remcob

A book like the "Neuromancer" really needs to be understood in it's time. If you read it now, it will look very clichéd. But that's because it introduced and popularized those very clichés!

A lot of really significant cultural works have become so absorbed in our common culture, and sometimes improved upon, that the original starts to look cliché and even naive. Is there a name for this effect?

7z is a generic lossless compressor. These algorithms tend to specialize in exact repeating patterns in sequences. Good for english prose, but images tend to have non-exact patterns in two dimension. A dedicated algorithm like FLIF will be much better at picking those up. I would expect an order of magnitude difference in performance between 7z and FLIF.

For the curious: the way these non-exact pattern compressors work is by subtracting out the predicted pattern and then storing the remaining signal, which will be smaller numbers and therefore take less bits to store.

As others have explained in this thread: Yes, easily, the projection is that population will top off soon and start declining. It is unlikely (frankly ridiculous) that we will go extinct in the next few decades because of overpopulation.

Also, be careful with the we/them separation. 'They' will lift themselves out of poverty, just like 'we' did.

This is a rude misconception. Family size is mostly related to child mortality rates. In better off places the population growth is neutral or negative and evidence suggests this is causal, so better circumstance lead to _less_ growth, not more. (Source: Hans Rosling's Factfullness)

This only looks at energy consumed during production, not at (long term) availability of raw materials. It seems reasonable to assume that long term the cost of energy and recyclables will go down and that of non-renewables such as oils and ores will go up.

Per capita amounts are more relevant here:

442B in Germany is 5K per capita. 3498B in the US is 11K per capita.

So the spending difference is even more in US's favour.

I generally avoid meat for this and other reasons, but the Kea for sure don't seem to care about animal rights:

> The video confirmed what many scientists had long suspected: that the kea uses its powerful, curved beak and claws to rip through the layer of wool and eat the fat from the back of the animal. Though the bird does not directly kill the sheep, death can result from infections or accidents suffered by animals when trying to escape.

source: https://en.wikipedia.org/wiki/Kea#Sheep

> Each time a node has an internal event, it hashes that event with k hash functions and increments its bloom filter.

Isn't this effectively the same as generating k random indices? In other words, would the bloom filter work equally well if we pick k random indices to update?

In regular bloom filters this won't work because we want identical elements to collide, but here that doesn't appear to be a concern. I can see how we might want identical nodes to collide (i.e. have the same node use the same indices over and over), but that is not what is going on.

> It's a 10-year-old recombination of 20-year-old tooling. Merkle trees and hashcash. That's it.

This is an unfair characterization. Every technology can be described as a recombination of what existed before. Bitcoin solved the distributed consensus problem in a completely new way. This was a long open problem.

> There's no research, no innovation.

There's a ton of blockchain motivated research going on in signature schemes, distributed consensus, formal verification, zero knowledge proofs, verifiable delay functions, game theory, governance, economics, etc.

> For this reason it is safest to break sodium lights under water [...]

This sounds like dangerous advice. Sodium reacts exothermically with water producing NaOH. In practice this means it will violently boil and splatter strongly alkaline lye around. (Source: this is how my high school chem prof lost some of his eyesight.)

It.. doesn't work like that. Suppose someone sells you a private key, or even a hardware wallet, claiming that it "contains much Bitcoin". There's no way for you to be sure that the seller doesn't have a backup. Or is selling ten other copies to other people.

It's like trying to sell an account by selling the password to it. The first thing you do is change it so the previous owner can not get in anymore. Before you did that you can not be sure that you are the only one with access. On Bitcoin the only way to change the password is to move the funds to a new address.

Not necessarily. Because C allows the pointer manipulation, the compiler can in general not make assumptions about pointer aliasing. This prevents some optimizations.

In Rust, the compiler has more information/control over memory layout/lifetime and can therefore make stronger optimizations.

Automatic vectorization is an area where this helps a lot, and raytracing can benefit a lot here. 20% sounds reasonable to me.

Likely they can recover something like `private_key XOR nonce`. If you know a few bits of the nonce, you can reveal bits of the private key. But this is not something you can brute-force.

In other words, the nonce bits leak information about the private key bits. You can't magically create this information by trying all possible values.

Title makes it sound like a bug is discovered. The 'trapdoor' is in the protocol by design. A lot of zero knowledge proof protocols have an initial setup phase in which values are created that need to be forgotten for the system to be secure. These values are sometimes know as 'toxic waste'.

ZCash has a good write up on how this went in their zero-knowledge proof based system: https://z.cash/technology/paramgen

The main criticism by the researchers seems to be that the process around this setup was insufficient to demonstrate security.

> A lot of people now import devices from countries with different frequencies causing a mess.

If they are already using devices that are not compliant with EU regulations, then I don't see how stricter regulations would solve this. This sounds like an enforcement issue. What prevents people from importing the very same non-compliant devices under the proposed regulations?

In fact, I can see more of this import happening as the domestic devices will now be more expensive and crippled.

We have existing regulations stating what is allowed on which frequencies.

Is there evidence that the devices prohibited by the proposal lead to widescale breaking of these regulations? And if so, is there evidence that this causes harm?

It's harsh to assume that people can not be trusted with SDRs and therefore we should ban them outright. I'd like these sorts of blanket bans to come with a strong evidence backed reason.

I am currently in the process of a obtaining a ham radio license specifically so I can experiment and innovate with SDRs and IOTs and be confident that I do so in compliance with the regulations. This proposal will make it much harder for ham radio enthusiasts and innovators to experiment and learn in a compliant way.

^ Above submitted as feedback

Multiplication in decimal systems is inefficient. FFT multiplication works by converting the numbers to and from a more efficient representation where multiplication is O(n) and embarrassingly parallel (convolution vs pointwise multiplication).

If I'm not mistaken addition is also trivial in this representation, and since no other operations are required, the entire computation can be done in the FFT space.

This definitely works in a residue number system, where a chinese remainder transform is used instead of an FFT. (CRT and FFT are algebraically related)

In short, you can create a massive parallel cluster of computers computing parts of the result without interaction, and then in the end combine the results using a single huge pass of FFT.

This is very similar to Montgomery reduction [1]. Both do a multiply taking low bits followed by a multiply taking the high bits to reduce a number.

Montgomery reduction requires an augmented number system with conversions to and from. The article's approach works on numbers directly, but if I understand it correctly the multiplications are twice the width.

Are these two methods algebraically related? Would it be possible to get fastmod working without the double width?

[1]: https://en.wikipedia.org/wiki/Montgomery_modular_multiplicat...

"An app designed for readers." Ironic.

Is there court precedence of a vendor (for example a bank) being held responsible for customer loss due to relying on SMS for 2FA?

The insecurity of SMS needs to have way more awareness and vendors need to feel it in their pockets for this to improve. There are better alternatives like TOTP.

FLIR Camera has many good uses. I bet in a hot or cold climate you can quickly win it back in heating/cooling costs by optimizing insulation.

They are designed for close up work, almost microscope like as you can see in the demo video. I wonder how well they work for a living-room sized scene.

$100k in equipment, but I bet none of that went to an oscilloscope to actually quantify the problems they are imagining.

The graph in the article was made using a 30m dish, which I would not consider DIY territory. The referenced article is a study on what size dish is required with a RTL-SRD. Their conclusion is a minimum of 6m dish is required with simple post processing. They conclude that with improved hardware and better post processing 3.5m dish should be usable.

I'm thinking phased array. It's a more scalable method to getting large apertures and it can constructed to target multiple pulsars at once.

Ever since I heard about the first Pulsar clock in Dansk [1] I've been wondering what it would take to DIY one. Unfortunately, documentation on the Dansk clock is nonexistent, giving me little to start on.

I hope ESA documents and open sources the work. If the required antenna/receiver sensitivity is not too ridiculous, this might make a cool project.

[1] https://en.wikipedia.org/wiki/Pulsar_clock

I wish we could show real bandwidth, latency and drop rate instead. Even better, have industry agree on minimum values for one bar to four bars of connectivity. And gradually upgrade these requirements over time.

All these names just confuse consumers and distract from solving the actual problems. The radio technology stack is mostly an implementation detail.

Gravity on Venus is 0.9 that of Earth, but it's atmosphere on the surface is 90 times as dense. Surface gravity doesn't say much about atmospheric density (the mass of the atmosphere does).

I like to go one step further and use a first pass to convert bytecodes to a list of labels. Use a pointer in this list as the program counter. Then the dispatch becomes as simple as:

    goto *pc++;

Here's an example of the technique for a Brainfuck VM implemented in the Ethereum VM assembly. (a recent competition entry of mine):

https://g.solidity.cc/submissions/wicketh.eth/3bb28842856230...

This technique also makes it easy to do some peephole optimizations and combine multiple bytecodes into a single label. The above example also does a lot of that.

Brainfuck is actually a great language to built a VM for. You have a simple interpreter up in no time, but you can go really far in optimizing it. LC3 looks comparatively messy (more state, more instructions, etc), but more representative of real CPUs.

Look at the two examples where k is used with β being a String or a List. And now combine the two examples, this would force β to be simultaneously String and List -- impossible. A type parameter is not what we want here.

The return type of the continuation is better captured by something like the Bottom type from Haskell:

https://wiki.haskell.org/Bottom

In particular, "Bottom is a member of any type", which is what solves my example above.

They are 100Hz, as I found out while trying to do an optics experiment, but they never bothered me directly. I think it has to do with LEDs going completely dark, where fluorescent just has slight fluctuations. Or maybe it is a 50Hz vs 100Hz flicker thing.