madars

Project website: https://satcom.sysnet.ucsd.edu/

Paper (CCS '25): https://satcom.sysnet.ucsd.edu/docs/dontlookup_ccs25_fullpap...

Of course, but the choice is standard and thus the answer is 9. I can define a non-standard sqrt(x) which sometimes gives the positive root and sometimes the negative one, and then sqrt(sqrt(16)) could be -2 or undefined (if I defined sqrt(16)=-4) but that's just silly - the only reasonable interpretation for what the calculator should show for sqrt(sqrt(16)) is simply 2.

There is no choice here - each inverse is uniquely determined. That's similar to how 3 and -3 are both square roots of 9 (i.e., solutions to x^2=9), but sqrt(9)=3 as it denotes the principal square root, which by convention is always the non-negative value. Of course, in a different context we might design functions to have multi-valued properties, like atan2(x,y) != atan(y/x) in general (atan2 takes quadrant in account and returns full range [-pi, pi], atan only returns principal values in [-pi/2, pi/2]) as practical applications benefit from preserving quadrant beyond just the principal inverse (or not failing when x=0!)

arcsin(arccos(arctan(tan(cos(sin(9)))))) = 9 (in degrees mode - when regular trig functions output pure numbers, those numbers get interpreted as degrees for the next function and similar for inverses - calculator style), because each intermediate lands in the principal-value domain of the next inverse (e.g., arctan(tan(x)) = x when x \in (-90°, 90°) and the intermediates happen to be in those ranges). Specifically, sin(9°) ≈ 0.156434, cos(0.156434°) ≈ 0.999996, arctan(tan(0.999996°)) = 0.999996°, arccos(0.999996)≈0.156434°, arcsin(0.156434)≈9°.

Anubis is MIT licensed so you should be able to modify it freely: https://github.com/TecharoHQ/anubis/blob/main/LICENSE It would be very funny if they did that though.

By the way, you can also construct heap in linear time - instead of doing n consecutive insertions, at least half of which require log n work, you can apply sift-down operations from bottom up, beginning with the last non-leaf node and working backwards to the root.

That way, roughly half the nodes are leaves (requiring no work), a quarter are at the second-to-last level (requiring at most 1 comparison/swap), an eighth at the third-to-last level (requiring at most 2 comparisons/swaps), and so on. Summing up 1 n/4 + 2 n/8 + ... gets you O(n) total complexity.

See https://en.wikipedia.org/wiki/Heapsort?useskin=monobook#Vari...

Integrated Intel GPU and no graphical system, just KMS VT (text console). That's what made it so frustrating - only displaying a console should not result in kernel panics under CPU load! Admittedly, the experience was anecdotal and years ago and I heard Debian is doing less of a RHEL-style "frankenkernel" now.

>I've never had a Debian system break without it being my fault in some way.

Debian is great but I can't say this is a shared experience. In particular, I've been bitten by Debian's heavy patching of kernel in Debian stable (specifically, backport regressions in the fast-moving DRM subsystem leading to hard-to-debug crashes), despite Debian releases technically having the "same" kernel for a duration of a release. In contrast, Ubuntu just uses newer kernels and -hwe avoids a lot of patch friction. So I still use Debian VMs but Ubuntu on bare metal. I haven't tried kernel from debian-backports repos though.

The blog kept redirecting to the home page after a second, so here's an archive: https://archive.is/SE78v

You don't even need to import foreign doctors: just increase the number of residency slots.

All modern Zcash wallets are shielded-only.

The article doesn't show pictures of the unit itself, focusing on the CT scan, so here's the recall notice: https://www.anker.com/a1263-recall

A great resource for visually seeing progress is https://sam-jaques.appspot.com/quantum_landscape (click "Prev"/"Next" to see other years) - it makes very clear that incredible progress is happening - this is a log-log plot.

The second Chrome drops uBlock Origin (as part of their "Manifest V3 without blocking Web Request" plan), I'm off to an alternative browser. Enough is enough.

A bit more context: BuyVM is a legitimate business, popular with hobbyists, and has features that are hard to get elsewhere (e.g., BGP sessions). They do take a pro-free speech stance but they are a far cry from bulletproof hosting ("shady operators"). An imperfect comparison at a massively bigger scale would be Cloudflare's prominence in certain contexts.

This comment (currently downvoted to hard-to-read grey-on-grey) is an excellent example why you should always enable "showdead" in your profile, use user CSS to make downvoted comments readable again (e.g., https://news.ycombinator.com/item?id=41514726), and browse https://news.ycombinator.com/active instead of the homepage once in a while.

You can opt out of it if you use API to register your WireGuard public key. Specifically, you pass in hijack_dns=false https://schnerring.net/blog/use-custom-dns-servers-with-mull...

TLS does this with GREASE (Generate Random Extensions And Sustain Extensibility) - https://www.rfc-editor.org/rfc/rfc8701.html . HN discussion: https://news.ycombinator.com/item?id=39416277 (19 points, 8 comments)

Go's implementation of JSON format for protobufs also does this: https://protobuf.dev/reference/go/faq/#unstable-json

> To avoid giving the illusion that the output is stable, we deliberately introduce minor differences so that byte-for-byte comparisons are likely to fail.

This reads like standard AI slop. A giveaway is structured Attempt/Methodology/Efficacy pattern repeated all the way through the article, while top level categories are overlapping. (ZeroGPT: "Your Text is Likely generated by AI/GPT".)

This comic is evergreen https://i.redd.it/d3423w2g5ur21.jpg (source: https://old.reddit.com/r/comics/comments/bcdlbf/hello_old_fr...)

They have RSS feeds for new/updated papers: https://eprint.iacr.org/rss/

Looks great! Could you add eprint.iacr.org (Cryptology ePrint Archive)?

All those promises are "parchment guarantees", to use Madison's term. For example, social security distributions were promised to be tax exempt, and they were from the institution of the program in 1935 up until 1983 - talk about reliance interest!

Yeah, replace x.com with xcancel.com or nitter.poast.org, e.g., https://xcancel.com/Robert_Palgrave/status/18562734059656934...

This is fantastic. One rarely discussed use case is avoiding overzealous "alignment" - you want models to help advance your goals without arbitrary refusals for benign inputs. Why would I want Anthropic or OpenAI to have filtering authority over my queries? Consider OpenRouter ToS - "you agree not to use the Service [..] in violation of any applicable AI Model Terms": not sure if they actually enforce it but, of course, I'd want hardware security attestations that they can't monitor or censor my inputs. Open models should be like utilities - the provider supplies the raw capability (e.g., electrons or water or inference), while usage responsibility remains entirely with the end user.

Oxford Music Online has a great article that goes into a lot of detail - https://doi.org/10.1093/gmo/9781561592630.article.44010 and fittingly remarks that

>The organ is, together with the clock, the most complex of all mechanical instruments developed before the Industrial Revolution. Among musical instruments its history is the most involved and wide-ranging, and its extant repertory the oldest and largest. Despite its essentially indirect and therefore relatively inflexible production of sound, no other instrument has inspired such avowed respect as the organ, ‘that great triumph of human skill … the most perfect musical instrument’ (Grove), ‘in my eyes and ears … the king of instruments’ (Mozart, letter to his father, 17–18 October 1777).

Stephen Malinowski's famous YouTube channel of animated graphical scores has a great selection of organ pieces: https://www.youtube.com/watch?v=JXaCmhAMnHs&list=PLA20FE13E9...

St. Cecilia, pray for us!

Neat! As a human one can recognize that this question embeds a computation and use a standard trick - explicitly ask the LLM to use a program to generate the answer. LLM's are great at generating code but not necessarily that great at executing it "in head" (e.g., "what's the numeric integral of foo?" vs "write a Python program that computes foo"). Some instances of this are noticed by models themselves (I guess by now all know that they are bad calculators so would whip out code to do multiplication) but still a lot of them remain. Concretely, Claude 3.7 with "How is ahnentafel number 67 related to me? Use a program to help you." gets to "your father's father's father's father's mother's mother", whereas without the hint it indeed trips up in arithmetic and logic errors.

For those unaware of the reference, a famous example was Tesla. HN discussion: https://news.ycombinator.com/item?id=33621562 (Tesla has used space characters in internal emails to identify leaks), see also a general discussion of related techniques: https://en.wikipedia.org/wiki/Canary_trap

After seeing "Jigar Kumar" cognitive exploits on xz mailing list a maintainer would be excused (and I'd even say, encouraged) to just ignore pressure tactics altogether. It's an open source project - if it works great, if it breaks, you get to keep both pieces.