HN user

petsfed

1,678 karma
Posts0
Comments619
View on HN
No posts found.

I mean, if you'd asked law makers in 1850 if we needed speed limits on the roads between towns, you'd be laughed out of the room. Does that mean we shouldn't have speed limits? Or does that mean that nobody in power dreamed that it would be a serious or wide-spread enough problem that just suing each individual for damages was insufficient? Sometimes, we even have to invent new ways to express the damages so that suit can even be brought.

As I recall, the main issue with that was that because it used facial recognition, the labor burden of enforcing that was significantly lower. If its just human beings looking at every visitor and trying to decide if they match a description, the venue has to decide "has this person done anything so egregious that all this extra effort is worth it?" which makes the tactic self-limiting.

With facial recognition, enforcing a trespass order becomes nearly zero cost, so it can be applied for basically any reason. I can sort of get to understanding the tactic for "this lawyer is actively suing us", but if its "this person said something mean about us online, and we can get a facial recognition match from their profile picture", it seems like a wild abuse.

Which is why that whole Radio City Music Hall situation was such a good illustration of the actual harm of facial recognition systems. If a potentially bad action is only kept "good" because the high cost (in labor or lucre) causes discernment in its application, then removing the cost will necessarily remove the discernment, almost guaranteeing bad actions.

Business owners should have the right to bar someone from the premises, and legal recourses to enforce that right. But enforcing that right should be sufficiently cost prohibitive that enforcing that right does not grant the business outsized power to limit the public's rights to e.g. express negative opinions of that business.

The AirPods Effect 1 month ago

Almost as if there's a limit to how many demands strangers can reasonably place on a person before we as a society agree that the person should put up boundaries (like, say, putting on headphones and outright ignoring the demands) and even go out of our way to encourage strangers to respect those boundaries.

I'm not calling you out, IncandescentGas, you're right and you're doing a good thing. I'm just saying its ironic that you jump to the defense of somebody who has made it clear that they don't believe others deserve the same courtesy you are providing them.

Well, no.

First, increased demand drives increased prices. This is the least controversial axiom of modern economic theory. So if you add a huge power consumer to a market, all consumers in that market will have to pay more. You can mitigate that some if that new, big consumer builds their own power facility, but the fact still remains that the local price in fuel (oil, coal, etc) or materials for renewable generators (turbins, solar panels, etc) will increase. Again, because demand increased.

Second, its one thing for things to cost more in a market that has a booming economy and plenty of high paying jobs. Home prices in the Bay Area are horrifying, but the poverty line for a family of 4 is $80k, which sort of grounds things. If energy costs go up by $100/year in the Bay Area, nobody notices. But if energy costs suddenly skyrocket in Great Falls, Montana (poverty line for 4: $33k) or similar that lacks a vibrant economy, the residents don't have much choice but to tighten their belts over the suddenly larger electric bill that has done basically nothing to actually revitalize their economy.

Or maybe the problem is that exchanging labor for scarce necessities only makes sense when labor itself is scarce.

I had a similar experience playing, of all things Operation Flashpoint: Dragon Rising. You play as a variety of US Marines helping Russia defend some island from a Chinese invasion so as to control some recently discovered oil deposits.

And at one point while playing this already somewhat implausible situation, I thought "that random PLA soldier I just mowed down would not have, in reality, signed up to be the invading force in a petro-conflict". It kind of killed first-person shooters for me for a while, because I accidentally humanized a poorly-programmed bot in a ridiculous, made-up scenario into making me feel bad for them.

Questions like "how would you search for a substring?" are so incredibly dependent on what you're doing on a day-to-day basis, and what you're doing with the data once you've split it. Just because .split(...) is in all the tutorials doesn't mean the codebase you've worked on for the last 5 years actually uses that specific call with any regularity, and it may well be the case that your codebase does use regexs more often (maybe for query-portability purposes).

I write bare metal firmware, primarily in C, and I've had to make it a point to explain, in most every interview I do, that I've only ever used malloc(...) in tutorials. "In my world, malloc is a 4-letter word". So while I know what it does, and how it works, I actually have to google its usage, and I'm not as keyed into its pitfalls, because every system I've ever worked on could not afford the risks associated with dynamic memory allocation.

All of this to say, bad interviewers go looking for a specific answer, good interviewers go looking for good process. All of the jobs I've held are ones that accepted that I was rusty on this or that specific call, but could think about the system holistically.

My point is that American suburbs are so aggressively designed for cars that even considerable cost and inconvenience (driving 30-40 minutes) is acceptable in the face of much MUCH lower prices (even if the prices are so low because the quality is commensurately lower as in e.g. Levi jeans since they started selling in Walmart). If we hadn't let our communities be planned around driving to begin with, then building a big box would be less effective.

People are so habituated to driving, and the walking/biking experience outside of inner cities so unpleasant, that the corner stores simply cannot compete. Of course a big box store is going to have better prices, they win on economies of scale. But we've allowed our cities to be designed around the total separation of residence and commerce, so that you have to drive an interminable distance on unpleasant, congested roads just to get to an ersatz corner store. May as well drive to Walmart.

The funny thing about the big box stores is that they can apply so much pressure to their suppliers (and their landlords) that the price difference outstrips differences in quality and convenience.

Once upon a time I lived within easy walking distance of my job, a big box retail store in one of these clusters. Elsewhere in the same cluster was a Market of Choice, which is basically a Whole Foods but for Oregon residents who are too cool for such a mainstream store. Despite the higher costs, I vastly preferred shopping at Market of Choice because I could easily walk there and the food was generally higher quality. Generically, I would consistently shop at stores I could bike or walk to, outside of that shopping center, and frequent restaurants I could get to.

Meanwhile, where I live now, there's a Haggen (an Albertson's branded Whole Foods competitor) across the street from a Safeway (which is the brand Albertson's uses in this particular market for their standard store). They carry identical store-brand items, but the premium store charges 5-10%, because they know that people will pay more just for a better shopping experience.

Obviously I don't speak for everybody, but I think a lot more people would accept paying a little bit more in exchange for not driving so much, but part of the problem is that, thanks to market effects and economies of scale and so on and so forth, they have to pay a lot more, for the same products.

I'd go further and say that modern suburbs are designed to completely isolate commerce from residence, in a way that doesn't just necessitate driving, it necessitates driving long distances to centralized commercial centers. So you can't just walk around the corner to a shop and get what you need, you have to drive, sometimes for miles, to get to a big-box grocery store. And because people didn't want to drive all that way for one set of things, just to drive another long distance for another set of things, you ended up with clusters of shops (or more frequently, clusters of big box stores), all in these centralized locations.

You can see this all over the east Bay in the San Francisco Bay Area, where its just miles and miles of residences, with all of the commerce on 880, 580, or all the various names for 185/238, from Gilroy to Vallejo. If you're more than a few blocks from any of those roads, you're driving or taking a bus to get to any shops. And East Bay surface streets are not exactly pleasant.

Just to buttress and embroider around your point that a fab is not a small business:

If there was a realistic way even to go from bare wafers to non-trivial custom chips in a small-batch fashion, you can bet there would be a cottage industry around it. I would love to live in a world where I could manufacture custom silicon as easily as I can manufacture a custom PCB or custom mechanical part.

But as it stands, quick-turn, rapid-proto "micro" fabs are obscenely expensive, to the extent that if you aren't absolutely certain you need the performance gains from custom silicon, justified by years of R&D that confirms the inadequacy of a multi-chip solution, then the idea is killed before any layout engineer is contacted.

Microfabs are either operated by research institutes, or they're booked solid for years, and basically printing money.

... which sort of assumes that the global downturn and the local downturn are completely unrelated.

Like, if a rent hike pushes out the tenant who has been there the longest, who has the most consistent revenue stream, in other words is the surest bet, then that, all by its lonesome, should be a pretty clear risk indicator to the bank.

I was working retail in Eugene, Oregon during the 2014 University of Oregon grad student strike. I had a little bit of a chip on my shoulder because I was working retail with a master's degree in physics (because I did not have the endurance to complete a PhD, but had not yet accepted that fact).

My then-partner was part of the strike. One of the strike demands was higher wages as teaching assistants. And while I worked 40 hours a week, for $11/hr, I made considerably more and worked fewer hours than her. She put in probably 30 hours a week just on her teaching load, plus an additional 30 hours split between explicit course work and dissertation work.

It's crazy that a job that requires excellent marks while completing a 4-year degree pays worse, has worse working conditions, and is considerably more competitive to get into than a job selling office supplies.

One of the other things the grad students were demanding (which they only sort of got) was paid parental leave, because they did not fail to notice that most of their professors were in their late 30s or early 40s before they could afford to stop work long enough to start a family. It was very rare for two academics to have children together, because of the heinous, career ending financial cost to having children when you were young enough that their high school graduation date was before your expected mortality.

Generalized in the sense of "this interface works as well on a tablet as a desktop computer", or "we can also generate ad revenue with this operating system" or "there should be constant invasive AI integration, even for users who don't want to and should not use such features, and who would pay a premium to avoid it if possible".

Not specialize in the sense of "here's your civil engineering operating system, which is different from your structural engineering operating system, and neither bear any similarity to your gaming operating system".

Not to white-knight microsoft here, but I think the problem they run into with every product is that because of their ubiquity, they rapidly reach saturation with most every specialized product they sell. You cannot grow a business if your market is saturated, even if you're the only one selling. So they have to find a way to expand their market. With specialized tools, that's done by generalizing, right? And anyone who has ever driven a screw with a swiss-army knife can tell you, generalized tools never work as well as dedicated tools. Thus, Word ultimately sucks. Windows ultimately sucks. Github ultimately sucks. They are all of them trying to be everything for everyone, because the alternative is just mumbling along, being really good at being tools, but being really bad at conveying profit to their creators.

There are a million ways to express the fact of the hormonal backlash without including a quote that makes it sound like killing will improve your sex life.

In context, its correct, that's not up for dispute. The question is "does it add anything to the context?" and more importantly "could a student misconstrue its inclusion as something else?"

You'd think that, being so educated on the hormonal backlash from experiencing trauma, that cops and the greater judicial system would be more forgiving of e.g. emergent hypersexuality in rape victims after experiencing a rape that Grossman calls out there. But you would be wrong, because even if Grossman wants his students to understand that concept for their own health, he wildly misunderstands the culture he helped create where the police view themselves as a thin blue line holding back the manifold forces of Chaos Undivided.

So he had a pretty good (not perfect) run up until the final 1/3, then had a staggering turn that only the author thought was profound or earned?

That's a man who lived his craft right there.

I've been saying this for years.

Yes, we have a gun violence problem. But notably, we do not have a heavy weapons problem. By and large, gun crimes are committed with guns that can be purchased legally somewhere inside the US.

So if the silver bullet to the gun violence problem is taking away all the guns (please do not misunderstand me, I DO NOT BELIEVE THAT TO BE THE CASE), then step 1 is to limit what guns can be purchased anywhere in the US.

But this whole 3d printer farse reveals something we sort of already knew: if people want to have guns (or have weapons in general), they're going to find a way. If you want to address the gun violence problem, you have to find a way to make people not want to kill, nor own guns, that's unrelated to how difficult/expensive it is to get guns. And you're going have to do that in the shadow of the constitution.

I don't know if its still true, but I recall reading once that CEQA had never been used to actually prevent or even slow the building of a dam or a mine or something. It had only ever been used to hobble otherwise neutral development. Its a good idea in theory, but I feel like the plaintiff ought to be able to articulate what environmental impact they are concerned about and maybe require a study from them in support of that claim too.

My number 1 complaint with Ubi games is that they all feel the same. Sure, in this one you stab, and in that one you shoot, and in that one over there you stab AND shoot, but it's all fundamentally the same. You've got a drone or a bird or a droid to tag enemies for you, and there's a straightforward shopping-list style crafting mechanic. There's also some vehicle combat, but its very limited, and its pretty rare that you're part of a larger group of vehicles attacking together - at best its a group of enemy vehicles coming after you (and the comedy of errors of those enemy vehicles crashing into each other trying to get to you, because apparently they didn't turn on pathfinding while in "alert" mode...). The whole thing looks like its chasing the annual-release pattern of Call of Duty, and the major sports franchises.

Lock-Picking Robot 6 months ago

Locks do not provide real physical security, they just keep honest and lazy people out.

Really?! I had no idea! I had such a miniscule understanding of what portion of the threat space locks address that the second sentence in the very fucking post you're replying is this:

Locks are very good at discouraging honest people and lazy, opportunistic people.

I'm so ignorant of the threat space, that the sentence immediately following that one goes:

They are not very good at discouraging generally skilled and motivated people, or people who are specifically interested in what's behind a specific door.

I guess you're right, the two sentences I wrote 76 characters before the one you're shitting all over as evidence of my ignorance have absolutely no bearing on the context of the statement I made. They just exist entirely disconnected from any other sentences in that same post. I bow to your superior intellect and analytical skills.

Lock-Picking Robot 6 months ago

Eh, its more that any one security tactic will almost certainly not cover the entire threat space.

Locks are very good at discouraging honest people and lazy, opportunistic people. They are not very good at discouraging generally skilled and motivated people, or people who are specifically interested in what's behind a specific door.

Locks are no obstacle if the intruder is willing to use social engineering. But if all they're trying to do is get into my garden shed, they're going to have to manipulate me or my spouse. Or somehow get past my dog. Meanwhile, my dog has absolutely no bearing on a bad actor getting access to my bank account. But similarly, bringing the full might of the best electronic security to bear to protect a chainsaw and a rake seems a bit excessive. And sort of beside the point, since I've not built my garden shed to withstand creation of an additional door (by e.g. a sawzall or a fireaxe).

I mean to say that the only time I've ever needed to diagram a sentence to figure out what was being said was while taking Philosophy 1010, because the cheapest translations available of e.g. The Republic was a bit too opaque for me.

There's certainly a lot to be said about the manifold interpretations of Platonic Idealism; what I'm saying is that when we've historically introduced new philosophy students to things like Jowett's translations ("But tell me, Zeno, do you not further think that there is an idea of likeness in itself, and another idea of unlikeness, which is the opposite of likeness, and that in these two, you and I and all other things to which we apply the term many, participate-things which participate in likeness become in that degree and manner like; and so far as they participate in unlikeness become in that degree unlike, or both like and unlike in the degree in which they participate in both?"), there's also a grammatical issue. Yes, I can deconstruct that and reassemble it in more colloquial terms. The problem is that for a lot of students, they don't develop interest enough to engage in the deconstruction until after they've gone through the arduous process of reading that and thinking "WTF?!"

The real barrier to students reading Plato has historically (and correctly) been the dismal quality of translations available. I always hated reading plato because the translations available to me were significantly more concerned with carrying into the modern day the wonky syntax and sentence structure of ancient Greek philosophical writing, and less concerned with translating the underlying ideas into language understandable by a 19 year-old engineering major who can barely spell their own name.

My point was not "oh, social media bullying is some kind of special case compared to other ways kids today bully their peers". My point was "modern bullying is different from historic bullying, and dismissing modern bullying as the same as historic bullying is intellectually lazy"

I remember when a bully would have to go up to you themselves to mete out whatever harassment, and you could avoid a lot of it by just being aware and avoiding that particular person.

Juxtapose that with today, where any one bully can create dozens of accounts to bully in a swarm, and the bully has constant access to you from your own pocket. Also, a person in Minsk or Timbuktu or whatever couldn't just come up to your house in the middle of the night to harass you out of boredom.

This "we could do X before computers, why are we trying to ban X-with-computers now?" line of arguments is just intellectually lazy. If a bad behavior was well moderated in the past because it was labor or resource intensive, the sudden removal of those constraints is a material change that demands revisiting. Put another way, if a constraint stops working, we should change constraints, not just do the old constraint with a confused expression on our faces.

Nowadays, when I hear "German Engineering", I internally translate it to "German love of complexity and bespoke/manual manufacturing".

The extreme depreciation of BMWs and Germany's loss to the Allies in WWII are both aspects of the same phenomenon; that fact is very funny to me.

The leadership of the Whites were not the moderate monarchists who just wanted Nicholas to abdicate to literally any functioning adult. They were the “Orthodoxy, Autocracy, and Nationality or death!” types. Their explicit goal was a restoration of pre-Revolution autocracy, whose brutal dysfunction was the explicit reason for the February revolution in the first place. The Whites were not good people, and it’s a mistake to characterize them as simple, noble anti-communist fighters. Most of the White leadership that survived into WWII went beyond just collaborating with the Nazis on invading Russia, but were onboard for all of the Nazi program save for “Ukraine belongs to Germany now”.

Don’t misunderstand me, Stalinism was worse for Russia than the Czars, but there’s really no White-victory scenario where it’s all sunshine and roses and limited democracy. That option went out the window with the October revolution.

All I’m saying is that there is no better illustration of how bad War Communism got than the fact that people looked at the literal pogroms and said “maybe that’s not so bad”.