manwe150

Even so, it'd be better to lose a small fraction of trains than make them all late, since have 2x trains at 2x intervals operates at much lower efficiency (with the first train being over capacity, taking much longer for people to get off and on, and the second train half empty -- unloading feels almost quadratic in the size of the stack of people to push and pop). Of course, yes, that assumes you can ensure it's only a small fraction of trains being skipped and that each train is big enough and scheduled frequently enough to move the expected passengers. Those seem like different problems to solve though, once the trains aren't allowed to run late.

Ionizers are described as "air filtration using electrostatic attraction", while the ESP page I linked as the alternative specifically lists being filterless as its distinguishing feature, and is described as typically using lower voltage in home applications.

That doesn’t seem to follow from the Wikipedia article on them, which seems to say one is the same as the other. Though there’s a third option with a similar name that is used without a filter https://en.wikipedia.org/wiki/Electrostatic_precipitator

I’ve often wonder if cancelling trains (and busses) would make them on time more in practice too, by queuing theory. If you just let them run late, that starts to compounds until you get a Bernoulli distribution (exponential arrival delays) as the system becomes unstable. If canceling a train means the next one can run actually on time, then the system is forced back towards stability with predictable reliable times. Sure you had to miss one slot, but in the alternative scheme where everyone is late, you miss every slot — and average wait times are doubled in an exponential process compared to a periodic system that sometimes skips an entire train intentionally.

Isn’t that just a marketing term for ionizer?

That’s the design appeal of ionizers like BlueAir (not most ionizers, but specially those designed to supposedly prevent ozone production). They comparatively sip power while getting the same measured reduction in particulates and ozone.

Until either the enter key wears out and breaks, or we get pervasive sandboxes that act according to the users interests. It seems like Windows Vista UAC all over again for a new generation.

You know how to feed the site certificate into gpg before every login to check for url phishing attacks? I usually have to ask my password manager to do that (which didn’t exist 25 years ago)

I’ve never seen an implementation where passkeys deleted my password or could lead to a situation where it was locked out. They usually just bypass the password+2fa dance when functioning, and otherwise just don’t matter if they don’t function. (Well perhaps except home depot, where it gets mad at me every time I use a password and tries to install a new copy of the same broken passkey as last time, for my convience)

I asked Fable to write me a small CRUD app recently so I could keep tabs on my todo list and edit it. It chose to write the database as YAML files stored in git. When I asked why not use a database, it defended the choice by saying it would let me edit the files by hand whenever I wanted and sync them to where ever I wanted and get reliable git merges maintaining consistency, so I didn’t have to rely on remote database operations for that.

That’s assuming the reporter reproduced it. I’ve seen so much slop these days. Sometimes people post whole conversations in which the final answer from the AI is that the initial report was impossible to fit to the data after reproducing it, then continues in circles arguing that proves the original reported issue was real and the refuted cause isn’t actually refuted (eg it said there was a bug in function A, in code which never used function A — and that example was with the top tier models just released last month)

About 1700sf with taller than normal ceilings. It’s unnecessary bright in the evening, but blends with sunlight in the day. I feel like our brains must interpret light differently depending on the angle it reaches the eye. I used to rent a place that was 900sf with normal ceilings, which had 4 of those tiny sort of ones in the living room, and it was painfully bright in the way they concentrated the light into spots so we never used them (unfortunately they didn’t have dimmers)

I realized I did accidentally almost double my final number, so 500-1000W would be more accurate for the way this condo was built (depending on which extra lamps we are including—does every understair closet light count or just the main rooms, at the low end estimate)

A lot of bug reports aren’t valid, or handle a case that can’t realistically happen or realistically be handled (eg what do you do if you detect a crash while the prior crash is crashing and the logging pipe is throwing errors)

I feel very torn as a maintainer on this, since the only way to respond to the increased noise from AI has been to have AI do the research for me to extract all the links and line numbers I used to have to find by hand to explain why the PR needs more effort to be completed. But I also will be highly dismissive of any submitter who just posts AI text without cleaning it up first. It feels hypocritical, but the alternative is that I just can’t respond to most people instead due to limited bandwidth. I mark if something comes directly from the LLM though and to try to express my degree of confidence in its claims.

Yes almost 60 sounds right, and I keep the dimmers at about 50% usually since it’s a small condo. But it also very /evenly/ illuminated. The builders chose this, not me. Your typical incandescent hallway or stairwell probably used to have 1-2 lights (60-120W) to cast enough light to get around. Now that same hall in modern LED style might have 4-6, because it eliminates shadows, not because it makes it brighter in a given spot. Same for the kitchen—it probably used to have one central 100W light and 50W over sink, and now a modern style build in the same space might have at least 6 LED ceiling lights+6 under cabinet lights

No, that would fundamentally break the way (good) hashing works if I could add known data and get it to reveal the secret data

Across my house, I think I have about a kilowatt or two of LED lights, mostly in the form of ceiling cans at 16W each. Running everything full brightness would thus be about $0.30-$0.50 per hour. It’s the paradox that the same rooms probably would have used the wattage with the old incandescents, it just would have had a lot more shadowy corners.

Seriously? It is also logical to kill anyone accused of a crime, in case they actually might have really committed it and might again. Is trial even worth the risk that one guilty man should go free?

Also seems like it isn’t clear that they ever owned it. There appears to be a mistake in the contract asserting mutually inconsistent clauses, but it appears that both parties agreed on the not-owning interpretation for nearly 30 years, which might already be a far simpler contract resolution (depending on jurisdiction, it can almost immediately binding as soon as both parties accent to that reading) than also having waited out the statute of limitations

I’m not sure if either of you are being honestly factual or snarky, but I seem to recall this hypothesis is actually quite highly contested in social circles, and not at all as strongly supported as the parent seems to state it as fact

Why not do both though? I often now write my human summary, then paste in also the content you could get by asking a bot (with markers for which is which). You’re perfectly welcome to ignore the bot text, or ask your own bot, but bots are pretty slow, so I also don’t want to wait for it to “decompress” that 300 bit blob to get the detailed page back. But my human summaries are also only intended to be 50 bits — compressed again to just pass the signal I mean to convey, and not the whole prompt needed to fetch the right info from the right place to substantiate the claim.

TLDR the length was the same curtesy of tl;dr before, just now with a different name.

Several of my coworkers — it’s not that unusual that if you can max out a couple accounts, the companies will obviously notice you (as a high cost customer), and sometimes offer more

That only makes sense if money is a durable good destroyed by use. But money is improved by use and lost when put under a mattress. In economics terms, MV=PQ, and your proposal sets V low, which harms Q (goods available for sale)

It was about lisp ;)

Secret mode: ./julia —-lisp

I challenge you to dissemble your bespoke furniture and have it still usable after reassembly

It seems pretty much how law (specifically common law) has always worked though. If Congress thinks the law should apply to them, then Congress's job is to write a law that applies to them. The president and judicial branches are not supposed to make new laws, but rather to enforce the laws as written.

My reading of the EPA's statement here is they also wanted to mildly warn the companies that if they ever in the future want to contemplate connecting to the grid (eg. to bring down costs), they should be aware that bypassing compliance now could have expensive consequences for them later. Of course, Congress (local or federal) could also change the law in the much shorter term, so there's risk to them either way if they design without any forward thought for how to comply. On the other hand, if this inquiry to the EPA was merely requesting clarity on whether they are exempt from the reporting paperwork, but are using otherwise standard and compliant designs, then maybe the law is already working generally as intended.

Use a Mac or iphone? The opt-dash key combo has mapped to em-dash for a long time, and I think you can now just long press the dash key too. The opt key doesn’t scale beyond the hundred or so extra characters it gives space to live on the keyboard, but it sure is a lot better than memorizing Unicode code points, and the rest I can get from the Fn (or Fn-E), which activates a Unicode search panel just about everywhere. I don’t know which Linux toolkits do something similar, and Windows Character Map seems to be somehow even worse than I remember from several decades ago.

I don’t get that argument. Most of the time by the end of the session the comments from the agent encode tricky details that I told the agent to write down so it stops making “simplifying” assumptions. Thus comments at the end of a couple days of agent-only coding, when I start to actually read and edit the prose, contain the details which aren’t possible to know from reading the local code. It may help that my last couple sessions before I start reading the code myself are variations on telling the agent to self-review and improve the comments in specific ways, so the comments left are only those the agent thought remain meaningful at clarifying unexpected interactions between the local code and other code that needs to be referenced to understand it.

I think the parent meant it is more interesting to pose new problems than solve them. Posing a new conjecture along the path to solving something is a close cousin, but still seems more bounded than proposing something novel to prove—if only because proving that something novel is also actually interesting is subjective and thus difficult for a different reason.

The difference is Linux has fork, while Windows doesn’t, so Linux cannot do accurate accounting (without wasting a lot of memory) while Windows can.