AlotOfReading

The stdlib's use of unsafe is the process of formal verification [0]. Unsafe usage is still a decent rule of thumb for dependencies you should pay more attention to.

[0] https://github.com/model-checking/verify-rust-std

Microsoft has written more application development frameworks than you can shake a stick at. They've also failed to gain traction with virtually all of them, even internally.

You can have floats with bounded absolute error, and floats are a universal standard that makes all the projection calculations much easier. Speaking of projections, the error between lat/long coordinates and the physical surface of the earth they correspond to is also non-uniform. Looks like you can get a bit under 1m resolution with absolute error in single precision if I did the math right.

There are both descriptive and normative uses of gender. To use a less charged example, it's not prescriptive to identify as American. It's not prescriptive to say other people identify you that way, even if your passport says Canadian.

An example of using the category normatively would be saying someone isn't American because they burn the flag. My experience is that most of the people using "gender" normatively don't differentiate it from sex.

If we interpret the grandparent charitably, they're saying the centuries-long history of people suggesting Africans couldn't have built the pyramids to justify their own racial bias [0] also applies to modern people. It doesn't seem like an outrageous position, nor does it preclude ignorance being involved.

[0] https://www.smithsonianmag.com/history/how-white-southerners...

It's not possible to put a 5c nano on a keychain. They're intended to be kept in the slot at all times.

I used to keep a yubikey in a spare slot on my laptop. One day it fell out and subsequently escaped through an unnoticed hole in my backpack.

I've never lost a password because my backpack was overly abused.

That's a bad take. For one, illegal immigrants are a minority of the foreign born population. Secondly, even if the statistics are by some miracle not tortured out of the numbers, still obviously not a sufficient explanation. San Francisco for example has a small undocumented population (primarily from Asia), yet also has huge affordability issues from high rental demand and income disparity. There's no world in which Mexican fieldworkers are competing with FAANG developers for $4k/mo apartments in SoMa.

I have a bush of pink blueberries, and I greatly prefer them to the "blue" ones I've also planted. It's a novelty for guests and a nice seasonal treat. Minmaxing the nutritional content of every bit of fruit doesn't seem like a fun way to live.

Thesis by publication is only one way, and not even the most common in many fields. I can't access the actual text of this thesis, but the abstract sounds more like a monograph and I don't see any author publications before the thesis that would lead me to think otherwise.

That's assuming the game initialization order is deterministic. Using the hash of the combined state of seed and string avoids that assumption without giving up determinism.

"Writing" is a tricky term. Indigenous groups in what's now the US had property records, laws, and symbolically represented stories that could be read by others. What they didn't have was a system of symbols that can fully encode human speech (and vice versa). The latter is the typical definition of "writing" and it's not required to have the former.

On an unrelated note, Gwynne's book is fine as a fantasy story, but it's very badly regarded from the perspective of narrative history. Hämäläinen's Comanche Empire is a much better book arguing a largely similar position. Don't take that as applying to later books by the same author, sadly.

    ...then shuffled that array 115792089237316195423570985008687907853269984665640564039457584007913129639936 (2^256) times uniquely, each time recording the resulting shuffled array.

Correct as an analogy for block ciphers, but note that there are (2^256)! unique permutations of the input space. You're selecting an unimaginably small slice of possible keyed pseudorandom permutations.

The higher Himalayas are largely unpeopled as well, especially above 4000m. The only time I met people above those altitudes were at night in camps/settlements. I'm sure the EBC route is more crowded, but that's one very small trek in an enormous mountain range.

    Oftentimes that comes from road taxes, so it's really car use subsidizing the car-centric projects.

The various vehicle taxes haven't fully covered roadway expansion and maintenance for years. The national average is somewhere in the neighborhood of 60% of costs, I believe.

C has had a catalog of undefined behavior since C99, Annex J. There's sample code demonstrating all cases too.

It's worth noting (as the report itself does) that these sorts of lists can only document explicitly undefined behavior, and not the much larger universe of implicitly undefined behavior. So they're not strictly all possible UB, just the most important cases.

It's already a federal crime to lie to the census bureau, it's just not prosecuted for discretionary reasons.

That's what OceanGate of imploding submarine fame was trying to be.

Polyethylene, like they use in food containers. Virtually nothing sticks to it unless specifically designed.

I know it's not what you mean, but Lincoln dramatically shrank the size of the Confederate government.

   It's fairly trivial, but still significantly harder than computing a single CRC.

You can do it with a single GF(2) multiplication, ignoring the complications of reflection and such. A normal CRC is just the special case of making the remainder 0 (again ignoring complications). You can also brute force it, but that's a bit slow for 64 bit CRCs and well, nanoseconds vs minutes in your example.

    Noise is not the only thing they should be protecting against.

Sorry, can you point to the comment where I tried to defend AMD's use of CRCs in this particular application? I think I've made it pretty clear that I don't think they're appropriate for cryptographic applications. I was just talking about the math.

Different tools for different purposes. You probably don't want to be using your mac scheme for noise resistance, because then you're paying a cost in either buffer space, PDU size, or retransmits, and your error correction capabilities are nil. CRCs allow some error correction (albeit rarely used and inefficient for multibit errors vs FECs), good bit error detection properties, and are cheap. It's common to use both at different layers of a protocol stack.

Affine as in substructural linear types. They correspond to linear logic [0], and affine logic is named such because the way it's defined corresponds to affine functions. You don't literally need to scale your pointers though.

[0] https://en.wikipedia.org/wiki/Linear_logic

The entire point of my post was that it's trivial, exactly as difficult as computing the CRC in the first place. Not sure why that was controversial.

Nevertheless, they're still useful protection against noise, and you usually want to detect it right as you're pulling protocol messages off the wire. Placing checksums in the last field of each message (as Ethernet does) simplifies the hardware implementation.

Computing a CRC is equivalent to attacking it. The checksum is the value that produces a certain fixed constant when appended to the data. This is why you'll often see checksums as the last field in a message. It allows for hardware to verify the entire message by checking if the CRC of the bytes equals that fixed constant without having to parse it.

You have to put on some very narrow lenses to argue that "most" serious software is written for the JVM. Operating systems, compilers, browsers, databases, planes, cars, the transaction processing for at least one major payment processor, major cloud services, etc predominantly use other languages.

I don't think there's any single language that most serious software is written in.

The stones don't get tired, but the humans running the visitor center and keeping the tourists in line do. Operating a highly visited historical site like Stonehenge takes significantly more work than people realize.

It's a bit annoying getting frontier models to even work on IOCCC code because of "safety", but even if you get around that LLMs just aren't very good at it. Obfuscating code involves a level of creativity and deviousness that LLMs struggle to meet.

Keep in mind that I primarily consume ebooks, but I generally find books cheaper unless we're counting piracy. You can get physical books from used bookstores for remarkably cheap and ebook/new book prices are kept as close as publishers can get away with.

That said, there are clear advantages to books. You can't page through an ebook nearly as well as a physical reference book. That's admittedly somewhat balanced by the existence of search. Physical books can also pay much more attention to the aesthetics of print and layout. Eink readers and epub/mobi/az3 formats are atrocious for this, whereas iPads with PDFs are somewhat better. There's still works that can't be captured in those formats though, like pop-up books, raised/embossed/textured printing (which I've seen used in poetry), or illuminated works. And books don't need power.

I'm not sure that's a great takeaway. Lots of problems are undecidable and have reasonably effective solutions in practice (e.g. finding bugs -> testing, static analysis, etc). Mind you, I don't expect we'll find anything like that for transformers, but there's a surprisingly large gap between what's possible in general and what's possible in the cases we care about.

No, this is in the same ballpark as ideas like big-O notation. The paper is saying that transformers can recognize a language with exponentially fewer symbols than other kinds of systems, i.e. they're more succinct.

It's exactly as related to real models as computer science is to real computers.