A: it's less likely to get injured in an impact B: as weird as this sounds, mechanical hard drives with spinning discs don't work well at high altitudes, like Everest Base Camp. Apparently many hard drives fail at over roughly 3500 meters altitude. With that said, none of my computers or apple ipod/creative zen have had troubles with extended operation -- several days at a time -- at 11,000' elevation, and it's not a problem if they're not running. (I didn't previously know the hard drive cases were vented to atmosphere, although I guess it makes sense.)
I'll try and give an answer, but I think it's pretty complex. What it boils down to is that many/most people think drinking isn't okay but it's unstoppable because we've been doing it for 3000 years. As a country we did try and stop it in the 1920's. Pot is generally regarded as stoppable because it requires actual growing, and it doesn't grow wild in many places in the US.
A more interesting approach to this whole problem is depressants vs. stimulants: many people have claimed that the transition from the late Middle Ages to the Industrial revolution was actually the transition from massive use of alcohol -- everyone drank beer all the time -- to the massive use of caffeine, as tea or coffee. Once governments figured this out, they promoted use of stimulants -- nicotine cigarettes given out freely by governments to armed forces, likewise methamphetamines in many armies, because of the increase in productivity. So you see a society where massive use of mind-altering drugs that make people more motivated, is encouraged in an organized manner, while use of depressants is discouraged. Pot and opiates are strongly prosecuted, alcohol production and distribution is regulated and discouraged. If there weren't so much money to be made, I think it'd probably be even more discouraged.
You want a tiny bit of play because if you just crank it down blindly, there are circumstances where non-equilibrium heating means you have negative clearance and things either bend or one valve doesn't close/seat fully -- both of which are *bad*. That's why they spec valve clearance as a range, with a non-zero least-clearance, on every valve train I've ever worked on.
On the Ford, I was disinclined to do anything but follow the directions carefully, since the exhaust valves were tubular and filled with liquid sodium metal (well, liquid once the engine was hot, at least), and any sort of mistake could lead to serious excitement, like, say, a sodium fire in the engine compartment. On the '71 Datsun, however, I couldn't care less, because I bought it for $200, so I did the valve lash hot once, let the car cool down, wrote down all the clearances when it was cold, and from then on just relashed it to those numbers rather than the quoted hot numbers.
(Why non-fully-seated valves are *bad* -- if it's the intake valve that sticks down, combustion leaks into the intake manifold and carburetor, which eats all the fuel/air mixture for the whole engine and is rough on the air filter, and for either valve but particularly the exhaust valve, the valve seat wicks away a lot of the heat the valve's picked up so you risk burning the valve edges. And lemme tell you what, a burnt valve that ends up fracturing does really rotten things to an engine.)
>I think the problem is that Operating Systems like Windows have to be designed with a wide user base in mind, so they have to have features that only 10% of the users would use.
Sort of. It's more that Windows is designed with a wider base of user *experience* in mind -- they hand you everything and you use it. A la carte, the *nix way, is great if the user knows enough to go decide what's needed. My linux system can load drivers for stuff Windows has never heard of: Amiga file system management, USB-to-serial IC's. But 90% of the people who use computers will never need any of that, so the Windows system of one-package-to-rule-them-all, one-package-to-bind-them works great. But just try to get support or drivers working on Windows for any hardware that's not sold at Best Buy. (I bought a Philips webcam a while back. It works with Windows98. There is no other version of Windows that can work with it. But a tiny bit of tweaking and my linux systems, one from 9 years ago and one brand-new, could both handle it.)
>I'm not a mechanic, and I have little to no experience under the hood, but are a lot of cars really designed this poorly?
Other people have already talked about the specific case of the battery behind the wheel. Things I've seen on cars I've worked on: having to remove the wheel to change the oil filter, on a Saturn; having to remove part of the power steering booster to change the rearmost spark plug, on an Oldsmobile; and having to wrap the CV boots with plastic bags before removing the oil filter so it doesn't drip on them and dissolve the rubber seals, on a Subaru. I've been told that on some rear-engine Porsches you had to remove the engine to change the spark plugs, and on some '85-90 Corvettes you had to remove part of the intake manifold to change the spark plugs. On my dad's '64 Ford, there were no hydraulic lifters, so every 3000 miles or thereabouts, I had to relash the valves -- manually adjust for the wear in the valve train. I had to do that on my '84 Nissan, actually, but then all the clearances were quoted cold, so that wasn't too bad. On my '71 Datsun, they were quoted hot, so you'd run the engine, then quickly pull off the valve cover and start measuring clearances between really hot pieces of metal, trying to adjust them accurately. But the '64 Ford was the king of annoyance, because the adjustment was specified WHILE THE ENGINE WAS RUNNING. You want a bad time: try adjusting a nice hot threaded bolt with a locknut, while it's jerking through about 15 degrees of movement 400 times a minute, while hot oil is spraying out of the valve train lubrication lines, and you have to feed a feeler gauge between the bottom of the bolt and the top of the pushrod during the brief moment they're not in contact. Oh, and the cam was sufficiently aggressive that at idle the car was continuously backfiring through the carburetor so there were occasional blasts of flame from right in front of you. Compared to that, what's a little hassle like removing a wheel to replace the battery? I was so glad to see that car go, even if it did have the hottest engine Detroit ever made.
>First, vehicle information is very proprietary. Why is it that cars can't report status information via a simple USB connection?
Build yourself one of these. It's an OBDII-to-USB converter. It still requires *extensive* software on the computer side, but you're already talking about having that. On-board vehicle diagnostics are fairly complicated, but there are plenty of programs that handle it, many for free.
I agree it'd be nice to have sensors to detect fluid levels... but until sensors are more reliable, you might start relying on something that has broken and end up in trouble. In my car, the wiper washers are plumbed such that when you're just about out of washer fluid, the rear window pump stops working, which is clever and avoids having to add an extra sensor.
It would be nice to see more electronics that used USB connectors to get power, if they can stay under half an amp. It'd also be nice to see more wall-warts with some sort of load-detection circuitry and a solid-state relay so if they're not being used, they disconnect the transformer/switching PS from the AC so it doesn't sit there sucking up power.
I don't know how much you know about transcription/translation, so I'll do a very quick summary. DNA -> RNA, a very closely related polymeric molecule. The RNA then gets translated by transfer RNA molecules -- themselves made of RNA (which is why we think RNA might've been the original stuff of life, because it can both hold information and perform transformational changes to other molecules.) The transfer RNA molecules, tRNA, basically have three RNA's sticking off one side, that are complementary to the messenger RNA produced from the DNA template (a mirror image of a mirror image reproduces the original image) and on the other side a single protein for which they code. Protein assembly consists of linking those proteins carried by the tRNA together. So think of this as two systems that interact: DNA is the memory, protein is the implementation, and the tRNA is what translates from the memory to the implementation. We could presumably figure out a way to hijack this if we could build enzymes that built different tRNA molecules: A: by putting on different, 'unnatural' proteins onto the tRNA's, which would allow us to make enzymes that could do things no current enzymes can do. (To some extent this already happens: bacteria produce/use weird proteins in their cell walls, including proteins that are mirror-images of the sort of proteins seen throughout the rest of the living world. But the possibilities are much greater than that, because enzymes are thermodynamic factories that can do amazing chemistry if you can just find a way to judo chemicals A and B into chemical C by building an enzyme with exactly the right surface topology to make A and B stick together the way you want them to so they turn into C, and being able to supply extra electron density right where you want it, rather than just having to settle with the current 20-something proteins we have, would massively help in such engineering.) B: by putting on different, 'unnatural' RNA's on the other end of the tRNA, to match to engineered DNA, so you could potentially replace the entire DNA/RNA system with something that works better (fsvo 'better'). The thing that's nice about this is if you can mess with the tRNA, you can use the existing DNA toolchain to build weird proteins, or use the existing protein toolchain for playing with different DNA.
This research is an early step down that road: finding engineered DNA that can work like existing DNA. They're reverse-engineering, which will presumably allow them to find what parts of a base nucleotide you can change, while retaining the ability to use the cell's existing toolchain.
I'm not in this field anymore so bear in mind my knowledge might be very dated. Generally speaking, viruses that insert their DNA into eukaryotic DNA don't have a particular place that they do so: they get their DNA into the cell, and it then inserts itself randomly in some bit of exposed DNA. See, eukaryotic DNA is very tightly bound to accessory proteins that protect/maintain it and hold it in some sort of to my knowledge poorly understood large-scale organizational scheme that constitutes a chromosome, so it's not like you can get to just anywhere on the DNA, but the parts you CAN get to might not be consistent, depending on whether that particular DNA bit is being transcribed at the moment, or repaired, or what have you. So what happens is that viruses stick their DNA all *over* the place, and the vast majority of them are indeed in no-ops or unread/untranscribed sections and just sit there -- which is where all the endogenous retrovirus stuff we read about comes from. Complete replication of the DNA is rare -- it only happens when the cell needs to divide for some reason. Small-scale scanning and replication is very common because a cell's day-to-day enzyme turnover requires it. But that small scanning is less likely to hit the area where the virus DNA is because of the sheer size of the genome. Eventually it'd be nice to migrate to a whole different genetic code and support enzymes, because then viruses would be instantly nonviable, but that's a long, long ways off. However, this research is the first step: if we had a wholly different DNA, and re-engineered the enzymes that make transfer RNA, which convert the messages read from DNA into protein, we could retain all the protein-handling enzymes we have, and everything they make, and 'only' swap out the DNA and RNA suite. That's still an enormous problem, but it's like 0.01% of the problem of trying to engineer new proteins and a whole metabolism based on them.
This could be really useful in the long-term: if we could substitute replacement codons that work with most of our existing DNA, it's one step to building really tough DNA. Right now, there are a lot of damage mechanisms like adjacent thymines linking resulting from exposure to chemicals or shortwave radiation, and replacement codons engineered to not be suseptible to these could make, say, protracted exposure to radiation outside the Earth's protective atmosphere more viable. Of course, then we'd have to engineer a whole set of enzymes to synthesize those new codons, which is an extremely hard project, but finding things that work as replacement base pairs, now, gives us time to study how they might fail and figure out what the best candidates are.
Well, ideally, a projectile would be shaped like a zep, right? A toroid with a teardrop cross-section would be better than just shooting bearing housings, but a *good* system for launching things would launch minimized-drag objects and no toroid really can compete on that basis. But this system can only shoot toroids. What I'd like to build is a coilgun where the primary and core are wrapped around the projectile, so it doesn't need a hole through it, but I haven't gotten those to work very well -- they seem to work much better using a single massive pulse of DC, like a railgun. Still working on that... The other problem is that with both systems you want your mass to be as close to your core as possible to maximize the energy transfer efficiency, by avoiding the air gap loss, and any good aerodynamic shape is going to have lots of air gap loss since it only reaches its max width at one point. Frustrating when the laws of physics dictate contradictory shapes.
>If everyone would just follow them, then the world would run like a well-oiled machine.
And the thing is: they're right. The world WOULD run great if everyone just did exactly what they did. But it would also run great if everyone did something totally different yet. The problem is that there are lots of people doing lots of different things, differently, for what they consider perfectly good reasons. I think the last bit of that is critically important, though: everyone, even the idiots who do $something_I_hate, think they're doing the right thing. It's that little jump that's difficult for fundamentalists because they don't think someone else can have a justification or rationalization that is both different and valid. That's where a lot of the world's conflict comes from. Engineers try and argue you into believing what they do, because they're convinced they're right and they can change your mind. Religious nutjobs threaten or use violence to get you to believe what they do because they're convinced they're right and they insist you change your mind. (Which is why I prefer engineers. That and very few of them believe that God is on their side. Well, okay, actually they do: they just call God 'logic'. But that's less hostile.)
The other thing about engineers is that I think it's easier for people who are led to engineering to get sucked into a project without considering all the ramifications of the result. Like the movie Real Genius, they can get so worked up building something that they overlook or refuse to consider its dangers: a consequence of being highly focussed. Similarly fundamentalists, who can go from "this is for the Good of all mankind" and get wound up in the goal, and justify doing awful things to get to an end they think is worthwhile or necessary. It's not like determination, or focus are bad things: it's just that people who have that can be a great deal more constructive or destructive than people who spend all their time worrying about the consequences of their actions.
A previous poster pointed at Engineer's Syndrome, and I see some similar tendencies.
Engineers -- and I'm speaking as someone who is doing an engineering job, surrounded by engineers, and from a family of engineers -- tend to favor experience more than empathy. They tend to think that if they're convinced something is right, it's for good reason, and once they're convinced, it takes some work to change their minds. More particularly, if they're convinced, they're unlikely to use someone else's experience as a guideline: they're less likely to put themselves in someone else's shoes to regard a problem from that standpoint.
My own definition of Engineer Syndrome is encapsulated in the phrase, that I actually heard from one of my dad's coworkers once, "If you would've thought about this problem as much as I have, you'd agree with me." The level of premise and and patronization enclosed in that one sentence is staggering, but when it comes right down to it, I think many people drawn to engineering feel that way at some point or another. The consequence of this is that if someone else *doesn't* agree, the person suffering from ES thinks the other person is either stupid or stubbornly wrong, and either way, is a fool whose opinion is not to be regarded.
Likewise, engineers come from a background where things are provably correct (mathematics) or experimentally verifiable (most of the rest of science and engineering) and take that sense of certainty and apply it in areas where it isn't applicable -- sociology, politics, art, places where it really does come down to opinion, where there isn't actually a right and wrong, just preference.
The fundamental difference is that engineers do tend to rely on things that are provably correct or experimentally verifiable, whereas religious extremists are predicating invisible omnipotent entities. But the point is: if you have people who have this engineering set of mechanisms and filters for dealing with the world, and who believe in invisible omnipotent entities, they're going to have similar behavior to people who are drawn to engineering.
Mine are nowhere *near* supersonic and are comparatively unstable in flight. Even if they weren't, the idea projectile shape for drag reduction is basically like a Zeppelin, definitely not a toroid. Unfortunately, this design relies on a toroid as the shorted secondary winding of a linear motor, basically, so I'm stuck with it. That's okay: it'll still go through a sheet of drywall.
Batteries are optional. I've made a couple coilguns out of a christmas paper wrapping tube filled with welding rod (cut to different lengths so it's solidly packed at the bottom tapering to 50% fill or so at the top) with a big coil of magnet wire around the bottom, that's attached directly to 110V AC (with a relay in there to turn it on/off.) The drag is it only fires ring-shaped projectiles -- but boy do they fly. Some time it'd be nice to build one that launches spheres rather than toroids, but that seems to require a lot more work.
Not exactly the same, but in some ways more impressive: the company my now-ex-gf was working for, designed and built an ultrasonic imaging device that they could feed into your femoral artery and snake up into your beating heart to image the insides real-time. No cracking the chest open, no shutting the heart down and rerouting the blood through an external pump.
It gets better. They could click the ultrasound transducer into high power mode and selectively kill small sections of the heart that were beating incorrectly -- were hindering, rather than helping, the contraction of the heart, which apparently happens when scar tissue from a previous heart attack reduces or blocks the nerve signal propagation across the heart muscle -- so they could do actual surgery to correct heart malfunctions, with nothing more than a tiny cut on your leg.
Her job was plating quartz substrate with layers of copper and gold, then cutting it into tiny sections that acted as the piezo elements for the transducer. She then hand-soldered wires onto each of the hundred-some elements, using wire about the diameter of a human hair, because the whole transducer was like 2mm in diameter and only a couple times that long. It was amazing. (So was the plating process, which involved solid chunks of copper and gold that weighed a kilo or more, each one. They were pretty careful about the gold ones, but let me have several of the used copper ones, so I have table coasters way cooler than either the infamous AOL cd's or hard drive discs: massive machined chunks of copper with one side plasma-etched away.)
I was at a small, but unaccountably well-connected college, sending email to my dad, who was working at a big tech company. Usually I just copied whatever route he'd chosen, and then marvelled as my mail got to him in LESS THAN TWO HOURS -- the very idea! At the time, it was still pretty unusual that two people would both have access to email, so I actually showed off to my friends -- "hey look at THIS!" Well, one of my friends knew more than me, so he taught me about uuhosts -- a way to find out what was connected, for the times when my email was just vanishing because something, somewhere, was offline. So I used it. The next day I got some Very Crabby Email from a sysop who tore me a new one for using a satellite uplink to send personal email to Japan and back.
It felt like having a switchboard operator yell at me. I was *mortified* and I didn't even know for sure what I'd done.
It's possible I'm wrong, but that's what I've been told by coworkers who design bandgap references. I don't do chip design or know that much about it: I know they're heated, but coworkers indicated they were actively cooled as well, if necessary.
Peltier devices on-chip have been used for a while, whenever temperature variations are intolerable. Some examples: Analog Devices AD595 thermocouple amp, which uses in-chip thermal calibration to ensure a cold junction of known temperature, and many voltage regulators and switching supply controllers that use temperature-controlled bandgaps as their voltage reference.
'bout 1/2, if that. *slow*. The reason American cars have so much power is because mechanical friction increases linearly with speed, but the power required increases as the cube of speed. (Air resistance increases as the square, but power as the cube, before someone jumps in to correct me: force rises as the square, as a result of air resistance, and power as the cube.) So the point of having big engines isn't to climb, it's to climb *faster* or to have a higher top speed. Most cars cruising at 100 km are only using like 20 hp, if that. As an ex-bike-racer I can beat most cars at stoplights, but once we get above 45-50kph they leave me behind.
I agree there are some pretty steep hills in Vermont. Likewise in Seattle, San Francisco, and where I'm at, the mountains of Colorado. But according to various departments of transportation and everything I've managed to find and measure, Baldwin is the steepest paved street in the world. Those pictures don't really portray just how steep it was. It's hard to convey the feeling of riding up something that steep, and I ride my mountain bike up stairs -- multiple flights of stairs. It's REALLY steep. Although there are some little roads in Wales that are just terrifying: you come over the top of a hill and you literally cannot see the road because it drops off so steeply it's obscured by the hood of your car, and that's in a tiny economy car.
>I think this is going a bit far. Not everything is useless junk after all. Some people load up drywall and other building supplies for home improvement projects. Also, having some cars that can only go 4 MPH uphill on an interstate is asking for trouble. Do you really think the car gets 56 MPG if it has a top speed of 4 MPH going up an incline?
People who want to haul drywall will hopefully have enough sense to buy something else: this is so not a car for a single-car family. Many of my coworkers own tiny economy cars or one of a half-dozen hybrids parked out front, and at home they have a big ol' pickup that they drive when they need drywall. What's scary is that they actually save enough on gas in the tiny car to pay for a '70's pickup. It's a nice idea: they recycle old cars, basically, and use the appropriate car for the job, and it doesn't actually cost much (except for that whole buying-a-new-hybrid thing) but it does take up a lot of space.
I'm *sure* that car doesn't get anywhere near 56mpg on climbs. Then again, mine doesn't either: I get about 30mpg on average, but more like 5mpg on climbs. When I drive up to my grandparents' place, which is at about 11,000 feet elevation, I get like 15 mpg, and on the way back home I get like 50, because I'm coasting half the way. The thing is: even floored, throttle wide open, an engine that small will only suck up maybe 4 gallons per hour, if that. I'm basing that on how much fuel a VW engine at 3000 rpm uses, in VW-powered aircraft: this engine will be racked out at more like 6000 rpm on a climb, but it's only half the size of a bored & stroked VW. A big ol' Hemi 440, in contrast, can slurp over 30 gallons an hour at 4000 rpm on a climb.
This isn't the right car for a family that lives in the country and only has one car, or a family with lots of kids, and probably not even for a family that lives in a hilly area like yours. But as a single-person commuter car in urban/suburban environments, this would work as well as many other crappy little econoboxes that are selling well in lower-income demographics.
>Besides the car was woefully underpowered requiring near lead footing to use on the highway...
I was following a SMARTcar the other day on my way to work. It was doing 65mph up a steep hill at 6500 feet elevation. It might not have as much power as a Silverado: I wouldn't want to haul boat trailers with it. But it had more power than it needed for more difficult driving conditions than 90% of them will ever see.
While I enjoy your comment, I'd like to point out that 30 hp is a lot of power. I managed to get up Baldwin Street on a bicycle -- that's 2/3 horsepower, and there aren't steeper streets in Vermont or anywhere else. Granted, the bike and I, together, only weigh 70 kilos, but with 45x more power, and lower gearing, it'd do just fine... at the expense of speed. One of my friends has a Pinzgauer, that weighs 3000 kilos and can haul 14 people through waist-deep water full of rocks and up a muddy slope on the other side. It has a 65 horsepower engine. Any amount of power can get you up any hill if you have low enough gearing. So the problem isn't that it can't go up hills with heavy loads, it's that fatass American families won't buy a car that can only do 4 mph up the hill to their home when they've packed the back with useless junk they bought at the mall and stuck their enormous selves in every seat.
If you're thinking about trying this, do a single-layer board with a dozen BGA sites and find yourself some cheap BGA's to try and practice. It's a bear. You can remove screwed-up ones the same way you put them on: one heat gun on the bottom on low and one on top on high. Again, it's a bear, and carefully removing all the solder with solderwick from every bga pad is slow and demanding. The main problems we see are: ground plane connectivity with the bga bump, because the ground plane pad isn't warm enough to reflow well so you get intermittents (boards that work only when you push down on the bga) which we can sometimes solve with flux and lots of heat and, as I said, poking at the chip to make sure it's bouncing rather than rocking... which leads to shorts between adjacent BGA bumps. What we've done is to figure out -- on our microSMD's -- a way to tell if there are shorts by continuity testing between points elsewhere on the board. It would be a lot easier if you're doing your own layout: use via-in-pad or adjacent to pad (much cheaper) on every pad, and then you can check continuity from the bottom side and see if there are shorts before you light it up. I can tell you from personal experience that one serious short will end up delaminating the pads from the substrate, so when you try and remove the BGA you rip off all the pads adjacent to the short as well, scrapping the board, unless you are a true wizard of pad-and-trace repair.
Another thing I've done in an emergency -- a unique chip, a customer failure that we had no option but to get working to replicate the failure -- was to cut off small chips of solder of known length, with a fixture involving a pair of nail clippers nailed to a board so I could feed solder through the clippers until it hit a stop, clip, and repeat, to get exactly the same length of solder each time. Then I'd touch each bit of solder, lift it, and touch it to a pad on the bottom of the chip, approximating rebumping the chip (since we have no facilities for doing that the right way, with a mask and precalibrated solderballs that are gently reflowed.) It worked, after two tries, but boy was that unpleasant, high-stress work. The point being: you can reuse both the boards and the chips if you're very careful, spend a lot of time, and do a lot of checking before powering anything up.
>Such a board is possible (and relatively easy) to design by an amateur - but very expensive to make as it would have to be 6 layers and require soldering BGA chips.
I have a few tips from personal experience. You can get multilayer boards built fairly inexpensively if you can justify having four made at one time: you might be looking at under $80/board for a 6-layer (although I'm not positive about that. I know you can get 4 layer done for under $60/board.) It's possible, although unpleasant, to reflow your own BGA's. You need a microscope with a tilt-head. Draw the BGA package outline in the layout software as a silkscreen, making sure it's at least as large as the actual package, or even better, draw several outlines of increasing size. Align the BGA visually within the closest package size, double-check by looking at the edge with bright illumination and a microscope to make sure you're basically on-pad, then gently reflow it down with a heat gun. It works best if you can preheat the board from the bottom with one heat gun on low, then do the reflow from the top with the second one. I'm doing this at work with microSMD, which are way, way smaller than BGA -- chips 3mm on a side with 12 bumps on the bottom. After a bit of work I have a 70% success rate. The main thing I've found is that while you're reflowing, you'll see the chip move as the capillary action of the solder pulls it into place. Very, very lightly touch the chip on one edge with a probe. If it rocks, the center isn't yet reflowed and it's pivoting on the as-of-yet-solid bumps. When the whole chip bounces like a spring on all the melted bumps, rather than rocking, then it should be good.
>The rationale for having me boot my computer apparently was that it may be a bomb, not that my contents might be suspicious. The logic of having me sit in front of them and power on a bomb just to find out if it is, in fact, a bomb still escapes me to this day.
Do you remember that old Daffy Duck cartoon where he's checking ammunition by hitting it with a hammer and if it doesn't go off, he writes 'dud' on it? Maybe you've just discovered why the TSA is hiring so many people and advertising at local community colleges for job applicants.
That's why I said 'correlate' in the last sentence -- I don't think that treating the weak better is the reason civilization exists. I think civilization exists because cooperative group effort makes that group outcompete other groups with less cooperation. As such, civilization is a competitive evolutionary strategy: it exists, and expands, because people who do it survive longer and can better care for their children. However, since it is a cooperative group effort, I think that over time giving more power to weaker members is an emergent quality, although that might have to do more with group psychology and the (mostly human, only sometimes observed in other animals) tendency to try and force life to be fair that we exhibit. In short, people want life to be fair becausse it makes them feel more in control of their futures, and they will pay, in labor or money, to try and make it fair. Since humans have empathy, we will as a group do the same thing for the weaker members of the group and over time pull them into the group. It's weird behavior, frankly, but we've been doing it fairly consistently for all of human history, and it is, well, nice. So, more power to us as a race: we're not unrelentingly horrible.
They're using the sodium chloride as a thermal reservoir -- heating it and relying on its high temperature to make up for its so-so specific heat. Water's specific heat isn't much different, but it's difficult to contain as steam. So they heat up the salt -- or anything else -- and let it gradually cool down, extracting heat from it by vaporizing water and reclaiming the energy through turbines. That way they can produce power all night off the heat saved during the day. It's not a bad idea if they have a good insulated container for the molten salt. It introduces a lot of waste because of the cumulative inefficiency of heat transfer between the different systems, but it allows a system based on this to provide more reliable energy -- energy that's closer to being on-demand, rather than just when the sun is shining strongly enough.
Slashdot Mirror
A: it's less likely to get injured in an impact
B: as weird as this sounds, mechanical hard drives with spinning discs don't work well at high altitudes, like Everest Base Camp. Apparently many hard drives fail at over roughly 3500 meters altitude. With that said, none of my computers or apple ipod/creative zen have had troubles with extended operation -- several days at a time -- at 11,000' elevation, and it's not a problem if they're not running. (I didn't previously know the hard drive cases were vented to atmosphere, although I guess it makes sense.)
I'll try and give an answer, but I think it's pretty complex.
What it boils down to is that many/most people think drinking isn't okay but it's unstoppable because we've been doing it for 3000 years. As a country we did try and stop it in the 1920's.
Pot is generally regarded as stoppable because it requires actual growing, and it doesn't grow wild in many places in the US.
A more interesting approach to this whole problem is depressants vs. stimulants: many people have claimed that the transition from the late Middle Ages to the Industrial revolution was actually the transition from massive use of alcohol -- everyone drank beer all the time -- to the massive use of caffeine, as tea or coffee. Once governments figured this out, they promoted use of stimulants -- nicotine cigarettes given out freely by governments to armed forces, likewise methamphetamines in many armies, because of the increase in productivity. So you see a society where massive use of mind-altering drugs that make people more motivated, is encouraged in an organized manner, while use of depressants is discouraged. Pot and opiates are strongly prosecuted, alcohol production and distribution is regulated and discouraged. If there weren't so much money to be made, I think it'd probably be even more discouraged.
You want a tiny bit of play because if you just crank it down blindly, there are circumstances where non-equilibrium heating means you have negative clearance and things either bend or one valve doesn't close/seat fully -- both of which are *bad*. That's why they spec valve clearance as a range, with a non-zero least-clearance, on every valve train I've ever worked on.
On the Ford, I was disinclined to do anything but follow the directions carefully, since the exhaust valves were tubular and filled with liquid sodium metal (well, liquid once the engine was hot, at least), and any sort of mistake could lead to serious excitement, like, say, a sodium fire in the engine compartment. On the '71 Datsun, however, I couldn't care less, because I bought it for $200, so I did the valve lash hot once, let the car cool down, wrote down all the clearances when it was cold, and from then on just relashed it to those numbers rather than the quoted hot numbers.
(Why non-fully-seated valves are *bad* -- if it's the intake valve that sticks down, combustion leaks into the intake manifold and carburetor, which eats all the fuel/air mixture for the whole engine and is rough on the air filter, and for either valve but particularly the exhaust valve, the valve seat wicks away a lot of the heat the valve's picked up so you risk burning the valve edges. And lemme tell you what, a burnt valve that ends up fracturing does really rotten things to an engine.)
>I think the problem is that Operating Systems like Windows have to be designed with a wide user base in mind, so they have to have features that only 10% of the users would use.
Sort of.
It's more that Windows is designed with a wider base of user *experience* in mind -- they hand you everything and you use it. A la carte, the *nix way, is great if the user knows enough to go decide what's needed. My linux system can load drivers for stuff Windows has never heard of: Amiga file system management, USB-to-serial IC's. But 90% of the people who use computers will never need any of that, so the Windows system of one-package-to-rule-them-all, one-package-to-bind-them works great. But just try to get support or drivers working on Windows for any hardware that's not sold at Best Buy. (I bought a Philips webcam a while back. It works with Windows98. There is no other version of Windows that can work with it. But a tiny bit of tweaking and my linux systems, one from 9 years ago and one brand-new, could both handle it.)
>I'm not a mechanic, and I have little to no experience under the hood, but are a lot of cars really designed this poorly?
Other people have already talked about the specific case of the battery behind the wheel. Things I've seen on cars I've worked on: having to remove the wheel to change the oil filter, on a Saturn; having to remove part of the power steering booster to change the rearmost spark plug, on an Oldsmobile; and having to wrap the CV boots with plastic bags before removing the oil filter so it doesn't drip on them and dissolve the rubber seals, on a Subaru. I've been told that on some rear-engine Porsches you had to remove the engine to change the spark plugs, and on some '85-90 Corvettes you had to remove part of the intake manifold to change the spark plugs. On my dad's '64 Ford, there were no hydraulic lifters, so every 3000 miles or thereabouts, I had to relash the valves -- manually adjust for the wear in the valve train. I had to do that on my '84 Nissan, actually, but then all the clearances were quoted cold, so that wasn't too bad. On my '71 Datsun, they were quoted hot, so you'd run the engine, then quickly pull off the valve cover and start measuring clearances between really hot pieces of metal, trying to adjust them accurately. But the '64 Ford was the king of annoyance, because the adjustment was specified WHILE THE ENGINE WAS RUNNING. You want a bad time: try adjusting a nice hot threaded bolt with a locknut, while it's jerking through about 15 degrees of movement 400 times a minute, while hot oil is spraying out of the valve train lubrication lines, and you have to feed a feeler gauge between the bottom of the bolt and the top of the pushrod during the brief moment they're not in contact. Oh, and the cam was sufficiently aggressive that at idle the car was continuously backfiring through the carburetor so there were occasional blasts of flame from right in front of you.
Compared to that, what's a little hassle like removing a wheel to replace the battery? I was so glad to see that car go, even if it did have the hottest engine Detroit ever made.
>First, vehicle information is very proprietary. Why is it that cars can't report status information via a simple USB connection?
Build yourself one of these. It's an OBDII-to-USB converter. It still requires *extensive* software on the computer side, but you're already talking about having that. On-board vehicle diagnostics are fairly complicated, but there are plenty of programs that handle it, many for free.
I agree it'd be nice to have sensors to detect fluid levels... but until sensors are more reliable, you might start relying on something that has broken and end up in trouble. In my car, the wiper washers are plumbed such that when you're just about out of washer fluid, the rear window pump stops working, which is clever and avoids having to add an extra sensor.
It would be nice to see more electronics that used USB connectors to get power, if they can stay under half an amp. It'd also be nice to see more wall-warts with some sort of load-detection circuitry and a solid-state relay so if they're not being used, they disconnect the transformer/switching PS from the AC so it doesn't sit there sucking up power.
I don't know how much you know about transcription/translation, so I'll do a very quick summary. DNA -> RNA, a very closely related polymeric molecule. The RNA then gets translated by transfer RNA molecules -- themselves made of RNA (which is why we think RNA might've been the original stuff of life, because it can both hold information and perform transformational changes to other molecules.) The transfer RNA molecules, tRNA, basically have three RNA's sticking off one side, that are complementary to the messenger RNA produced from the DNA template (a mirror image of a mirror image reproduces the original image) and on the other side a single protein for which they code. Protein assembly consists of linking those proteins carried by the tRNA together.
So think of this as two systems that interact: DNA is the memory, protein is the implementation, and the tRNA is what translates from the memory to the implementation.
We could presumably figure out a way to hijack this if we could build enzymes that built different tRNA molecules:
A: by putting on different, 'unnatural' proteins onto the tRNA's, which would allow us to make enzymes that could do things no current enzymes can do. (To some extent this already happens: bacteria produce/use weird proteins in their cell walls, including proteins that are mirror-images of the sort of proteins seen throughout the rest of the living world. But the possibilities are much greater than that, because enzymes are thermodynamic factories that can do amazing chemistry if you can just find a way to judo chemicals A and B into chemical C by building an enzyme with exactly the right surface topology to make A and B stick together the way you want them to so they turn into C, and being able to supply extra electron density right where you want it, rather than just having to settle with the current 20-something proteins we have, would massively help in such engineering.)
B: by putting on different, 'unnatural' RNA's on the other end of the tRNA, to match to engineered DNA, so you could potentially replace the entire DNA/RNA system with something that works better (fsvo 'better').
The thing that's nice about this is if you can mess with the tRNA, you can use the existing DNA toolchain to build weird proteins, or use the existing protein toolchain for playing with different DNA.
This research is an early step down that road: finding engineered DNA that can work like existing DNA. They're reverse-engineering, which will presumably allow them to find what parts of a base nucleotide you can change, while retaining the ability to use the cell's existing toolchain.
I'm not in this field anymore so bear in mind my knowledge might be very dated.
Generally speaking, viruses that insert their DNA into eukaryotic DNA don't have a particular place that they do so: they get their DNA into the cell, and it then inserts itself randomly in some bit of exposed DNA. See, eukaryotic DNA is very tightly bound to accessory proteins that protect/maintain it and hold it in some sort of to my knowledge poorly understood large-scale organizational scheme that constitutes a chromosome, so it's not like you can get to just anywhere on the DNA, but the parts you CAN get to might not be consistent, depending on whether that particular DNA bit is being transcribed at the moment, or repaired, or what have you. So what happens is that viruses stick their DNA all *over* the place, and the vast majority of them are indeed in no-ops or unread/untranscribed sections and just sit there -- which is where all the endogenous retrovirus stuff we read about comes from. Complete replication of the DNA is rare -- it only happens when the cell needs to divide for some reason. Small-scale scanning and replication is very common because a cell's day-to-day enzyme turnover requires it. But that small scanning is less likely to hit the area where the virus DNA is because of the sheer size of the genome.
Eventually it'd be nice to migrate to a whole different genetic code and support enzymes, because then viruses would be instantly nonviable, but that's a long, long ways off. However, this research is the first step: if we had a wholly different DNA, and re-engineered the enzymes that make transfer RNA, which convert the messages read from DNA into protein, we could retain all the protein-handling enzymes we have, and everything they make, and 'only' swap out the DNA and RNA suite. That's still an enormous problem, but it's like 0.01% of the problem of trying to engineer new proteins and a whole metabolism based on them.
This could be really useful in the long-term: if we could substitute replacement codons that work with most of our existing DNA, it's one step to building really tough DNA. Right now, there are a lot of damage mechanisms like adjacent thymines linking resulting from exposure to chemicals or shortwave radiation, and replacement codons engineered to not be suseptible to these could make, say, protracted exposure to radiation outside the Earth's protective atmosphere more viable. Of course, then we'd have to engineer a whole set of enzymes to synthesize those new codons, which is an extremely hard project, but finding things that work as replacement base pairs, now, gives us time to study how they might fail and figure out what the best candidates are.
Well, ideally, a projectile would be shaped like a zep, right? A toroid with a teardrop cross-section would be better than just shooting bearing housings, but a *good* system for launching things would launch minimized-drag objects and no toroid really can compete on that basis. But this system can only shoot toroids. What I'd like to build is a coilgun where the primary and core are wrapped around the projectile, so it doesn't need a hole through it, but I haven't gotten those to work very well -- they seem to work much better using a single massive pulse of DC, like a railgun. Still working on that... The other problem is that with both systems you want your mass to be as close to your core as possible to maximize the energy transfer efficiency, by avoiding the air gap loss, and any good aerodynamic shape is going to have lots of air gap loss since it only reaches its max width at one point. Frustrating when the laws of physics dictate contradictory shapes.
>If everyone would just follow them, then the world would run like a well-oiled machine.
And the thing is: they're right. The world WOULD run great if everyone just did exactly what they did. But it would also run great if everyone did something totally different yet. The problem is that there are lots of people doing lots of different things, differently, for what they consider perfectly good reasons. I think the last bit of that is critically important, though: everyone, even the idiots who do $something_I_hate, think they're doing the right thing. It's that little jump that's difficult for fundamentalists because they don't think someone else can have a justification or rationalization that is both different and valid. That's where a lot of the world's conflict comes from. Engineers try and argue you into believing what they do, because they're convinced they're right and they can change your mind. Religious nutjobs threaten or use violence to get you to believe what they do because they're convinced they're right and they insist you change your mind. (Which is why I prefer engineers. That and very few of them believe that God is on their side. Well, okay, actually they do: they just call God 'logic'. But that's less hostile.)
The other thing about engineers is that I think it's easier for people who are led to engineering to get sucked into a project without considering all the ramifications of the result. Like the movie Real Genius, they can get so worked up building something that they overlook or refuse to consider its dangers: a consequence of being highly focussed. Similarly fundamentalists, who can go from "this is for the Good of all mankind" and get wound up in the goal, and justify doing awful things to get to an end they think is worthwhile or necessary. It's not like determination, or focus are bad things: it's just that people who have that can be a great deal more constructive or destructive than people who spend all their time worrying about the consequences of their actions.
A previous poster pointed at Engineer's Syndrome, and I see some similar tendencies.
Engineers -- and I'm speaking as someone who is doing an engineering job, surrounded by engineers, and from a family of engineers -- tend to favor experience more than empathy. They tend to think that if they're convinced something is right, it's for good reason, and once they're convinced, it takes some work to change their minds. More particularly, if they're convinced, they're unlikely to use someone else's experience as a guideline: they're less likely to put themselves in someone else's shoes to regard a problem from that standpoint.
My own definition of Engineer Syndrome is encapsulated in the phrase, that I actually heard from one of my dad's coworkers once, "If you would've thought about this problem as much as I have, you'd agree with me." The level of premise and and patronization enclosed in that one sentence is staggering, but when it comes right down to it, I think many people drawn to engineering feel that way at some point or another. The consequence of this is that if someone else *doesn't* agree, the person suffering from ES thinks the other person is either stupid or stubbornly wrong, and either way, is a fool whose opinion is not to be regarded.
Likewise, engineers come from a background where things are provably correct (mathematics) or experimentally verifiable (most of the rest of science and engineering) and take that sense of certainty and apply it in areas where it isn't applicable -- sociology, politics, art, places where it really does come down to opinion, where there isn't actually a right and wrong, just preference.
The fundamental difference is that engineers do tend to rely on things that are provably correct or experimentally verifiable, whereas religious extremists are predicating invisible omnipotent entities. But the point is: if you have people who have this engineering set of mechanisms and filters for dealing with the world, and who believe in invisible omnipotent entities, they're going to have similar behavior to people who are drawn to engineering.
Mine are nowhere *near* supersonic and are comparatively unstable in flight. Even if they weren't, the idea projectile shape for drag reduction is basically like a Zeppelin, definitely not a toroid. Unfortunately, this design relies on a toroid as the shorted secondary winding of a linear motor, basically, so I'm stuck with it.
That's okay: it'll still go through a sheet of drywall.
Batteries are optional. I've made a couple coilguns out of a christmas paper wrapping tube filled with welding rod (cut to different lengths so it's solidly packed at the bottom tapering to 50% fill or so at the top) with a big coil of magnet wire around the bottom, that's attached directly to 110V AC (with a relay in there to turn it on/off.) The drag is it only fires ring-shaped projectiles -- but boy do they fly. Some time it'd be nice to build one that launches spheres rather than toroids, but that seems to require a lot more work.
Not exactly the same, but in some ways more impressive: the company my now-ex-gf was working for, designed and built an ultrasonic imaging device that they could feed into your femoral artery and snake up into your beating heart to image the insides real-time. No cracking the chest open, no shutting the heart down and rerouting the blood through an external pump.
It gets better. They could click the ultrasound transducer into high power mode and selectively kill small sections of the heart that were beating incorrectly -- were hindering, rather than helping, the contraction of the heart, which apparently happens when scar tissue from a previous heart attack reduces or blocks the nerve signal propagation across the heart muscle -- so they could do actual surgery to correct heart malfunctions, with nothing more than a tiny cut on your leg.
Her job was plating quartz substrate with layers of copper and gold, then cutting it into tiny sections that acted as the piezo elements for the transducer. She then hand-soldered wires onto each of the hundred-some elements, using wire about the diameter of a human hair, because the whole transducer was like 2mm in diameter and only a couple times that long. It was amazing. (So was the plating process, which involved solid chunks of copper and gold that weighed a kilo or more, each one. They were pretty careful about the gold ones, but let me have several of the used copper ones, so I have table coasters way cooler than either the infamous AOL cd's or hard drive discs: massive machined chunks of copper with one side plasma-etched away.)
I was at a small, but unaccountably well-connected college, sending email to my dad, who was working at a big tech company. Usually I just copied whatever route he'd chosen, and then marvelled as my mail got to him in LESS THAN TWO HOURS -- the very idea! At the time, it was still pretty unusual that two people would both have access to email, so I actually showed off to my friends -- "hey look at THIS!"
Well, one of my friends knew more than me, so he taught me about uuhosts -- a way to find out what was connected, for the times when my email was just vanishing because something, somewhere, was offline. So I used it. The next day I got some Very Crabby Email from a sysop who tore me a new one for using a satellite uplink to send personal email to Japan and back.
It felt like having a switchboard operator yell at me. I was *mortified* and I didn't even know for sure what I'd done.
It's possible I'm wrong, but that's what I've been told by coworkers who design bandgap references. I don't do chip design or know that much about it: I know they're heated, but coworkers indicated they were actively cooled as well, if necessary.
Peltier devices on-chip have been used for a while, whenever temperature variations are intolerable. Some examples: Analog Devices AD595 thermocouple amp, which uses in-chip thermal calibration to ensure a cold junction of known temperature, and many voltage regulators and switching supply controllers that use temperature-controlled bandgaps as their voltage reference.
>Fair enough. What was your MPH though?
'bout 1/2, if that. *slow*. The reason American cars have so much power is because mechanical friction increases linearly with speed, but the power required increases as the cube of speed. (Air resistance increases as the square, but power as the cube, before someone jumps in to correct me: force rises as the square, as a result of air resistance, and power as the cube.) So the point of having big engines isn't to climb, it's to climb *faster* or to have a higher top speed. Most cars cruising at 100 km are only using like 20 hp, if that. As an ex-bike-racer I can beat most cars at stoplights, but once we get above 45-50kph they leave me behind.
I agree there are some pretty steep hills in Vermont. Likewise in Seattle, San Francisco, and where I'm at, the mountains of Colorado. But according to various departments of transportation and everything I've managed to find and measure, Baldwin is the steepest paved street in the world. Those pictures don't really portray just how steep it was. It's hard to convey the feeling of riding up something that steep, and I ride my mountain bike up stairs -- multiple flights of stairs. It's REALLY steep. Although there are some little roads in Wales that are just terrifying: you come over the top of a hill and you literally cannot see the road because it drops off so steeply it's obscured by the hood of your car, and that's in a tiny economy car.
>I think this is going a bit far. Not everything is useless junk after all. Some people load up drywall and other building supplies for home improvement projects. Also, having some cars that can only go 4 MPH uphill on an interstate is asking for trouble. Do you really think the car gets 56 MPG if it has a top speed of 4 MPH going up an incline?
People who want to haul drywall will hopefully have enough sense to buy something else: this is so not a car for a single-car family. Many of my coworkers own tiny economy cars or one of a half-dozen hybrids parked out front, and at home they have a big ol' pickup that they drive when they need drywall. What's scary is that they actually save enough on gas in the tiny car to pay for a '70's pickup. It's a nice idea: they recycle old cars, basically, and use the appropriate car for the job, and it doesn't actually cost much (except for that whole buying-a-new-hybrid thing) but it does take up a lot of space.
I'm *sure* that car doesn't get anywhere near 56mpg on climbs. Then again, mine doesn't either: I get about 30mpg on average, but more like 5mpg on climbs. When I drive up to my grandparents' place, which is at about 11,000 feet elevation, I get like 15 mpg, and on the way back home I get like 50, because I'm coasting half the way. The thing is: even floored, throttle wide open, an engine that small will only suck up maybe 4 gallons per hour, if that. I'm basing that on how much fuel a VW engine at 3000 rpm uses, in VW-powered aircraft: this engine will be racked out at more like 6000 rpm on a climb, but it's only half the size of a bored & stroked VW. A big ol' Hemi 440, in contrast, can slurp over 30 gallons an hour at 4000 rpm on a climb.
This isn't the right car for a family that lives in the country and only has one car, or a family with lots of kids, and probably not even for a family that lives in a hilly area like yours. But as a single-person commuter car in urban/suburban environments, this would work as well as many other crappy little econoboxes that are selling well in lower-income demographics.
>Besides the car was woefully underpowered requiring near lead footing to use on the highway...
I was following a SMARTcar the other day on my way to work. It was doing 65mph up a steep hill at 6500 feet elevation. It might not have as much power as a Silverado: I wouldn't want to haul boat trailers with it. But it had more power than it needed for more difficult driving conditions than 90% of them will ever see.
While I enjoy your comment, I'd like to point out that 30 hp is a lot of power. I managed to get up Baldwin Street on a bicycle -- that's 2/3 horsepower, and there aren't steeper streets in Vermont or anywhere else. Granted, the bike and I, together, only weigh 70 kilos, but with 45x more power, and lower gearing, it'd do just fine... at the expense of speed. One of my friends has a Pinzgauer, that weighs 3000 kilos and can haul 14 people through waist-deep water full of rocks and up a muddy slope on the other side. It has a 65 horsepower engine. Any amount of power can get you up any hill if you have low enough gearing. So the problem isn't that it can't go up hills with heavy loads, it's that fatass American families won't buy a car that can only do 4 mph up the hill to their home when they've packed the back with useless junk they bought at the mall and stuck their enormous selves in every seat.
If you're thinking about trying this, do a single-layer board with a dozen BGA sites and find yourself some cheap BGA's to try and practice. It's a bear. You can remove screwed-up ones the same way you put them on: one heat gun on the bottom on low and one on top on high. Again, it's a bear, and carefully removing all the solder with solderwick from every bga pad is slow and demanding.
The main problems we see are: ground plane connectivity with the bga bump, because the ground plane pad isn't warm enough to reflow well so you get intermittents (boards that work only when you push down on the bga) which we can sometimes solve with flux and lots of heat and, as I said, poking at the chip to make sure it's bouncing rather than rocking... which leads to shorts between adjacent BGA bumps. What we've done is to figure out -- on our microSMD's -- a way to tell if there are shorts by continuity testing between points elsewhere on the board. It would be a lot easier if you're doing your own layout: use via-in-pad or adjacent to pad (much cheaper) on every pad, and then you can check continuity from the bottom side and see if there are shorts before you light it up. I can tell you from personal experience that one serious short will end up delaminating the pads from the substrate, so when you try and remove the BGA you rip off all the pads adjacent to the short as well, scrapping the board, unless you are a true wizard of pad-and-trace repair.
Another thing I've done in an emergency -- a unique chip, a customer failure that we had no option but to get working to replicate the failure -- was to cut off small chips of solder of known length, with a fixture involving a pair of nail clippers nailed to a board so I could feed solder through the clippers until it hit a stop, clip, and repeat, to get exactly the same length of solder each time. Then I'd touch each bit of solder, lift it, and touch it to a pad on the bottom of the chip, approximating rebumping the chip (since we have no facilities for doing that the right way, with a mask and precalibrated solderballs that are gently reflowed.) It worked, after two tries, but boy was that unpleasant, high-stress work. The point being: you can reuse both the boards and the chips if you're very careful, spend a lot of time, and do a lot of checking before powering anything up.
>Such a board is possible (and relatively easy) to design by an amateur - but very expensive to make as it would have to be 6 layers and require soldering BGA chips.
I have a few tips from personal experience. You can get multilayer boards built fairly inexpensively if you can justify having four made at one time: you might be looking at under $80/board for a 6-layer (although I'm not positive about that. I know you can get 4 layer done for under $60/board.)
It's possible, although unpleasant, to reflow your own BGA's. You need a microscope with a tilt-head. Draw the BGA package outline in the layout software as a silkscreen, making sure it's at least as large as the actual package, or even better, draw several outlines of increasing size. Align the BGA visually within the closest package size, double-check by looking at the edge with bright illumination and a microscope to make sure you're basically on-pad, then gently reflow it down with a heat gun. It works best if you can preheat the board from the bottom with one heat gun on low, then do the reflow from the top with the second one.
I'm doing this at work with microSMD, which are way, way smaller than BGA -- chips 3mm on a side with 12 bumps on the bottom. After a bit of work I have a 70% success rate. The main thing I've found is that while you're reflowing, you'll see the chip move as the capillary action of the solder pulls it into place. Very, very lightly touch the chip on one edge with a probe. If it rocks, the center isn't yet reflowed and it's pivoting on the as-of-yet-solid bumps. When the whole chip bounces like a spring on all the melted bumps, rather than rocking, then it should be good.
>The rationale for having me boot my computer apparently was that it may be a bomb, not that my contents might be suspicious. The logic of having me sit in front of them and power on a bomb just to find out if it is, in fact, a bomb still escapes me to this day.
Do you remember that old Daffy Duck cartoon where he's checking ammunition by hitting it with a hammer and if it doesn't go off, he writes 'dud' on it? Maybe you've just discovered why the TSA is hiring so many people and advertising at local community colleges for job applicants.
That's why I said 'correlate' in the last sentence -- I don't think that treating the weak better is the reason civilization exists. I think civilization exists because cooperative group effort makes that group outcompete other groups with less cooperation. As such, civilization is a competitive evolutionary strategy: it exists, and expands, because people who do it survive longer and can better care for their children. However, since it is a cooperative group effort, I think that over time giving more power to weaker members is an emergent quality, although that might have to do more with group psychology and the (mostly human, only sometimes observed in other animals) tendency to try and force life to be fair that we exhibit. In short, people want life to be fair becausse it makes them feel more in control of their futures, and they will pay, in labor or money, to try and make it fair. Since humans have empathy, we will as a group do the same thing for the weaker members of the group and over time pull them into the group. It's weird behavior, frankly, but we've been doing it fairly consistently for all of human history, and it is, well, nice. So, more power to us as a race: we're not unrelentingly horrible.
They're using the sodium chloride as a thermal reservoir -- heating it and relying on its high temperature to make up for its so-so specific heat. Water's specific heat isn't much different, but it's difficult to contain as steam. So they heat up the salt -- or anything else -- and let it gradually cool down, extracting heat from it by vaporizing water and reclaiming the energy through turbines. That way they can produce power all night off the heat saved during the day.
It's not a bad idea if they have a good insulated container for the molten salt. It introduces a lot of waste because of the cumulative inefficiency of heat transfer between the different systems, but it allows a system based on this to provide more reliable energy -- energy that's closer to being on-demand, rather than just when the sun is shining strongly enough.