Oh, geez, you're right. I was thinking the stuff in the glue *was* butadiene. Sigh. Yeah, alternating single/double bonds makes for a great polymer starting material, whereas terminal hydroxy groups don't do much of anything but blind you if you drink too much of them.
Gyah. No wonder I'm not in chemistry anymore. SO totally wrong.
Butenediol would be interesting: I'm assuming it's 2-butene, not 1-. 1-butenediol would have the start for a resonant structure down there at the end, with the pi orbitals of the oxygen and the pi orbitals of the double bond lined up, so it should be somewhat willing to polymerize. and blah blah blah.
>The chinese replaced the original glue with another, why?
They replaced an expensive chemical with a cheap one. That's called the free market in action. If your goal is profit, well, let's hear it for the free market. If your goal is having children live to a reasonable age, still able to count to ten with bifocal vision, if you know what I mean, then yeah, regulation starts to look a lot more attractive.
In case anyone was wondering -- or just so I can use my chemistry degree for like the second time since graduation ten years ago -- 1,4-butanediol is an industrial chemical synthesized in *enormous* quantities because it's the precursor to a whole slew of useful materials, mostly rubber polymers. BUNA rubber gets the first half its name from butanediol. It was developed during WWII by both German and American/Canadian chemists because of a worldwide rubber shortage, since submarines kept sinking all the cargo ships. The stuff is derived from grain alcohol, easily and cheaply, and some bacteria can be coerced into producing it through large-scale fermentation.
In contrast, 1,5-pentanediol is significantly more difficult to make and doesn't have anywhere nearly the demand or volume production, hence its higher expense and the temptation to substitute the cheaper, more readily available material that's almost just the same (except for the metabolites.)
This is also why I don't trust herbal remedies that come out of China. That one carbon makes only a little difference in this case, but there are others where it'd be the difference between as effective as herbal remedies ever are and *dead*, and who are your surviving relatives going to pursue when it's a Chinese company that made the stuff?
Basically the same thing here: my cousin didn't have to take anywhere near what a civil/structural engineer has to take. It was statics, not dynamics or fluid dynamics or the like, and even then it was sort of Statics For Poets. Their expectation was that architects should be designing stuff that could be built safely, and subsequent engineers would figure out how to implement it.
What you describe is the case here in the USA, as well: the local building code office professional engineers analyze it not only to make sure it's not obviously unsafe, but also in light of local hazards an architect from somewhere else might not have considered. An architect from the Midwest might not think about earthquakes, while one from California might not think about tornado or snow load. That's what the local city/county engineers do.
When my cousin got her architecture degree, they required her to take mechanical engineering courses in statics: load calculations for cantilevered vs. supported beams, the like. Architects that are fresh out of school -- or at least the architecture schools I know about -- do indeed consider gravity and infrastructure. It's possible that this wasn't the case when Gehry was in school, or that he's been designing for so long he's stopped looking at the nuts-n-bolts, or that he's so famous he doesn't *have* to care about failures, but in any case, any design that's submitted for building goes straight to the local county/city building department's engineers and they go over the whole thing. Where I live, a lower-income suburb of a small city, I have to submit engineer-signed drawings to build a garage, and the city engineer then goes over those drawings briefly. To sum up: many competent civil and structural engineers all signed off on the plans for this building.
If you own or can rent an insulated dewar you can get liquid nitrogen from a local welding supply company. It's pretty cheap. LN ice cream tastes reasonably good and boy is it fast to make.
Note: do *not* try this with liquid oxygen. Just sayin'.
Yeah, those guys have raised their prices since last time I looked at them. I thought I had another place that'd do it for $30 for a 4-layer but I can't find them right now. Do you *need* multilayer for what you're doing? Can you stack two-layer boards? I've done that for proto work and it's worked pretty well if you have through-plated vias. Put your vias on 0.1" centers and use headers for interconnect -- more of a PITA but you can stick a lot of passives on the inside-facing layers, giving you more room on the tops, and that way your decoupling caps are *right* below your IC's. We've had situations where we actually got better performance, inductance-wise and power-loss-wise, using this than with a comparable multilayer using standard vias.
Cheap 2-layer: http://www.batchpcb.com/ -- 10-12 days, proto: $10 + $2.50 per square inch. Cheap multiplayer: http://www.myropcb.com/ -- 21 days, 4-layer: $40 for proto, $0.38 per square inch + $120 setup for large batches. Myro will also do flex stuff with copper on polyimide, which is useful and unusual.
>If you're trying to produce an artificial intelligence to run the robot then the low level electronics aren't terribly important to you.
Right up until the moment when a power spike caused by a big relay opening causes the robot arm to punch a hole through a wall you used to like...
Which, granted, is a good time for the observers, but maybe not what the programmer was hoping to accomplish.
(Under some design circumstances, power relays should have a nearby diode to shunt the back-EMF caused by the relay inductance. If your black box design doesn't happen to include that diode, and you don't think low-level electronics are terribly important, you could find out some very exciting things. I speak from personal experience on this particular subject.)
Well, okay, I admit I've smelted iron in my back yard... but it's amazing how much you can pack on a two-layer board if you're patient, and circuit board plotters/CNC mills have come down dramatically in price. Once you have that kind of technology, you can start doing multilayer boards by stacking them, if you're willing to deal with the frustrations of having to hand-solder every via (and in the case of stacked boards, not being able to rely on contact on one inside-facing board, although that's okay if it's a shielding ground plane and you can get a wire soldered onto the edge.) But given that you can get multilayer SMT fabbed for you for $30 for a small board, it's hard to justify DIY. Obviously it depends on circuit complexity, but designing a recording electrocardiogram that writes FAT32 to a hard drive -- a nice mix of analog and digital -- is easily manageable using homebrew design, free software, and cheap board fab houses. (I personally think that SMT is *far* easier to work with than through-hole, but then again I have a cheap stereomicroscope.)
Speaking from personal experience, this sort of building-block approach to electronics can let a person down. I've designed a bunch of electronics, small simple chunks that do things like translate a logic-level signal to a relay that can switch ovens or computer-controlled A/D converters. By themselves, they work, but when you start just stacking them together like black boxes, their cumulative errors start to bite you -- or you put in a black box that contains a switching regulator and the line noise on the output wipes out everything downstream. If you don't know what they're actually doing, you don't know what their side-effects are going to be. The amount of post-regulator processing required to make a switching regulator look like a good, pure voltage source would be bulky enough to make that black box significantly less useful, and all that processing might not be required for 95% of possible loads.
Likewise, my coworkers do analog design of IC's, and even though we have a design reuse library for the company, every design they do is basically ab initio because another similar design does something they don't need and as a result uses up vital silicon space, and they can't simply remove just that bit.
A talented designer could use building blocks to build something great. A lousy designer could use those same blocks to build something dangerously unsafe -- they facilitate only design, not quality. Speaking as a lousy designer, I think it's a much better idea to actually do the work in analyzing the problem and coming up with an adequate design, and the good designers, in my experience, already *have* a head full of black boxes, for which they understand the limitations and how they interact.
It's some sort of sad, ironic commentary on my way of doing things that my 'fast' computer -- a 500MHz AMD -- does, indeed, just get used to check email/browse the web. That's *it*. In contrast, two of my 'slow' computers -- a P166 and another, maybe 266, have all the heavy stuff: relays sticking out the back that are interfacing with stepper motors, serial-port-attached one-wire thermometers, lots of custom programming to implement PID controllers, software to talk via GPIB to my oscilloscope and function generators. I guess it's kind of like keeping the '65 Jag for special occasions and using the dump truck for a daily driver.
In the late '90's I was working at a contract manufacturing place, diagnosing failed motherboards. We had these enormous quad-CPU PA-RISC boards that had 5000 components on them. Someone somewhere in the supply chain had gotten a reel of caps that had been loaded in the reel backwards -- I have no idea how. Anyway, the pick-n-place machines stuck them down, just like normal, some 500 per board or so, all with reverse polarity. We were building that type of board at about 30 per hour, and it took almost an hour for the first boards to get to functional (power-up) test. At that point, the RP caps all vaporized at the same moment. It was like a miniature war zone: the caps would blow out tiny flaming chunks of stuff, leaving little spirals of smoke, while tiny flames shot upwards and downwards out of the test racks. It was awesome, although I'm sure glad I wasn't standing right over the first boards when they went. They burnt holes deep into the 22 layer circuit boards. Then we had the job of finding, removing, and replacing, by hand, 500 caps on each board, and if we missed even one: fwoom! there goes another board. I don't think we managed to get a single board from that lot repaired and out the door.
Much smaller than your experience, but very impressive nonetheless.
This was about five years ago and he'd implemented a 16-bit (instruction and data) computer entirely through TTL, with wire-wrap. It included enough to do RS232 so you could telnet into it, although it certainly didn't have anything like an expandable bus. They'd written a tic-tac-toe game for it, and it was pretty good.
The really odd moment was overhearing the hardware guy talking to one of the software guys, who was bemoaning the lack of a logical shift-right as opposed to a bitwise shift-right in the assembly code. The hardware guy sat down, drew a couple of things, and said, "yeah, we can add that with four gates." Wouldn't THAT be nice, to be able to spend two hours wiring, and add a new assembly instruction to your processor?
I wish I could find links: they're all members of the Denver Mad Scientists' Club, but I can't find anything on their homepage.
Babbage's engine was hard to make *then* because they didn't have good machining capabilities. Nowadays you can make a small difference engine out of LEGO blocks -- it doesn't get much more POTS than that. (notice the geek cred in the domain name, btw.)
It's surprisingly easy to draw wire. I've done lots of it. The original stuff was done without drawplates: they filed a notch in a plate, then put a second plate against it, clamped them firmly, and pulled, then used the next, smaller notch. You get a half-circle of wire that tends to curl but it's doable. Insulation is *much* harder -- making something that's flexible, tough, and has a reasonable dielectric, and getting it to stick to the wire, is *hard*. Drawing 30 gauge copper tubing is easy in comparison.
Voltaic piles are easier to make than Leyden jars. If you're bored, you can light an LED with a stack of small pieces of aluminum and pennies, separated by lemon-juice-soaked paper towels. It took me about 7 of each to get a red LED to light. If people had known what to do they could've made voltaic pile batteries in Egyptian times -- separate copper and silver chunks with spit- or saltwater-soaked papyrus sheets.
There were early relays made from glass tubes with wires and piles of steel filings. An electric charge on one wire attracted filings, which bridged to the other wire. You could use those to make primitive high-power diodes as well, by messing with the geometry of the wires -- again, stuff that any culture with some competence in glass could've done (and that's pretty old.) The problem was always one of basic research and not knowing what to try.
At least that's what our department head, the guy with advanced degrees in engineering and marketing, says. His claim is: companies that buy other companies who do something similar end up diluting themselves and losing maneuverability. Apple's already designing hardware *and* operating systems *and* lots of applications. Do they need to spend money on *more* applications, when those applications are currently being managed by someone else who knows how to market them, and whose marketing helps drive Apple's sales effectively for free?
Using a set of tools with a known accuracy to produce another set of tools with better accuracy is a solved problem -- difficult to implement, but very possible with care. Consider that most everything we as a civilization have built, was built with poorer tools. In this particular case, a good replicator would base its physical measurements on a physical constant, probably wavelength of light. A Fabry-Perot interferometer is unbelievably accurate, able to measure the deflection of a brick wall when you lean on it, and all it requires is a laser and some mirrors. If the machine can make a laser, it can make a replica of itself that has the same accuracy it has.
The reason molecular-level construction would be nice is that many of our manufacturing processes rely on altered bulk characteristics of materials. Alloys outperform raw metals, semiconductors are silicon with precise impurities, and engineering metals are heat-treated or forged to give them macromolecular structures with specific properties that merely throwing atoms into a general form can't duplicate.
When we can print an entire engine -- crankshaft of iron with molybdenum and chromium that has high nitride content right at the bearing surface, printed bearings of layers of copper and lead alloys for bearings, block of zinc/copper aluminum gradually tapering to high-carbon iron around the cylinders -- then we'll see some serious changes in manufacturing and quality/reliability.
I'm trying to build something using a wirefeed TIG welder currently, so it'll deposit any metal you want. The problem is that it has to have something to start with -- something to complete the circuit. So you can't do undercut forms, which really limits what you can create. (Plus, my TIG welder has problems building up charges or oxides or something on the tungsten so it only kicks the arc on about 95% of the time, which is basically useless for something like this, but at least it's easy to detect automatically.)
like this? It prints chocolate and it's made of LEGO bricks. Saul Griffith made one for his Master's Degree thesis project. I'm making a somewhat larger one right now. The table's moving around but I don't yet have the chocolate melting setup working.
For the record, *some* banks won't take just any signature. My mom's bank calls her when she starts using a different pen, to make sure it's her. (Yes, they're small and she's been with them for 50 years.)
Well, that just sucks. So they've got, basically, a low-pass filter with a rolloff so that their sampler never sees anything above 8khz? Phoo. My oscilloscope dithers the clock rate so even though it's 350 MHz, it acts cleaner because it moves back and forth across a signal, to evade aliasing artifacts. My thought was that multiple data streams from different cellphones would act similarly, but if the data's shorted to ground before it ever sees the ADC's front end, well, just phoo.
Yeah, it'd be crap. My hope is that by using multiple microphones and relying on the fact that they're bad in different ways, you'd be able to derive a fair amount of quality. Obviously, information that none pick up is lost. But oversampling to partially compensate for poor signal/noise is done with a lot of other noisy signals. Signals isn't my forte either, by a long shot. It's just an idea I thought was interesting, so I started playing with using some dsp software mixers.
I've spent some time designing and programming (but will never finish) something similar to this, just using cellphones' audio capabilities. Imagine getting twenty random people at a concert to call into a server and leave their cellphones running, recording the concert from twenty different points. From the overall stream, you should be able to derive an excellent, local-noise-removed bootleg, and from a bit of playing with signal intensities you should be able to figure out where the individual recorders were and do some nice sound balancing.
We're all carrying these great little computers: we should start doing networked or collaborative stuff with them.
But your starting example brings up a question. I wasn't interested in Civil War Reenactors until I read that there was such a thing, and then I got curious. Now I want to go read about it. From your post I get the feeling that you think most people aren't like this, that most people just don't care about anything they don't already know about. Do you think this is true? Coz, for my group of friends and family, it sure isn't -- if I say "hey, I just read a really cool book about Oranges, they'll all want to read it.
I guess maybe Wikipedia wants to be just for people who only want to read about things they already know.
Slashdot Mirror
Oh, geez, you're right. I was thinking the stuff in the glue *was* butadiene. Sigh. Yeah, alternating single/double bonds makes for a great polymer starting material, whereas terminal hydroxy groups don't do much of anything but blind you if you drink too much of them.
Gyah. No wonder I'm not in chemistry anymore. SO totally wrong.
Butenediol would be interesting: I'm assuming it's 2-butene, not 1-. 1-butenediol would have the start for a resonant structure down there at the end, with the pi orbitals of the oxygen and the pi orbitals of the double bond lined up, so it should be somewhat willing to polymerize.
and blah blah blah.
>The chinese replaced the original glue with another, why?
They replaced an expensive chemical with a cheap one.
That's called the free market in action.
If your goal is profit, well, let's hear it for the free market. If your goal is having children live to a reasonable age, still able to count to ten with bifocal vision, if you know what I mean, then yeah, regulation starts to look a lot more attractive.
In case anyone was wondering -- or just so I can use my chemistry degree for like the second time since graduation ten years ago -- 1,4-butanediol is an industrial chemical synthesized in *enormous* quantities because it's the precursor to a whole slew of useful materials, mostly rubber polymers. BUNA rubber gets the first half its name from butanediol. It was developed during WWII by both German and American/Canadian chemists because of a worldwide rubber shortage, since submarines kept sinking all the cargo ships. The stuff is derived from grain alcohol, easily and cheaply, and some bacteria can be coerced into producing it through large-scale fermentation.
In contrast, 1,5-pentanediol is significantly more difficult to make and doesn't have anywhere nearly the demand or volume production, hence its higher expense and the temptation to substitute the cheaper, more readily available material that's almost just the same (except for the metabolites.)
This is also why I don't trust herbal remedies that come out of China. That one carbon makes only a little difference in this case, but there are others where it'd be the difference between as effective as herbal remedies ever are and *dead*, and who are your surviving relatives going to pursue when it's a Chinese company that made the stuff?
Basically the same thing here: my cousin didn't have to take anywhere near what a civil/structural engineer has to take. It was statics, not dynamics or fluid dynamics or the like, and even then it was sort of Statics For Poets. Their expectation was that architects should be designing stuff that could be built safely, and subsequent engineers would figure out how to implement it.
What you describe is the case here in the USA, as well: the local building code office professional engineers analyze it not only to make sure it's not obviously unsafe, but also in light of local hazards an architect from somewhere else might not have considered. An architect from the Midwest might not think about earthquakes, while one from California might not think about tornado or snow load. That's what the local city/county engineers do.
When my cousin got her architecture degree, they required her to take mechanical engineering courses in statics: load calculations for cantilevered vs. supported beams, the like. Architects that are fresh out of school -- or at least the architecture schools I know about -- do indeed consider gravity and infrastructure. It's possible that this wasn't the case when Gehry was in school, or that he's been designing for so long he's stopped looking at the nuts-n-bolts, or that he's so famous he doesn't *have* to care about failures, but in any case, any design that's submitted for building goes straight to the local county/city building department's engineers and they go over the whole thing. Where I live, a lower-income suburb of a small city, I have to submit engineer-signed drawings to build a garage, and the city engineer then goes over those drawings briefly.
To sum up: many competent civil and structural engineers all signed off on the plans for this building.
If you own or can rent an insulated dewar you can get liquid nitrogen from a local welding supply company. It's pretty cheap. LN ice cream tastes reasonably good and boy is it fast to make.
Note: do *not* try this with liquid oxygen. Just sayin'.
Yeah, those guys have raised their prices since last time I looked at them. I thought I had another place that'd do it for $30 for a 4-layer but I can't find them right now.
Do you *need* multilayer for what you're doing? Can you stack two-layer boards? I've done that for proto work and it's worked pretty well if you have through-plated vias. Put your vias on 0.1" centers and use headers for interconnect -- more of a PITA but you can stick a lot of passives on the inside-facing layers, giving you more room on the tops, and that way your decoupling caps are *right* below your IC's. We've had situations where we actually got better performance, inductance-wise and power-loss-wise, using this than with a comparable multilayer using standard vias.
Cheap 2-layer: http://www.batchpcb.com/ -- 10-12 days, proto: $10 + $2.50 per square inch.
Cheap multiplayer: http://www.myropcb.com/ -- 21 days, 4-layer: $40 for proto, $0.38 per square inch + $120 setup for large batches.
Myro will also do flex stuff with copper on polyimide, which is useful and unusual.
>If you're trying to produce an artificial intelligence to run the robot then the low level electronics aren't terribly important to you.
Right up until the moment when a power spike caused by a big relay opening causes the robot arm to punch a hole through a wall you used to like...
Which, granted, is a good time for the observers, but maybe not what the programmer was hoping to accomplish.
(Under some design circumstances, power relays should have a nearby diode to shunt the back-EMF caused by the relay inductance. If your black box design doesn't happen to include that diode, and you don't think low-level electronics are terribly important, you could find out some very exciting things. I speak from personal experience on this particular subject.)
Well, okay, I admit I've smelted iron in my back yard... but it's amazing how much you can pack on a two-layer board if you're patient, and circuit board plotters/CNC mills have come down dramatically in price. Once you have that kind of technology, you can start doing multilayer boards by stacking them, if you're willing to deal with the frustrations of having to hand-solder every via (and in the case of stacked boards, not being able to rely on contact on one inside-facing board, although that's okay if it's a shielding ground plane and you can get a wire soldered onto the edge.) But given that you can get multilayer SMT fabbed for you for $30 for a small board, it's hard to justify DIY.
Obviously it depends on circuit complexity, but designing a recording electrocardiogram that writes FAT32 to a hard drive -- a nice mix of analog and digital -- is easily manageable using homebrew design, free software, and cheap board fab houses.
(I personally think that SMT is *far* easier to work with than through-hole, but then again I have a cheap stereomicroscope.)
Speaking from personal experience, this sort of building-block approach to electronics can let a person down. I've designed a bunch of electronics, small simple chunks that do things like translate a logic-level signal to a relay that can switch ovens or computer-controlled A/D converters. By themselves, they work, but when you start just stacking them together like black boxes, their cumulative errors start to bite you -- or you put in a black box that contains a switching regulator and the line noise on the output wipes out everything downstream. If you don't know what they're actually doing, you don't know what their side-effects are going to be. The amount of post-regulator processing required to make a switching regulator look like a good, pure voltage source would be bulky enough to make that black box significantly less useful, and all that processing might not be required for 95% of possible loads.
Likewise, my coworkers do analog design of IC's, and even though we have a design reuse library for the company, every design they do is basically ab initio because another similar design does something they don't need and as a result uses up vital silicon space, and they can't simply remove just that bit.
A talented designer could use building blocks to build something great. A lousy designer could use those same blocks to build something dangerously unsafe -- they facilitate only design, not quality. Speaking as a lousy designer, I think it's a much better idea to actually do the work in analyzing the problem and coming up with an adequate design, and the good designers, in my experience, already *have* a head full of black boxes, for which they understand the limitations and how they interact.
It's some sort of sad, ironic commentary on my way of doing things that my 'fast' computer -- a 500MHz AMD -- does, indeed, just get used to check email/browse the web. That's *it*. In contrast, two of my 'slow' computers -- a P166 and another, maybe 266, have all the heavy stuff: relays sticking out the back that are interfacing with stepper motors, serial-port-attached one-wire thermometers, lots of custom programming to implement PID controllers, software to talk via GPIB to my oscilloscope and function generators. I guess it's kind of like keeping the '65 Jag for special occasions and using the dump truck for a daily driver.
In the late '90's I was working at a contract manufacturing place, diagnosing failed motherboards. We had these enormous quad-CPU PA-RISC boards that had 5000 components on them.
Someone somewhere in the supply chain had gotten a reel of caps that had been loaded in the reel backwards -- I have no idea how. Anyway, the pick-n-place machines stuck them down, just like normal, some 500 per board or so, all with reverse polarity. We were building that type of board at about 30 per hour, and it took almost an hour for the first boards to get to functional (power-up) test.
At that point, the RP caps all vaporized at the same moment.
It was like a miniature war zone: the caps would blow out tiny flaming chunks of stuff, leaving little spirals of smoke, while tiny flames shot upwards and downwards out of the test racks. It was awesome, although I'm sure glad I wasn't standing right over the first boards when they went. They burnt holes deep into the 22 layer circuit boards.
Then we had the job of finding, removing, and replacing, by hand, 500 caps on each board, and if we missed even one: fwoom! there goes another board.
I don't think we managed to get a single board from that lot repaired and out the door.
Much smaller than your experience, but very impressive nonetheless.
This was about five years ago and he'd implemented a 16-bit (instruction and data) computer entirely through TTL, with wire-wrap. It included enough to do RS232 so you could telnet into it, although it certainly didn't have anything like an expandable bus. They'd written a tic-tac-toe game for it, and it was pretty good.
The really odd moment was overhearing the hardware guy talking to one of the software guys, who was bemoaning the lack of a logical shift-right as opposed to a bitwise shift-right in the assembly code. The hardware guy sat down, drew a couple of things, and said, "yeah, we can add that with four gates." Wouldn't THAT be nice, to be able to spend two hours wiring, and add a new assembly instruction to your processor?
I wish I could find links: they're all members of the Denver Mad Scientists' Club, but I can't find anything on their homepage.
Babbage's engine was hard to make *then* because they didn't have good machining capabilities. Nowadays you can make a small difference engine out of LEGO blocks -- it doesn't get much more POTS than that. (notice the geek cred in the domain name, btw.)
It's surprisingly easy to draw wire. I've done lots of it. The original stuff was done without drawplates: they filed a notch in a plate, then put a second plate against it, clamped them firmly, and pulled, then used the next, smaller notch. You get a half-circle of wire that tends to curl but it's doable. Insulation is *much* harder -- making something that's flexible, tough, and has a reasonable dielectric, and getting it to stick to the wire, is *hard*. Drawing 30 gauge copper tubing is easy in comparison.
Voltaic piles are easier to make than Leyden jars. If you're bored, you can light an LED with a stack of small pieces of aluminum and pennies, separated by lemon-juice-soaked paper towels. It took me about 7 of each to get a red LED to light. If people had known what to do they could've made voltaic pile batteries in Egyptian times -- separate copper and silver chunks with spit- or saltwater-soaked papyrus sheets.
There were early relays made from glass tubes with wires and piles of steel filings. An electric charge on one wire attracted filings, which bridged to the other wire. You could use those to make primitive high-power diodes as well, by messing with the geometry of the wires -- again, stuff that any culture with some competence in glass could've done (and that's pretty old.) The problem was always one of basic research and not knowing what to try.
At least that's what our department head, the guy with advanced degrees in engineering and marketing, says. His claim is: companies that buy other companies who do something similar end up diluting themselves and losing maneuverability.
Apple's already designing hardware *and* operating systems *and* lots of applications. Do they need to spend money on *more* applications, when those applications are currently being managed by someone else who knows how to market them, and whose marketing helps drive Apple's sales effectively for free?
Using a set of tools with a known accuracy to produce another set of tools with better accuracy is a solved problem -- difficult to implement, but very possible with care. Consider that most everything we as a civilization have built, was built with poorer tools.
In this particular case, a good replicator would base its physical measurements on a physical constant, probably wavelength of light. A Fabry-Perot interferometer is unbelievably accurate, able to measure the deflection of a brick wall when you lean on it, and all it requires is a laser and some mirrors. If the machine can make a laser, it can make a replica of itself that has the same accuracy it has.
The reason molecular-level construction would be nice is that many of our manufacturing processes rely on altered bulk characteristics of materials. Alloys outperform raw metals, semiconductors are silicon with precise impurities, and engineering metals are heat-treated or forged to give them macromolecular structures with specific properties that merely throwing atoms into a general form can't duplicate.
When we can print an entire engine -- crankshaft of iron with molybdenum and chromium that has high nitride content right at the bearing surface, printed bearings of layers of copper and lead alloys for bearings, block of zinc/copper aluminum gradually tapering to high-carbon iron around the cylinders -- then we'll see some serious changes in manufacturing and quality/reliability.
I'm trying to build something using a wirefeed TIG welder currently, so it'll deposit any metal you want. The problem is that it has to have something to start with -- something to complete the circuit. So you can't do undercut forms, which really limits what you can create. (Plus, my TIG welder has problems building up charges or oxides or something on the tungsten so it only kicks the arc on about 95% of the time, which is basically useless for something like this, but at least it's easy to detect automatically.)
like this? It prints chocolate and it's made of LEGO bricks. Saul Griffith made one for his Master's Degree thesis project. I'm making a somewhat larger one right now. The table's moving around but I don't yet have the chocolate melting setup working.
For the record, *some* banks won't take just any signature. My mom's bank calls her when she starts using a different pen, to make sure it's her. (Yes, they're small and she's been with them for 50 years.)
Well, that just sucks. So they've got, basically, a low-pass filter with a rolloff so that their sampler never sees anything above 8khz? Phoo.
My oscilloscope dithers the clock rate so even though it's 350 MHz, it acts cleaner because it moves back and forth across a signal, to evade aliasing artifacts. My thought was that multiple data streams from different cellphones would act similarly, but if the data's shorted to ground before it ever sees the ADC's front end, well, just phoo.
Yeah, it'd be crap. My hope is that by using multiple microphones and relying on the fact that they're bad in different ways, you'd be able to derive a fair amount of quality. Obviously, information that none pick up is lost. But oversampling to partially compensate for poor signal/noise is done with a lot of other noisy signals.
Signals isn't my forte either, by a long shot. It's just an idea I thought was interesting, so I started playing with using some dsp software mixers.
I've spent some time designing and programming (but will never finish) something similar to this, just using cellphones' audio capabilities. Imagine getting twenty random people at a concert to call into a server and leave their cellphones running, recording the concert from twenty different points. From the overall stream, you should be able to derive an excellent, local-noise-removed bootleg, and from a bit of playing with signal intensities you should be able to figure out where the individual recorders were and do some nice sound balancing.
We're all carrying these great little computers: we should start doing networked or collaborative stuff with them.
>I've worked with people who are considered processional journalists
So *THAT* is what the right mean when they call journalists the Fifth Column!
I agree with what you're saying.
But your starting example brings up a question. I wasn't interested in Civil War Reenactors until I read that there was such a thing, and then I got curious. Now I want to go read about it. From your post I get the feeling that you think most people aren't like this, that most people just don't care about anything they don't already know about. Do you think this is true? Coz, for my group of friends and family, it sure isn't -- if I say "hey, I just read a really cool book about Oranges, they'll all want to read it.
I guess maybe Wikipedia wants to be just for people who only want to read about things they already know.