided up? then someone can easily come in and 51% the clusters one by one? Well, they have to work around that. Let's see.

> Additional validators then register with any registered member of the cluster.

Ok, so, if I have a cluster where I know all the Validators, I can collude with them to just ignore any new Validators from coming in?

> A validator receives all entries from the leader and submits votes confirming those entries are valid... Clients send transactions to any validator's Transaction Processing Unit (TPU) port. If the node is in the validator role, it forwards the transaction to the designated leader.

If I'm the leader, can I censor a transaction coming from a Validator? Or perhaps I see a transaction that will move the market, can I sneak in a transaction in my own benefit before broadcasting that one to my validators?

This just kind of a bank.

On a more humane note, I listened to Solana's podcast and I am annoyed that they were dunking on ETH and acted so egotistically "we are new like electric car... people say we are so smart". The blockchain speed/scale limitations are just hard to solve the right way, it's disingenuous to discredit truly brilliant people for your own tech. So maybe I'm dismissive of something that's lying underneath. For what it's worth, Vitalik recently has talked about many of these limitations.

Still, good on Solana for providing legible documentation.

PoH is essentially just PoW: more blocks mean more time has spent. I don't know why they just named it something different. Since SOL is not really decentralized, the validators can communicate a lot faster. So that's the difference with a usual blockchain where the latency is higher.

Hydro has historically been doing that pretty reliably. Most grid markets have access to some of it and the capacity can be conserved during day hours thanks to renewables. Add offshore wind, insulate buildings, and make utilities hand out more LEDs, those might just get us there.

I am not sure if it is true in general. But Saudi Arabia is an example. It is considered a reliable US ally, and in the meantime there are tensions about nuclear power generation there. I'd think a lot of countries fall in that category.

True. As long as we agree nuclear is not a Plan A, and it is not a silver bullet. Otherwise, it may cause hesitancy towards other impactful actions.

I don't think most people do. Well, there is the usual NIMBYism, but the energy industry has figured out how to rally the public, even if it is for something disastrous like fracking.

I think the *idea* that most decision-makers and advocates are against nuclear because of those two reasons (waste & explosion), is holding us back from having a constructive conversation.

Just my 2 cents, I haven't done any academic work on this.

Before anyone comes in to make a strawman argument, I'm gonna say myself why people like me are not fans of nuclear energy (fission).

I have absolutely no concern about nuclear waste or a chernobyl. There is a chance of those, but it will be small harm compared to other fuels.

Here are the big problems:

1. *Timeline:* Carbon reduction is like life savings. Start in the 20s and you will be way ahead. This state of the art demonstration which could not have happened in a better environment, will come online in the 30s. Mass adoption is so far away.

2. *Hesitancy:* Related to 1, some conservation and reduction effort may be halted now because people would think nuclear will fix it.

3. *Proliferation:* This technology cannot go to most places it is needed because there is a risk of atomic bomb making. It can mostly be deployed in China, India, and the US, which is huge, but won't address energy poverty.

These are the top 3 for me. Still a welcome news and a good _Plan D_ to get us from 2°C closer to 1.5°C.

The search too. Wikipedia articles and useful covid graphs. Excellent context inference. The assistant is surprisingly powerful when I ask engineering questions (what's the coefficient for this when that).

You're comparing Elixir & Ruby with Babel, not vanilla JavaScript. You are putting your code on non-standard JavaScript, that's on you.

So they are filing a document with SEC, it has a"Risks" section. They mention as a risk that the full self-driving might never work. I'm surprised if they had never mentioned that risk before, but the article seems to suggest that.