fulafel

7 ms·

121.
by fulafel 5mo ago
It won't be too little, it's still saving more people than died in wars and famines in the last 100 years. I don't really understand this "too late" failure of judgement unless you're assuming there's some
122.
by fulafel 5mo ago
> if it gets humid enough it stops It won't stop if it's ventilated with outdoor ambient air: 40C air can hold 51 g of water per m3 of air. 60C air can hold 130 g of water per m3 of air [1]. The curve is exponential. So, it wor
123.
by fulafel 5mo ago
Humanity needs to be in a serious hurry to ramp down fossil fuel use and production to curb the megadeaths. Eg the US has been going in the opposite direction for a while, net exporter of oil since 2021.
124.
by fulafel 5mo ago
Even if it's a wingnut dense place, there's good arguments for using a channel independent of your infra in a case like this. You (or Github themselves) don't know if their status page is pwned.
125.
by fulafel 5mo ago
The problem is more in language (or SIMD architecturally, depending on your POV). C semantics block too many of the necessary transformations that autovectorization would need to do.
126.
by fulafel 5mo ago
Autovectorisation is the main way SIMD hardware gets put into use, whether you think it's pretty poor or not. SIMD came to mainstream in 1995 Pentium MMX and has been proven rather difficult for compilers to target, but after 30+ years
127.
by fulafel 5mo ago
Futhark is a glimmer of light in the wasteland of C/C++ styled low level GPU languages.
128.
by fulafel 5mo ago
The more interesting question is whether GrapheneOS had this vulnerability. This published exploit sounds to be highly specialized to the specific build options.
129.
by fulafel 5mo ago
overflow check insns also in 6502 (bvs), x86 (into), m68k (trapv) - the gp was after something that wouldn't need to be checked every time, if interpreted as "no extra instructions" then that's a tall order. But the chec
130.
by fulafel 5mo ago
"If you turn off checks" is misleading, it's just incorrect by default in release mode: https://doc.rust-lang.org/book/ch03-02-data-types.html#integ...
131.
by fulafel 5mo ago
On its own we can't judge if this is a workable way to find vulns, as we don't know how many false positives you'd get if you ran it on all the code. (iow might be https://en.wikipedia.org/wiki/Base_rate_
132.
by fulafel 5mo ago
I can believe some people have implemented misconfigured thread pools, being worse off than with no thread pool at all. But this not an argument that justifies a "multitasking isn't a real choice for backends" position. (Then
133.
by fulafel 5mo ago
This take on native threads sounds overly pessimistic. 99% of backends do fine with the nr of native threads you can easily run in a single process at least on Linux. Not sure what the practical limit is now but 10k used to be no problem ma
134.
by fulafel 5mo ago
Tesla and SpaceX were both founded outside of recessions about 1 year apart, between the early 2000's recession and the great recession.
135.
by fulafel 5mo ago
Close: New versions go in unstable where development happens, testing is where things go to marinate for a while.
136.
by fulafel 5mo ago
That Exim seems to have a relatiely high density of RCE vulnerabilities that are being made public regularly. (The track record is considerably worse than other widely used Unix MTAs, roughly on par with MS Exchange.)
137.
by fulafel 5mo ago
Previously (2023): https://www.bleepingcomputer.com/news/security/millions-of-e... Previously (2020): https://www.exim.org/static/doc/security/CVE-2020-qualys/CVE... Previously
138.
by fulafel 5mo ago
These problems are very small compared to mounting death toll from the climate disaster. We are very late in ramping down fossils production but a far greater badness will come if we keep postponing it.
139.
by fulafel 5mo ago
> Go delivered most of the performance benefit at a fraction of the engineering cost. The biggest JS/TS shop on earth picked a harder, faster language for its flagship tool, and they did it because the effort calculus changed under
140.
by fulafel 5mo ago
What are some concise languages that are well received by humans (on par with Python)? Token efficiency might be a marked advantage. Clojure comes to mind at least.
141.
by fulafel 5mo ago
What would comparing rates across languages tell in the context of this benchmark? Are the tasks the same or robustly difficulty-normalized across the languages? Also somehow the 2 language comparison graphs (avg percentile and success rate
142.
by fulafel 5mo ago
This is terrible. Also terrible and underappreciated are the co2 emissions, we urgently need to outlaw fossil electricity generation, esp for data centers.
143.
by fulafel 5mo ago
Indeed, the body in the post made me think it was a url-less submission.
144.
by fulafel 5mo ago
Link to source blog post: https://gowers.wordpress.com/2026/05/08/a-recent-experience-...
145.
by fulafel 5mo ago
Right, so it blocked the first part of the chain. Which normally uses unprivilgeed network namespace to dp net admin inside that. I had been thinking of a RxRPC AF block for the second part of the chain which seems rarer. Systemd seems to h
146.
by fulafel 5mo ago
Both of the modules are (also) for desktop/workstation use. Though AFS could probably be retired generally.
147.
by fulafel 5mo ago
RxRPC is apparently an AFS (Andrew File System) thing.
148.
by fulafel 5mo ago
Both of these (copy fail and dirtyfrag) exploit obscure socket address families. Are these filtered by commonly used seccomp profiles in eg docker (assuming seccomp can express it)?
149.
by fulafel 5mo ago
This seems to be a different discussion than was going on up thread about: > in the next few years a "good enough" model will run on entry-level hardware
150.
by fulafel 5mo ago
Multi-token prediction is a refined form of speculative decoding. Researchers at Google came up with Speculative decoding in 2022: https://research.google/blog/looking-back-at-speculative-dec... (Fast Inference from Tr

More ›