yodelshady

Unpopular because it's victim-blaming and awful.

If party A operating multiple tonnes of high-speed steel hurts party B who is not, the dangerous party is party A. They might have been careful enough to discharge any direct liability, but they are *always* the party who bought multiple tonnes of high-speed steel into the situation.

There are at least three different actually deployed methods for doing so, all with working principles older than nuclear. If you want to be in an energy debate knowing them is not hard.

I have to downvote you, because requiring others to spend their time explaining basic technologies as a rhetorical tactic is *rude*.

The counter to that is - sometimes the toaster fuckers are right. See: civil rights, women's suffrage, LGBT.

My pet dream: Everybody Gets One, maybe two, things to be "radical" about. You reckon the earth is flat? Sure. You don't get that AND vaccines are micro-chipped AND an antisemitic 9/11 theory. Spend that token wisely.

Cheap, insolated land and a positive correlation between insolation and electricity demand makes PV just a no-brainer in the US.

I... put politely, don't understand how manufacturers can make that claim, it depends on exogenous factors. That said, you're right, the newest turbines are impressive structures and more consistent at their job.

Also that said, andy_ppp is right, or will be soon. If you want to make a dent in our fossil fuel needs on a cold windless day, you'll have giant globs of excess energy on warm windy days, that is simply orders of magnitude more than any practically-costed battery can store. At that point, who cares if electrolysis is only 30% efficient?

I'm not currently aware of any wind turbines in the UK being powered down due to lack of demand, so the 30% capacity factor is exclusively due to lack of supply, i.e. no wind to turn the turbines.

It's definitely worth looking at with another 3x wind capacity or so.

$500 M feels like the sweet spot for me actually. It's enough you can actually do something useful, unlike say "fusion never", but it's no so much as to be unauditable. If RR can't produce some goods with that amount (and, institutionally, they have the technical competence), you can ask why.

Re 1 - it really is quite astonishing how uninterested nuclear plants seem to be in burning their fuel.

It makes perfect sense if you consider the original reasoning - to make weapons, then progressively adapted to make them less useful for making weapons.

Still, I can't help but feel that fans of this mythical "innovative private sector" have a lot to answer for. Here's a 100-fold optimization in fuel use - not in fact a trivial cost, even for nukes - that's been on the table for more than fifty years, with the only drawback that nobody has literally given you the schematics. Now we're burning a shit-ton of gas and funding the tanks for the privilege.

JFC, The entire land area of Germany devoted to wind turbines would not produce that right now.

Germany heats with Russian gas, or nuclear, or does not.

On a related note, how much I wish 1440 / 1600 vertical had become dominant for PCs, rather than just "bigger number better".

It's so much more of a performance : sharpness sweet spot than 1080p - sorry, I can see pixel effects on a typical laptop - or 4k - which even desktops typically don't drive well.

> a lot of gas gets used for heating - nuclear isn't useful there.

Erm, no, that's just not true. Ironically, the best counter to that is... Russia, which successfully supplies a few million with extremely cheap hot water.

Note that the "cockroaches" slur originated from Queensland police: https://www.brisbanetimes.com.au/national/queensland/police-...

I won't be giving QL my tourist money in a hurry.

I don't know. I hate iOS with a burning passion, but the SEs/minis are the only devices that fit in my hand and have software support.

Android needs a way for OEMs to get some worthwhile profit in more than one market segment, and loading their own proprietary OSes isn't it.

It's honestly been 7 years since a had a smartphone I was excited about.

Controlling a tokamak: https://news.ycombinator.com/item?id=30379973

AlphaZero is RL as well.

Isn't that the sort of thing you thought of as a kid when thinking of AI, rather than "here's another way to serve ads and make hopefully-not-racist financial decisions"?

If you're looking for a podcast, the BBC have previously produced a crypto documentary that did not miss the salient fact:

https://www.bbc.co.uk/programmes/p07nkd84

Made some long journeys fly by.

per ofgem, wholesale prices have been greater than Hinkley's £92/MWh since October.

And that's before we add heating to the winter electric load. Conservatively, and I am really trying here, you're doubling grid load. If we go by my household, you're adding a zero.

I can find multiple >12 hr periods in this fairly mild winter where wind has essentially failed on demand (<5% satisfaction). I'm sorry but if someone says "storage" when talking about even that sort of shortfall they're lying.

It can't, source, gas prices in Europe now. Which we are paying. To a dictator who's overtly planning to invade the East. Because the sun isn't shining and batteries are three orders of magnitude too expensive.

Renewables get exactly as much credit for what might happen in future as Communists or NFT sellers get for how awesome their plans are. Nuclear already delivered, in some places still does, winter heat. Its detractors are responsible for gas reliance.

AGRs also perfectly demonstrate some of the more perverse elements of the nuclear economy.

They can be (at least were designed to be) refuelled online. An operator's dream - 100% uptime! And for the same reason, a security nightmare. So no exports.

Coal use is up: https://www.dw.com/en/coal-and-fossil-fuel-share-of-german-e...

Gas use is also up, at 400 €/MWh that goes to a dictator who is overtly planning invasion of Eastern Europe. The southern part of which is rapidly becoming a desert: https://www.kyivpost.com/ukraine-politics/desertification-of....

This is not brave, or principled, or sophisticated. It's a death toll with slightly higher plausible deniability.

I get that an NFT typically contains just a URL, because actual data is too expensive to store on the Enormous Work Disposal Unit(TM).

Surely it's at least possible that accessing the actual asset behind the URL is authenticated by the same public key that verifies your "ownership"? And indeed for the rest of the concept of ownership of the asset to be taken seriously - i.e. I the buyer can tell the hoster "I have made a local copy, or just plain feel like burning stuff, delete your copy please". Or "please move the asset to this hosting service instead".

Obviously that still has the same copying issues that the RIAA and MPAA and whoever else have failed to deal with so you probably aren't going to do any better. But, at least try?

I check you, Rust static-size arrays that implement common traits... as long as they're no more than 32 elements long.

Yes, this bit me hard on first Rust-based project, to the point of abandonment. (A list that grow and shrink is absolutely not the same thing).

a simple tuple of appropriately-sized integer types would only need 8 bytes, which coincidentally is what MS will need come 2043.

That does mean rolling your own "greater-than-or-equal" operation though.

Yeah, dismissing people who believe dumb things as dumb is a horrible thought-terminating cliche. Even the dumbest functioning human is really smart.

We're just much better at creating stories ("convoluted logic") than we are understanding our belief-making process.

Compared to last year: I moved from Python to Rust, VSCode to neovim, and having no work or social life to having some. I suspect the latter is the primary reason I only got to day 14 this year.

That said, it is definitely easier, by a mile, to throw away an approach that turns out to be faulty in Python, which means it's easier to try it in the first place. It's easier to simply write down the approach and figure out the implementation when you need to as well.

Interested in getting more out of vim, so far I'm quicker for some things, but I'm not convinced I'm quicker overall.

https://en.wikipedia.org/wiki/Ribosomal_frameshift if you really want to bend your brain a little.

As on other HN threads, transmissibility (Omicron is almost certainly more transmissible) has an exponential effect on healthcare demand and severity is linear (1). Plus if you're wrong about it being less deadly, no do-overs. Though, arguably, it's already way too late to do anything but look like you're doing something.

(1) Pseudo-linear? It's more like shifting the peak of a bell curve.

I would add for practical skills (including some research!) videos and podcasts seem to offer more feedback. Nobody in a book ever tells me what a flange or spline or baulk ring actually is, nobody in a video does either, but in the latter I get to see it and make my own, usually fit-for-current-purpose, inferences.

Closer personal example: I spent weeks trying to bully a supervised machine learning approach into a reinforcement learning one, because the 800-page reference book I used (that claims to cover all machine learning, and is well regarded!) in no way acknowledges the existence of this sub field. For whatever reason, and across multiple fields, I've never found static text to be good at "here's what you should be looking for", and I don't think it's reasonable to discount that knowledge as being valuable.

In fairness, that's a part of the point - it's a sustainable sabotage manual, for people who want to see retirement, which necessitates things that are hard to identify as malintent. Yeah, you could blow up the plant once, but your next action should probably be a plane out the country, and even with a ready supply of saboteurs, the vulnerability may well be patched.

Kind of like the Coventry problem (actually, identical to the Coventry problem).

I'm a little surprised at that reading, but fair enough.

I admit I only skimmed it, but I thought the point was "the employer" is not a homogeneous entity. "The employer" therefore doesn't make the wage decision. Managers with some extreme perverse incentives do.

What's bullshit is the idea that as soon as an enterprise does something inefficient, boom, they'll be outcompeted. They might, eventually, if no market failures or regulatory capture exists. So in reality, large and even small/medium organization with a successful pattern can tolerate a lot of bullshit.

Well TIL that printf won't lock stdout.

Calling stdout.lock() by itself makes little difference for me. Using BufWriter as you have I get ~80 MiB/s (slightly ahead of PyPy).

My implementation also has/had a String allocation in the loop, kicking that out got me to 120 MiB/s. I tend to write it as though the numbers/strings could be changed by a product manager at any time.

Final edit/errata: I get up to 140 MiB/s by appending .to_string() in the integer case instead of format!(), which makes some sense.

I didn't bother saving my original C implementation, but in retrospect I think I didn't use string concatenation, as I have with other languages. Redoing it with string-concatenation, C is slightly slower than Rust, 110 MiBs. Also makes sense as concatenation in C is slow.