hansvm

48 ms·

751.
by hansvm 2y ago
By all means. It's a pattern I use all the time, not just in Rust (often getting away with much less than 32-bit indices). You mentioned this and/or alluded to it, but my core complaints are: 1. You don't have much choice in
752.
by hansvm 2y ago
Rust (most of the time, I'm not arguing about failure modes at all right now, let's pretend it's perfect) drops items when you're done with them. If you use the item longer, you have a longer implicit lifetime. The memor
753.
by hansvm 2y ago
That's embarrassing. Thank you.
754.
by hansvm 2y ago
> Of course they are considered problematic in Rust. > And leaks are hard to code by accident in Rust. > ... I enjoyed this gem and its descendants from the comments. What I see instead, commonly, even in big rust projects, is that
755.
by hansvm 2y ago
> ignored such complexities as geography and migration That article doesn't look compelling when refuting a claim that depends entirely on geography and migration.
756.
by hansvm 2y ago
Totally unrelated, but since you brought it up a bit, I find recipes written in French to usually be much higher quality than similar recipes I find in English -- enough so that I'll alter my country and language for search engines whe
757.
by hansvm 2y ago
> all of the above NFL says something about it being a wash for arbitrary data. All results are going to be tuned to assumptions we have about our data in particular (not too many discontinuities, sufficiently well sampled, ...). > ne
758.
by hansvm 2y ago
Trains and buses in the US, where they exist, have ticket prices on par with the amount of gas a single person would use driving a used car. For very long trips they're a bit cheaper (up to 2x). For very short trips they're often
759.
by hansvm 2y ago
Determinism isn't the issue though. Many responses are fine. The displayed one is bad, whether chosen deterministically or not. Some alternatives: - Passed out drunk - Crashed in bed - Slacking because drunk ... The issue isn't a
760.
by hansvm 2y ago
That almost makes the metaphor more apt. Wine is the real deal, and brandy is the distilled approximation.
761.
by hansvm 2y ago
> on top of vanilla code > without requiring an extra DSL Give or take some nitpicking about where a language extension stops and a DSL begins, those things are the same. The only question is the syntax. The key thing the referenced p
762.
by hansvm 2y ago
Another strategy I like for groceries is walking to the store during my "lunch" break each day. It's never busy, you get used to the layout and can shop faster, it's easier to only choose ripe produce and to be selective
763.
by hansvm 2y ago
One small point: Token selection at each step is fine (and required if you want to be able to additively/linearly/independently handle losses). The problem here is the high inaccuracies in each token (or, rather, their distributio
764.
by hansvm 2y ago
I interview people every day who don't recognize glaring flaws despite the syntax highlighting and other linting in their favorite environment telling them something is definitely very wrong. Much like with LLMs, I'm firmly in the
765.
by hansvm 2y ago
HIPAA doesn't care about your POS TOS. It either applies or does not. That said, it's both less broad and more toothless than I'd like. If FB convinces you to install a tracking pixel (like button) stealing your private medic
766.
by hansvm 2y ago
This article seems a little ranty, but you could definitely be forgiven for thinking that iPads are general-purpose computing devices. There aren't any hardware limitations, and for years (haven't checked the last year or two), Ap
767.
by hansvm 2y ago
I thought about it a bit more: - That last point (attaining sqrt(2)) follows from Bolzano-Weierstrass. Just unroll the embedding into an n^2-dimensional space if you have n points, and that sequence of embeddings (as a technical detail, cen
768.
by hansvm 2y ago
I don't have any references off the top of my head, but I can state a result (which hopefully is easier to search) and then build to the thing I claimed: - There's a well-known result that says you can embed any finite, undirected
769.
by hansvm 2y ago
That cosine distance works at all as a concept isn't terribly shocking, especially given our habit of norming everything. Cosine similarity in a unit vector space is monotonic with euclidean distance, and we're using this stuff in
770.
by hansvm 2y ago
It's the "at the expense of others" thing that makes it more morally grey, and the chain of cause and effect is short enough that people sometimes get up in arms about it. For some other actions on some sort of badness scale,
771.
by hansvm 2y ago
Or wants to buy one.
772.
by hansvm 2y ago
I've seen a lot of their engineers here on HN defending Google's position, and very few of the anti-Google crowd here claim to be (x)googlers. That said, I know a number of xooglers (myself included) who don't believe for a m
773.
by hansvm 2y ago
Sure. On the one hand, everything adhered to the letter of the law. On the other, he used his money to get served before other people in an otherwise similar position would have been able to do. I personally view that as more of a failing i
774.
by hansvm 2y ago
I'm not advocating for C or against Rust though. I'm saying that GP's request to report errors instead of crashing is a perfectly fine opinion, and using Rust as an example of a language which already traps most instances of
775.
by hansvm 2y ago
> Which libraries in widespread use know how to detect all of their possible bugs due to invariant violations and report them as explicit error values? We're talking about the cases that are already being caught somehow (bounds chec
776.
by hansvm 2y ago
A big wide base with a center of mass magically very high off the ground would still be tippy though.
777.
by hansvm 2y ago
Adding to the sibling @mathgeek's comment, that's only true when there are no outside forces other than gravity. You can see that by taking the counter-argument to its extreme: with gravity of epsilon, even a gentle prod at the to
778.
by hansvm 2y ago
> By design, AR models lack planning and reasoning capabilities. If you generate one word at a time, you don’t really have a general idea of where you’re heading. I have one minor quibble here, which is that the limitation described isn&
779.
by hansvm 2y ago
That's not the only option, but it's very reasonable and almost certainly the thing I'd start with, maybe the thing I'd finish with. You then pass an appropriate sub-matrix to this thing which does a whole-matrix computa
780.
by hansvm 2y ago
That pattern of doing something for the tail (and possibly also the head, if you're on a platform where aligned loads are preferred or required) of your data is very common in SIMD linear algebra. It's so common that my default wa

More ›