491 karma · joined June 20, 2022
> Congratulations on empathizing with an authoritarian.
It speaks to the strength of different agency administrators if they can walk into the next oval office, grab the duly elected President by the arm, and say "stop hitting yourself, stop hitting yourself" over and over again. Putting a stop to that wouldn't be so controversial, I think.
> Stop being so gullible.
You are disregarding the election angle and instead misdirecting, the system of validating votes (according to Krebs' own assessment) is dependent on a system with publicly-known flaws.
I understand that the aim can be to enfranchise and enable more voters, but to that aim my statements are agnostic, except for revealing more facts about the case.
Here's an article supposedly of Krebs provenance, which implicitly lumps Trump himself in as a "malicious actor".
> can lead to uncertainty in the minds of voters; uncertainty that can be exploited by malicious actors
Maybe not something I would want said or repeated by my administration either, disregarding the veracity.
There's no date or byline either, so according to the authoritative FAQ, if this were to stand, it would be an admission of acting in bad faith.
Given federal government communications sprawl, it's quite a needle, pretty good performance in my opinion to root this out, disregarding sowing doubt about a federated election and who's will specifically it should / will service.
Voter inclusion (who should / may vote) is itself at issue, but even in the assessment here given DOGE findings unveils possible oversights, FWAB in the FAQ is cited to depend in part on SSNs and in light of the DOGE findings regarding 150+ year olds collecting social security, the security assessment itself does not describe a system that is definitively air-tight, or even terribly reassuring, if there's doubt in your mind about who voted, and how.
"lead safe mama" is an activist and filmmaker, who does not appear to have a scientific background, and who may primarily be motivated by taking on legal cases. [0]
Furthermore, in the toothpaste (informal, blogger-led) study it appears that the testing was performed by "Purity Laboratories" using "ICP-MS" instrumentation. [1]
According to "Purity's" own marketing it's admitted the "ICP-MS" may cause interferences, though they aren't specifically quantified. [2] Other laboratories primarily market the technique specifically for testing raw for contamination[3] and discusses "dynamic range", rather than what I speculate is "a complex matrix" though I myself am not technical enough to recognize the meaning there.
According to an instrument vendor, Agilent, there are other considerable number of other constraints and variables where the laboratory process itself could introduce issues[4] and alarmingly, I'm unable to locate any details regarding specificity or FPRs, and there's some evidence of targeted advocate edits on Wikipedia regarding the viability of this application of this technique. Furthermore these advocate edits specify a different methodology for health-related detection, rather testing wastes for result of contamination than testing products themselves. Even furthermore Agilent, one vendor that has strategically been selected for my review of technical details about the technique specifies "ICP-MS is an elemental analysis technique, meaning it is used to measure elements, rather than the molecules" but it's difficult to determine whether the technique can avoid creating FPs with molecules. The process itself seems like it is not fool-proof, it would take some expertise in plasma chemistry to understand and predict completely which I cannot provide, or probably some pretty intricate calibration details which are not available for this study.
It's conceivable that testing products represents a misuse of the technique, it at least differs in scope from the use-case described throughout the Wikipedia article. There's a section on Wikipedia regarding use of the technique for Pharmaceuticals but it is vague, speculative, and drifts substantially from the actual topic at hand into usage for forensics.
[0] https://tamararubin.com/2023/07/lead-safe-mama-llc-rates-for...
[1] https://tamararubin.com/2025/04/orajel-kids-anti-cavity-toot...
[2] https://www.puritylabsinc.com/heavy-metals-testing
[3] https://www.eag.com/wp-content/uploads/2023/11/High-Purity-R...
[4] https://www.agilent.com/en/product/atomic-spectroscopy/induc...
That is, from Mongo, you use Serde and wind up with only valid records operated upon, of a table of such values.
In Rust there's at least 5 types of everything, in order of strength:
- Value / unqualified / "owned"
- Generically, T
- Optionally mutable
- Mutable Reference - &mut T
- you can only have one of these for a given value
- Reference / Shared reference - &T
- you can have arbitrarily many for a given value
- Raw constant pointer - *const T
- you can have arbitrarily many, and they're not liveness checked
- Raw mutable pointer - *mut T
- you can have arbitrarily many, and they're not liveness checked
Now I say at least because things get appreciably more complicated when you find yourself dealing with lifetimes which apply to "References", those are indeed themselves types, but ultimately represent a compiler-executed calculus regarding liveness relative to some Value.They also can 'fan out' like a multiple-dereference pointer, but the tricky part is how the APIs for Types conform to these, for example;
Since there are 3 different types of things in a collection, then there are 3 different ways to iterate over them `iter()`, `iter_mut()`, `into_iter()` in increasing order of strength. Most of the breadth or early complexity arises from the urge to treat these as a distraction, rather than a fundamental aspect of systems code.
Crates / modules are a bit of a meme: https://www.reddit.com/r/rust/comments/ujry0b/media_how_to_c...
Bevy has done some work investigating build performance: https://bevyengine.org/learn/quick-start/getting-started/set...
I mean I guess a similar argument holds for colocating with telecoms that would have recently been cable or IPTV providers.
It'd be pretty tough to design around though if any of their caching infrastructure reveals viewership, or engagement data to intermediaries.
Flate2 is an odd crate so I'm not too surprised to hear there's issues with it.
It seems unusual to rely on an FFI for something that isn't extremely intricate, broad, or brittle like syscalls / drivers / runtimes, I'd think that one zlib implementation would be as good as another, but maybe I'm naive.
Thankfully crates makes it pretty plain who would see fit to build on top of that.
That's a lot of extrapolation, Marmite-flavored protein powder, no doubt?
> that comprising 3 full-body sessions a week for 12 weeks (Supplementary Materials, Table S1) [Table S1 describes 3 different bro-splits, comprised of 45 different assistance exercises and no planned progression in intensity, excepting "To failure"]
Ah yes, I see they've confused resistance exercise with resistance training, indeed a mistake.
Granted, it does seem like a fine time to avoid "fitness powders" in light of "The Rhisotope Project" https://news.ycombinator.com/item?id=40819617
I think it's OK, if there's a lion hollering you might need some gravity and momentum to hear it and they are certainly trying to find the new polarity.
I've been doing Leetcode in Rust which, statistically, is a poor choice. When I have asked, it writes solutions that look like python or C, unable to identify when iterators are the appropriate tool.
It's not authoritative reference material, which my background in IT trained me to strongly, strongly prefer to, even my own intuition in situations I've seen before.
For example since Vec can Deref to Slice, you can syntactically run Slice methods on a Vec; meaning Vec is a subtype of Slice.
[0] https://doc.rust-lang.org/std/ops/trait.Deref.html#deref-coe...
https://cockpit-project.org/guide/195/feature-virtualmachine...
Thanks for the tip, though.
Oh, boy, it's written in Rust!