The drones don't track anything. I don't even know if they have an IMU. All position data is acquired using a video motion capture system. I'd think it may be Vicon with Tracker software.
They were flying these things apparently over people's heads in a public demo of putting together a brick "building". The room was full of spectators. It almost looked like a live art exhibit.
Yeah, troll, because quadcopters' only use is to develop military tech. Must have never heard of people studying control engineering. Ultimately, proof is in the pudding. Either you make a device that works using your control scheme, or it doesn't. Now fuck off.
If you're cash strapped, I don't see you exactly splurging on NI stuff. Microcontrollers with decent A/D and D/A converters are plenty cheap, same goes for discrete A/D and D/A chips, and all the other stuff you need for signal conditioning and digital control. Never mind who the heck cares what brand you use for control equipment? I think this guy is just trying to leverage familiar names to prop up his scam.
Alas, his scam doesn't mean the idea is unworkable, or that the serious people who are producing reproducible results are faking it.
My problem is the opposite: it's not about the unnecessary stuff, it's about the cheap, non-invasive things they could do that they stay away from like if it was a transplant or something. My pregnant wife had to spend a couple of hours bleeding in the ER for a clueless resident to finally pick up the ultrasound and tell us "well, there's an apparently intact fetus with a beating heart, stuff happens, no need for a cleanup surgery, go home". They were already setting up an OR for a D&C -- what kind of an idiot does that before doing a basic ultrasound that takes 60 second start-to-finish?!
Another story: I have a solid family history (with tombstones and documented post-mortems to prove it) of plaque formation in coronary arteries, and the doofuses would drag their feet on noninvasive transthoracic ultrasounds of precisely the non-invasevely accessible parts of those arteries that were problematic in my family! I mean, come on, how stupid do you have to be.
Some guys I know decided they've had enough of bullshit and simply got an ultrasound for themselves to play with. It doesn't take 6 years of med school to do a decent doppler coronary artery exam. You simply need enough practice and access to rudimentary literature (and a modicum of intelligence). When you have the tool in your basement, you can get more hands-on time in a month than a resident gets in a year. Of course it helps if you're an engineer and can troubleshoot things and fix them when something breaks.
Tail recursion is not the problem I refer to. Merely the amount of garbage generated just making copies of stuff is what's killing it. Even Oracle's Java deals with it relatively poorly -- thus the presence of hardware-assisted garbage collection (at a price tag that'll make you weep). Open source implementations generally don't really have garbage collectors that address the architectures of today's CPUs. They are definitely not optimized for cache pressure, and are simply wasteful in their resource use. Even garbage collectors from Frantz and Chez Scheme seem to be relatively unimpressive when you cachegrind them. Mike Pall rather detailed plans for new garbage collector for LuaJit are one of the first designs that I've seen where all those pesky architectural requirements are addressed face on.
There's really no way to evaluate anything much in the store environment -- the most you can tell is that something is real crap, other than that you need it in your own place to check it out. They offer 30 day returns, and I did buy a $3k 5+1 system from them once, just to have it returned a couple weeks later. The bang for the buck was too little to warrant the expense -- we went with the wave system instead. Much smaller form factor, and the performance was more than adequate. You get quite a bit of sound from something that takes up less volume than a cuboid fitting an open 17" laptop (with screen straight up).
The hardest math courses I ever took was a series of hardcore numerical methods served by the math department, taught by a dyed-in-the-wool math professor who did know his math. You had as much materials available as you wanted to, and no exams -- just homework. I think 50% of the homework got you an A. The homework was easily 20 hours per week if you wanted it to be presentable and really something worth putting your name on. It was all about just getting the computations done, there was very little grunt style of theory -- it was all about analyzing the performance of the code as you implemented it, and getting it where it had to be. Of course you had to put down some equations to show why the code was performing the way it did, but that was really minor. Just applying the relatively simple math was lots of hard work -- I had an undergrad numerical methods course and it was a joke in comparison. I then took finite element modelling and had a good laugh at how they ignored the numerical stability aspects, and pretty much went with whatever Ansys had spat out. When you get to the nitty-gritty, applied math is hard. It's nothing like the purely theoretical pen-and-paper, armchair stuff some people excel at.
Myself, I only believe in physical security if you're attached to your kit, otherwise have sufficient insurance coverage and make sure the insurer is easy to work with. That's about it. Video security systems are useless when it comes to residential burglary. Not only are the cameras too low resolution to identify the suspects, a proper chain of custody needs to be followed as soon as the police comes to your place (if they even do). You can't just take the DVR and go to the police station.
That would be fine advice if it wasn't for how U.S. residential insurance works. They only pay amortized value of the old stuff upfront. They'll pay the balance once the work/replacement is done. This applies to most claims, AFAIK: burglary, weather, fire, etc.
I have their iPod dock, 2+1 computer speakers, and a wave sound system. The wave system will be 10 years old soon and has been problem-free and sounds great. The computer speakers perform very well, too. The iPod dock - same thing. I can't really say anything bad about this stuff. The only "problem" with the wave system was that the LCD backlight was flaky for 4 or 5 years, it started a couple years after purchase. It's been OK for 2 years now, though - it "fixed itself". I don't really know what people's problem with Bose is. We bought it because it made most sense compared to the form factor and features of other systems. Their UI is spartan, precisely the way I like it, I don't really need a 50 button UI and a lights show for a freakin' CD/amp/radio combination.
Lol, the same shit about grounded vs. ungrounded Faraday cages. Man, it makes absolutely, positively no difference at the frequencies involved. None whatsoever. Any grounding you would have is a big fucking inductor that is more of an open circuit at the frequencies involved, never mind probably a half-decent antenna as well. I have checked it for myself with WiFi, and it makes absolutely no measurable difference at all whether the Faraday cage is grounded or not. No matter how I'd measure things.
I agree, but it'd be wise to have it engineered before buying any components. Whether you do it on your own or hire a consultant RF engineer is of course up to you.
Exactly! And if one thinks that scientists don't take funding seriously, then well, one has no clue what it means to be a scientist these days. If you've got your Ph.D. and have subordinates, it's very likely that quite a bit of your time will be spent in various aspects of grant hunting and fund-schmoozing. It's a sad waste of brains, if you ask me, but that's how things are at the moment, at least in the U.S. The higher you go in responsibilities, the less time you'll have for science. Feynman knew exactly what he was doing when he abstained from all bureaucracy, meetings, etc.
People can fantasize all they want. Just because you have a flashy website doesn't mean there's anything of substance behind it. Just look at their timeline. It's a load of crap.
Oh give me a break, that's the second post here saying the same fantasy. They are not stupid, they know that there's no immortality on the table, just as they know that normal pace of medical progress can and does extend lives without them having to do anything special about it. They just want to support what's dear to them, in a way. Crossing the chasm between supporting life sciences and offering a "bid for immortality" requires a bit more support than a one liner post. Insightful, my ass. It's a troll post, that's all.
What I'm merely saying is that functional programming is more a matter of what the developer does, rather than what the language offers. Sure as heck I detest anything where the code is far flung from what is meant, so no, I do not advocate using plain C:)
It may be well-written code, but given that it's only quite recently that we have garbage collectors that can cope with pure-functional-style memory pressures, nobody in practice writes large LISP projects in functional style -- at least not projects that are not very recent. I can't wait for LuaJIT's new garbage collector to be done -- if it could be then "reshaped" to serve a LISP system like SBCL, it'd be a major win IMHO.
I wonder if they would have gotten better battery life by using a higher-powered ARM chip to emulate (JIT) the legacy firmware. It'd probably sleep most of the time.
I had a self-made circular slide rule in high school, then got a real one for college, just for the heck of it. Way easier to use than the linear one, IMHO. The self made one was, um, interestingly made. I scribed it on stainless steel, but based on a divider plotted on A3 paper using a Roland flat bed plotter. I think I wrote a QuickBasic program to generate the HPGL for the divider.
Slashdot Mirror
The drones don't track anything. I don't even know if they have an IMU. All position data is acquired using a video motion capture system. I'd think it may be Vicon with Tracker software.
They were flying these things apparently over people's heads in a public demo of putting together a brick "building". The room was full of spectators. It almost looked like a live art exhibit.
Yeah, troll, because quadcopters' only use is to develop military tech. Must have never heard of people studying control engineering. Ultimately, proof is in the pudding. Either you make a device that works using your control scheme, or it doesn't. Now fuck off.
If you're cash strapped, I don't see you exactly splurging on NI stuff. Microcontrollers with decent A/D and D/A converters are plenty cheap, same goes for discrete A/D and D/A chips, and all the other stuff you need for signal conditioning and digital control. Never mind who the heck cares what brand you use for control equipment? I think this guy is just trying to leverage familiar names to prop up his scam.
Alas, his scam doesn't mean the idea is unworkable, or that the serious people who are producing reproducible results are faking it.
You missed the key words:
This is far from a "Narrow Band" set of physical phenomena.
My problem is the opposite: it's not about the unnecessary stuff, it's about the cheap, non-invasive things they could do that they stay away from like if it was a transplant or something. My pregnant wife had to spend a couple of hours bleeding in the ER for a clueless resident to finally pick up the ultrasound and tell us "well, there's an apparently intact fetus with a beating heart, stuff happens, no need for a cleanup surgery, go home". They were already setting up an OR for a D&C -- what kind of an idiot does that before doing a basic ultrasound that takes 60 second start-to-finish?!
Another story: I have a solid family history (with tombstones and documented post-mortems to prove it) of plaque formation in coronary arteries, and the doofuses would drag their feet on noninvasive transthoracic ultrasounds of precisely the non-invasevely accessible parts of those arteries that were problematic in my family! I mean, come on, how stupid do you have to be.
Some guys I know decided they've had enough of bullshit and simply got an ultrasound for themselves to play with. It doesn't take 6 years of med school to do a decent doppler coronary artery exam. You simply need enough practice and access to rudimentary literature (and a modicum of intelligence). When you have the tool in your basement, you can get more hands-on time in a month than a resident gets in a year. Of course it helps if you're an engineer and can troubleshoot things and fix them when something breaks.
Tail recursion is not the problem I refer to. Merely the amount of garbage generated just making copies of stuff is what's killing it. Even Oracle's Java deals with it relatively poorly -- thus the presence of hardware-assisted garbage collection (at a price tag that'll make you weep). Open source implementations generally don't really have garbage collectors that address the architectures of today's CPUs. They are definitely not optimized for cache pressure, and are simply wasteful in their resource use. Even garbage collectors from Frantz and Chez Scheme seem to be relatively unimpressive when you cachegrind them. Mike Pall rather detailed plans for new garbage collector for LuaJit are one of the first designs that I've seen where all those pesky architectural requirements are addressed face on.
There's really no way to evaluate anything much in the store environment -- the most you can tell is that something is real crap, other than that you need it in your own place to check it out. They offer 30 day returns, and I did buy a $3k 5+1 system from them once, just to have it returned a couple weeks later. The bang for the buck was too little to warrant the expense -- we went with the wave system instead. Much smaller form factor, and the performance was more than adequate. You get quite a bit of sound from something that takes up less volume than a cuboid fitting an open 17" laptop (with screen straight up).
That guy has no background to pull it off. You'd need another Elon Musk who at least digs the technology.
The hardest math courses I ever took was a series of hardcore numerical methods served by the math department, taught by a dyed-in-the-wool math professor who did know his math. You had as much materials available as you wanted to, and no exams -- just homework. I think 50% of the homework got you an A. The homework was easily 20 hours per week if you wanted it to be presentable and really something worth putting your name on. It was all about just getting the computations done, there was very little grunt style of theory -- it was all about analyzing the performance of the code as you implemented it, and getting it where it had to be. Of course you had to put down some equations to show why the code was performing the way it did, but that was really minor. Just applying the relatively simple math was lots of hard work -- I had an undergrad numerical methods course and it was a joke in comparison. I then took finite element modelling and had a good laugh at how they ignored the numerical stability aspects, and pretty much went with whatever Ansys had spat out. When you get to the nitty-gritty, applied math is hard. It's nothing like the purely theoretical pen-and-paper, armchair stuff some people excel at.
Myself, I only believe in physical security if you're attached to your kit, otherwise have sufficient insurance coverage and make sure the insurer is easy to work with. That's about it. Video security systems are useless when it comes to residential burglary. Not only are the cameras too low resolution to identify the suspects, a proper chain of custody needs to be followed as soon as the police comes to your place (if they even do). You can't just take the DVR and go to the police station.
That would be fine advice if it wasn't for how U.S. residential insurance works. They only pay amortized value of the old stuff upfront. They'll pay the balance once the work/replacement is done. This applies to most claims, AFAIK: burglary, weather, fire, etc.
I have their iPod dock, 2+1 computer speakers, and a wave sound system. The wave system will be 10 years old soon and has been problem-free and sounds great. The computer speakers perform very well, too. The iPod dock - same thing. I can't really say anything bad about this stuff. The only "problem" with the wave system was that the LCD backlight was flaky for 4 or 5 years, it started a couple years after purchase. It's been OK for 2 years now, though - it "fixed itself". I don't really know what people's problem with Bose is. We bought it because it made most sense compared to the form factor and features of other systems. Their UI is spartan, precisely the way I like it, I don't really need a 50 button UI and a lights show for a freakin' CD/amp/radio combination.
Someone mod this to +5 informative. It makes most other rambling above completely pointless.
Lol, the same shit about grounded vs. ungrounded Faraday cages. Man, it makes absolutely, positively no difference at the frequencies involved. None whatsoever. Any grounding you would have is a big fucking inductor that is more of an open circuit at the frequencies involved, never mind probably a half-decent antenna as well. I have checked it for myself with WiFi, and it makes absolutely no measurable difference at all whether the Faraday cage is grounded or not. No matter how I'd measure things.
I agree, but it'd be wise to have it engineered before buying any components. Whether you do it on your own or hire a consultant RF engineer is of course up to you.
Exactly! And if one thinks that scientists don't take funding seriously, then well, one has no clue what it means to be a scientist these days. If you've got your Ph.D. and have subordinates, it's very likely that quite a bit of your time will be spent in various aspects of grant hunting and fund-schmoozing. It's a sad waste of brains, if you ask me, but that's how things are at the moment, at least in the U.S. The higher you go in responsibilities, the less time you'll have for science. Feynman knew exactly what he was doing when he abstained from all bureaucracy, meetings, etc.
People can fantasize all they want. Just because you have a flashy website doesn't mean there's anything of substance behind it. Just look at their timeline. It's a load of crap.
Oh give me a break, that's the second post here saying the same fantasy. They are not stupid, they know that there's no immortality on the table, just as they know that normal pace of medical progress can and does extend lives without them having to do anything special about it. They just want to support what's dear to them, in a way. Crossing the chasm between supporting life sciences and offering a "bid for immortality" requires a bit more support than a one liner post. Insightful, my ass. It's a troll post, that's all.
What I'm merely saying is that functional programming is more a matter of what the developer does, rather than what the language offers. Sure as heck I detest anything where the code is far flung from what is meant, so no, I do not advocate using plain C :)
It may be well-written code, but given that it's only quite recently that we have garbage collectors that can cope with pure-functional-style memory pressures, nobody in practice writes large LISP projects in functional style -- at least not projects that are not very recent. I can't wait for LuaJIT's new garbage collector to be done -- if it could be then "reshaped" to serve a LISP system like SBCL, it'd be a major win IMHO.
This is the post of the day for me. Couldn't agree more!
I wonder if they would have gotten better battery life by using a higher-powered ARM chip to emulate (JIT) the legacy firmware. It'd probably sleep most of the time.
I had a self-made circular slide rule in high school, then got a real one for college, just for the heck of it. Way easier to use than the linear one, IMHO. The self made one was, um, interestingly made. I scribed it on stainless steel, but based on a divider plotted on A3 paper using a Roland flat bed plotter. I think I wrote a QuickBasic program to generate the HPGL for the divider.
Yeah, because the SAT is the only thing one ever buys a calculator for. If you're good at it, you can get decent score on SAT without a calculator.