You voted for Jill Stein? Can I ask why? What actually appealed there? I mean, I am very, very much in favor of strong environmental policy, but I can't imagine voting for the current green party under Jill Stein.

It's tricky because the power companies can do it at an industrial scale for far cheaper, however they still will like economize too much and end up still being unreliable. It certainly would be nice to never have a power outage again.

If that's their one big advantage, then they can't hold a candle to battery storage, which "fire up" (figuratively speaking) whenever needed as well but, rather than a matter of minutes like the gas turbine, it can be a matter of fractions of a second. Now, you could also mention that, running on a fossil fuel, the gas turbines can potentially run for a lot longer than the batteries can hold a charge, but that A. was not included in your list of big advantages and B. may not be much of an issue since batteries like the one in the article can still run at peak output for four hours or so.

Well, bear in mind that, as I said, the primary purpose would be to get the first loads of water/foam to the site of the fire as quickly as possible. Even if all the drone systems did was that first load, it could still make a major difference in heading off a fire before it grows into something that actually needs large volumes of water. So the idea would be that the drones sit with preloaded tanks at remote, probably unmanned, hangers spread out through the area you want to protect in a much more distributed fashion than you can manage with manned stations. The hangers themselves could be basically autonomous with solar panels for power to keep the drone batteries charged. The main idea would be extremely fast first response.

Now, if those drones could continue to help fight the fire after that first response, all the better. However, that would just be a secondary benefit. For that purpose, there could just be larger tanks onsite at the hanger locations. Filled either by pumping from a local water source, or just filled up by multiple drone deliveries (the same system that would fill the drone tanks in a fire situation could simply go in reverse to fill the larger tanks from drone tanks).

Absolutely no need to get flatbeds with pumps and spare tanks out to a location under such a model, everything would already be onsite. You're still talking about a model where you fight a giant, raging forest fire. Once again, this is a model where you try to stop that from ever being necessary by stopping the fire in its tracks when it's tiny. Also, the "tank" in question would almost certainly not be a rigid metal tank (except possibly for pressurized foam instead of water). Almost certainly it would be more similar to the collapsible buckets used by helicopters now, just not open to scoop water since that would present a logistical challenge and allow evaporation (undesirable in something meant to sit possibly for years before use).

Now, if there were a reason to fill up from a water source in a remote lake or pond not near the hangar (for example it's much nearer to the fire), I would go with a drone delivered raft with a pump and apparatus with some spare tanks that the drones can keep swapping out. It could stay in a fixed location out of the way of other aircraft taking bucket loads of water and drones could simply keep coming back to it for full tanks and dropping off their empties. Once again though, this would only be supplemental to other firefighting efforts if the initial quick response doesn't stop the fire in its tracks.

Now, such a system might only stop the fire from growing some of the time. Considering the massive amounts of damage some of these fires can do, however, even if it's just 10% of the time, it could be worth it.

As presented, it did seem a bit rigid to me. Now, of course, there may have been leeway in actual practice, which is why I asked about those three scenarios. It's one thing to literally only accept the precise three (four depending on how you count contractions, which I tend to count as two words though the standard seems to be to count them as one) words "I don't know", but another to simply accept the admission that the given answer is just a process to reach an imperfect answer, for example. The original poster may have accepted a broader range of answers than the post suggested. Or maybe not. That's still their right as the one hiring I suppose.

I brought up the example of fairy-tale style tests like getting suitors to choose between a gold cask, a silver cask, and a lead cask where the "correct" answer is, of course, to choose the lead cask since it supposedly indicates a noble lack of desire for worldly things. Of course, there are a lot of problems with such tests from a perspective outside the realm of fairy tales and simplistic morality. For starters, in pretty much all of these tales, the ones giving the test are always filthy rich. So, you have the situation where some king or nobleman or whatever who almost certainly cares a great deal about worldly riches is judging people for their materialism. I mean, it's not really a paradox. It could be simple hypocrisy or even a non-hypocritical desire to actually try to maximize the daughter's happiness. It still comes off a bit judgy given who is making the test though. Then, of course, there's the cynicism aspect I mentioned. Anyone shrewd enough, regardless of greed level, recognizes that the right choice in these classical tests is the least greedy-seeming option. So, such tests are probably more likely to select for the most conniving weasel rather than the most virtuous. Then, of course there's the fundamental flaw of assuming that seeking some sort of material wealth is not virtuous. Clearly, it's normally considered to be virtuous to seek a paycheck to feed, clothe, and house your family, for example. For the example with the casks of different value, how does the test-giver tell the difference between "If I take the gold cask my fortune will grow even more!" and "Wow, think how many starving widows and orphans I can feed when I sell that!"?

Of course, a real world, and frankly tragic (among many pejoratives that could be used), version of that kind of fairy tale simplistic logic is an old standard over whether or not to stone a female rape victim to death. Basically, the way it went is that if she was in the city, and didn't cry out, she would be stoned to death (while presumably in the clear for public stoning at least, but probably facing other social consequences and/or punishment up to and including potentially being killed anyway, if she cried out and also could prove it because someone actually heard it and did not ignore her). In the country, she was clear of the crying out requirement (though still facing potentially terrible consequences). So, these rules follow the simple logic that she is complicit if she does not take specific action to prevent the rape and assuming that she will be heard in the city and not in the country. Nice and simple, and incredibly stupid (not to mention misogynistic, hyper judgmental, and frankly just murderous). Anyone with an iota of sense recognizes some basic problems with that requirement, for example that a person can have their mouth covered, be knocked out, be dragged into a building where screams simply won't reach the outside, be attacked in, or moved to, a place where everyone nearby is basically complicit and does nothing, be attacked in a place where everyone around are "good, decent" people who are still effectively complicit through apathy or self interest and simply ignore cries for help, etc. There's also the obvious knife to the throat/other threat of death with a warning to keep quiet, but for the hyper judgemental, that one is no problem since they would likely conclude that the virtuous should rather die than be raped, so if they didn't let themselves be killed, they are guilty. Same "logic" applies to those who realize that, if they draw attention to the attack, the consequences from society, even without being stoned to death, will be terrible and still may include a shallow grave somewhere. Obviously, for the hyper judgemental, that kind of reasoning is simply not virtuous. So, that's a bit of a ramble, but that's a real world example that's always bothered me tremendously. The "logic" involved is ridiculously rigid and based on a ridiculously simple model of the world. Now, someone would have adjudicated this, of course. Some doing the judging very well might have actually taken all of the real-world circumstances into consideration and made a more reasonable judgement than the stupidity of the actual law (note that when I say "more reasonable", that's still in a paradigm where women are being held criminally liable for being raped, so take with a kilogram of salt, which still would not be enough). There certainly also would have been those judging who simply say "the law is the law" and ignore actual fairness in favor of a rigid interpretation.

So, in a very long-winded way I am saying that overly rigid tests that don't really consider all the possibilities are problematic. Working with a rigid rule, but adapting and making judgement calls may ameliorate things somewhat, but may not make up for the basic problem of the rigidity of the rule in the first place. Basically all behavioral tests to see if someone will do the "right thing" in a manufactured (or poorly modeled) situation are inherently flawed. For the very specific "I don't know" test we're discussing, a lot hinges on how the answer actually gets judged and, ultimately, if the employer wants to apply such a test they can, but they get what they get, for better or for worse.

I've heard that about petty theft too. On the other hand, I was just reading a story about a mechanic who was sick of rims being stolen from cars on his lot and the police not being able to do anything. So he hit a tracker inside the tire and it led the thief who is being prosecuted for about $30K worth of stolen car parts (which he would have made a lot less selling than the appraised value, of course). That's just what they found at the time but it's almost certain that was just the tip of the iceberg. There certainly is petty theft that's just one offs, but there's also plenty where people are basically living off it and do again and again as basically their living. So treating it as one offs that aren't worth it to go after does not make a lot of sense.

You are right of course that law enforcement in the US and globally are very unlikely to actually do anything (either due to apathy or, in some cases due to actually profiting from crime directly or indirectly). The expense argument usually falls flat due to the sheer amount of theft and incidental damage that just one scammer can manage over a lifetime though.

Simultaneously though, section 230 is the only reason sites like Slashdot can even exist in today's world. Can you imagine if Slashdot or any other discussion site were liable for the comments of all of its posters? An all or nothing, throw the baby out with the bathwater approach is not the way to go here. Section 230 might need to be more thought out so that it can simultaneously provide some protection from/liability for scams and, regardless of section 230, the scammers themselves need to be caught and prosecuted more, but just getting rid of section 230 suddenly gives pretty much the whole Internet a massive barrier to entry that favors the tech giants with powerful legal departments.

Post editing is a bit of a tricky proposition. It can certainly cause a lot of confusion if a chat system allows editing posts that have already been replied to (or are in the process of being replied to since someone might be composing a lengthy reply that makes no sense if the edit comes before they post). Generally you should at least have notifications on all replies that the post they are replying to was edited and probably a way to see the original post.

What might be best is some sort of grace period during which you can edit as long as no-one has replied yet (and anyone who posts a reply within a certain time span gets a warning that the original was edited). Plus there should always be a notification that the post was edited from the original.

As far as such features on Slashdot... That's not really likely any time soon. Honestly I think that would annoy more people here than the people it would satisfy. At best maybe they might someday add a mechanism that allows errata follow up posts like yours to be embedded at the end of the original post, rather than being a new reply.

How would you have responded to something along the lines of "That's not something I know offhand but..." followed by an approximation method and an approximate answer? You stated that you wanted to hear "I don't know" and would sometimes ask how they would approximate it, but what if the admission of ignorance to the actual facts came along with an attempt to answer anyway?

Alternately, what if the admission of ignorance was less explicit and along the lines of "I can find out, but I don't know off the top of my head. That's the sort of thing I would normally consult reference materials for."?

Also, alternately, what if someone actually had a reasonably correct answer to the question? Did you actually bother to find out before you answered the question? What if they had answered "somewhere between 100K and 130K from what I last read" and they had read it in a newspaper article a few years ago? Would they have been wrong because they had not said "I don't know" because they didn't have perfect precision and up to the second knowledge?

Ultimately though, it sounds like this is a trick question really. While the guy who ranted and raved about it seems to have been out of line for a job interview, it certainly can be frustrating to deal with people evaluating you with trick questions with one "right answer" that the evaluator has decided on. There's a game called Mindtrap that my girlfriend will not play with me any more (and is a serious sore spot for her) because so many of its questions are riddle-like questions that actually have multiple possible answers that are absolutely correct and fit the question, but there's only one "right" answer that the game makers decided on. Or, then there's that one true/false question that I got in French class decades ago that went something like "Si vous voulez maigrir, mangez-vous des pommes de terre ?" I literally went up to the French teacher and translated the question to English: "If you want to lose weight, do you eat potatoes?" and asked her what the answer was and she just scolded me for asking for the answer for a test question. It was a _French_ class, not a diet class. The answer isn't even objective, you won't get a simple yes or no answer from most dieticians. I know I guessed the answer wrong. I can't even quite remember if the "correct" answer was true or false. I think it was false.

Anyway, the point is, I probably would have failed your interview (at least back then) because I would have taken the question honestly at face value and I would have tried to use all of the information at my disposal to come up with as close an estimated range as I could. I would have explained my work and not claimed that I somehow knew the exact number, but been clear on what I did know that lead to the answer. I probably would not have explicitly said "I don't know" because I didn't understand about interview trick questions back then.

These days, I'm a lot more cynical. I know now that if someone presents you with a gold cask, a silver cask, and a lead cask, you choose the lead cask because it's some bull*(%$ test of your "moral worth" or something like that. If you get asked to carry a glass of water without spilling a single drop, you drink it because it's also some bull*(%$ where drinking it somehow doesn't count as "spilling" it. If you get a multiple choice question with multiple answers that are actually right, you pick the answer that you guess the test-makers will have chosen as "right" (learned that I usually got a high rate of success if I chose the answer that I, personally, found to be the dumbest). If you get a question on a school test and you know a correct answer that you got from an article in Nature or another scientific publication, don't use that answer since it's not "right" even if it's true. Etc., etc.

Ultimately I think that may be a potential concern for your test question. Back when I was less cynical and more earnest, I would have almost certainly given the "wrong" answer, not really understanding that the question was a deception. These days, I would probably recognize the deception and picked the right "I don't know" (which from my way of looking at things is a dishonest answer since it's too absolute) because I would be deceptive in return.

That, of course, is just my personal philosophy on the matter. As someone doing the hiring, you can filter for whatever you want. For example, if you ask someone "I had a broom consisting of a handle and bristles, I replaced the bristles and later replaced the handle, when did it stop being the same broom?" you might get answers like "when you replaced the bristles", when you replaced the handle", or even "it never stopped being the same broom", even though the question itself implies that it stopped being the same broom at some point. You could even get a philosophical answer about the broom's identity never being truth in the first place, etc. You could choose any of those answers as the "right" answer and screen candidates by that depending on if they meet your own personal philosophy. So, that's your right I suppose, but not everyone is going to agree with your choice and it seems pretty subjective whether that's the best way to do it.

Certainly they typically use collapsible buckets in traditional firefighting. The idea of tanks was an alternative to the potential dangers and logistical issues you mentioned of an autonomous aircraft dipping one of those buckets into a lake or pond. I was considering a system where the autonomous craft that can sit in a remote hangar for years is either preloaded with water (or firefighting foam) where a tank would make more sense than a bucket since the contents would be prone to evaporation or even be under pressure. In such cases, either the autonomous aircraft would only be used once in an attempt to provide the quickest response possible and head off the fire before it even really starts, or it would need a system of reloading that does not involve dipping a bucket into a water source. For that, having a spot on the ground somewhere that it can drop off empty tanks and pick up a loaded one with the empty then being reloaded seems to make the most sense.

Exactly! Glad we got this all cleared up. :)

Literally from the GP comment that you replied to complaining that they did not subtract the electricity used to charge the batteries. It's the source you were literally complaining about. I am making no claim that their numbers were correct. It was clearly back of the envelope stuff. All I am saying is that they presented a model and some numbers and you specifically claimed that they left out a calculation that is actually unnecessary in the model they presented. You are free to present a better model, better numbers, etc. but as far as your specific complaint about needing to subtract the electricity needed to charge the batteries, you're simply wrong.

Why? They came up with 2500 MWh produced and $80/MWh with no difference in price for electricity used directly and electricity used from the battery. There is also 1200 MWh capacity on the battery storage.
Consider:
2500 MWh - 1200 MWh = 1300 MWh during the day and 1200 MWh in battery storage. If you sell that at $80/MWh, that's $104K during the day and $96K at night from the battery, which adds up to $200K.

If you store only 600 MWh from the battery during the day that's:
2500 MWh - 600 MWh = 1900 MWh during the day which is $152K and then $48K during the night from battery. That adds up to... $200K again.

If you store nothing during the day that's:
2500 MWH - 0 MWh = 2500 MWh during the day, which is $200k and then $0 during the night from battery. That adds up, curiously again, to $200K.

Now, I get that there are more complicated models for this where there are different day and night rates and there are some losses from storing power to battery and using it later versus using it right at once, etc. However this was presented as a back of the envelope calculation with ballpark figures and those factors would pretty much just be a rounding error in such a thing. Otherwise, if we're just looking at a simple model of a set amount of electricity per day and a set price, there is no point in subtracting the electricity used during the day to charge the battery. Just like there is no point in subtracting inventory in a 24 hour store that you keep in a back room and restock around evening and then sell overnight.

There is, of course, overcapacity as a consideration. Do you end up producing more electricity than you can sell? However, that's not really a consideration of the model at this point either. The assumption is that the amount of electricity generated is close enough to the amount sold that the difference is pretty negligible.

No actual disagreement with that as a general statement, but the GP pointed out just how low the usage actually is. Statements like the above have to have a point of diminishing return standard applied to them. With just about any effort you have to recognize that the last tiny bit may require a lot more effort or difficulty to achieve. As an example, consider the goal of removing 100% of the water from alcohol. You can approach that goal with multiple rounds of distillation, but the azeotropic limit will still leave at least 4.4% water no matter how much distillation you do (and it will still be almost impossible to reach even that just through distillation. There are other methods to go beyond that, but it gets more and more complicated.

So, for any such effort, there's a point you can label as good enough. That does not mean that you actually completely stop trying, just that if you have a list where you plan to complete one task before another, it does not make a lot of sense to pour massive amounts of effort into completing the last few percent of the first task before picking the low hanging fruit of the second task... I can use a fuller fruit picking analogy there, in fact. It does not make a lot of sense to let all your pears rot because you have your dozens of fruit pickers still working in the apple orchard because every few hours they manage to find three or four apples that were missed the first time.

While your point is quite correct, technically speaking they are also correct. That "annually" is not actually wrong, it's just completely pointless, meaningless, and extraneous.