less_less

13 ms·

91.
by less_less 3y ago
I'm not sure it ever got built! The Spectre stuff was found during the "how would we even begin to do this" phase. I've seen a fair amount of academic work about formally verifying RISC-V cores though.
92.
by less_less 3y ago
As a tidbit, this was part of how one of the teams involved in the original Spectre paper found some of the vulnerabilities. Basically the idea was to design a small CPU that could be formally shown to be free of certain timing attacks. I
93.
by less_less 3y ago
I think you'd want it to be a theorem (in Lean, Coq, Isabelle/HOL or whatever) instead of a constraint problem. So it would be more limited by developer effort than by computational power. Theoretically you can do this from softw
94.
by less_less 3y ago
> Yes, the police harass, sometimes I get harassed but that presence is what puts a lid on crime. > I’ll put up with some harassment in exchange for fewer criminals running free. Maybe you have a specific harassment:imprisonment ratio
95.
by less_less 3y ago
> ElGamal/DSA is fragile and difficult to implement securely in a way that doesn't reveal secret k. ECC (and somewhat particularly NIST curves[1]) is also very difficult to implement securely from side channel attacks[2]. IMHO
96.
by less_less 3y ago
> Meanwhile, minimizing the risk of CVD, the leading cause of death worldwide, does seem like a good idea, no? Maybe. I read a bunch of papers on cholesterol and statins a few years ago, and my recollection of the results is roughly as
97.
by less_less 3y ago
Squaring isn't backreferences, though you can square more than a literal. An example squaring is /(foo|bar*){2}/ which is equivalent to /(foo|bar*)(foo|bar*)/, not /(foo|bar*)\g1/. Backreferences aren&#x2
98.
by less_less 3y ago
But the site does allow squaring, and in fact also general exponentiation? Like you can write "fo{2}" to match "foo", where the {2} is squaring.
99.
by less_less 3y ago
Interesting. I think this problem is actually EXPSPACE-complete in general? But still has a straightforward algorithm. https://en.wikipedia.org/wiki/EXPSPACE
100.
by less_less 3y ago
I agree that the US is different from most countries in terms of how difficult it is to prove slander and libel. But I don't think your description is quite accurate. Then again, I'm not a lawyer. > The truth is a perfect def
101.
by less_less 3y ago
I don't agree with this characterization. This is not to say that foundational studies are never invalidated: I just don't think MM was one of them. > Michelson-Morley experiment found no changes in speed of light at all. Noth
102.
by less_less 3y ago
> Even better is to use the helium3-helium3 reaction, which completely annihilate to produce pure energy as the output (welcome to e=mc^2, enjoy your stay). It doesn't completely annihilate to produce pure energy. It produces heliu
103.
by less_less 3y ago
> OK, and? How does this change anything? You have to be a PhD to be a mathematician anyway. I don't understand your objection. Comprehensive math training is useful in areas other than academic research math, and it's not lik
104.
by less_less 3y ago
I agree that there are many aspects of how elite universities work that are not what's best for society. I don't think this is the best thinkpiece on it though. One small thing that jumped out at me is the bit about "we don
105.
by less_less 3y ago
The Golay code and closely related Leech lattice are super cool objects. I was looking into using them for post-quantum crypto, but in my application the Golay code was outperformed by LDPC, and the Leech lattice did slightly better but ma
106.
by less_less 3y ago
Grover's algorithm is not quite as powerful as popularly described. In particular, it (roughly) divides the number of times you need to run AES by the number of times you run it sequentially. So it effectively square-roots the time r
107.
by less_less 3y ago
Property rights are bad. Not in an absolute sense: property rights are obviously very important to how society functions. But "the US needs stronger property rights" is in general a harmful position. US limitations on property r
108.
by less_less 3y ago
Yeah, those vector permute instructions are super useful for both patterns. There are dedicated instructions for some specific permutations (shifting over by a constant number of bytes, and some interleavings) but you can easily end up nee
109.
by less_less 3y ago
You might already be able to get good acceleration with SSSE3 or AVX2 or NEON, which also has a 4-bit-input permutation instruction. The problem is that you're doing parallel lookup into many different tables, whereas NEON/SSSE
110.
by less_less 3y ago
Thanks for the info! It's not really my speed but it was interesting to read through the front page.
111.
by less_less 3y ago
Which instance is that, if you don't mind my asking?
112.
by less_less 3y ago
I don't think that machine was either general-purpose or error-corrected though. IIUC we can build at most a few error-corrected gates right now.
113.
by less_less 3y ago
> I don't think it's the case that finance is over produced. If it were, then the value of financial services would drop. I don't think this follows for all financial services. Overproduction leads to a drop in value if t
114.
by less_less 3y ago
Yes, of course. The quantity and quality of the output can certainly grow for a given resource input -- obviously with physical limits but we haven't generally reached those. It's possible to breed higher-yielding crops, design
115.
by less_less 3y ago
> When someone arbitrages prices back to where they should be, they are performing a service that everyone else benefits from, and are rightly compensated for this. Now, are finance people compensated too much for correcting price discre
116.
by less_less 3y ago
Soundness means that mutable matching doesn't break the type system. It's fine for the behavior to be well-defined and documented but unintuitive. But in a "safe" language, mutation during a pattern match shouldn'
117.
by less_less 3y ago
Maybe that, but also the Nvidia chips have *vastly* higher performance (see https://resources.nvidia.com/en-us-grace-cpu/grace-hopper-su... ). They claim 4 TB/s memory bandwidth and up to 989 single-precision TFlo
118.
by less_less 4y ago
Sure, AES-256 or ChaCha20 is probably an even better choice. I just mean to say that even a 192-bit key is comfortably out of reach of Grover-assisted brute-force for the foreseeable future, which would not be true if Grover effectively ha
119.
by less_less 4y ago
Yeah, this article seems pretty wrong to me. However! I believe that Grover's algorithm will not be a very big deal, at least in early-ish quantum computers. * The speedup is proportional to the depth of the quantum computation, measu
120.
by less_less 4y ago
> If you're talking about Nov 2021 when Apple's Gatekeeper servers went down, apps weren't just delayed, they were unable to be opened _at all_ unless you blocked DNS requests to the server or completely disabled your inte

More ›