Impossible? Impossible why? I don't see why this would be the case. In fact, I imagine that with minimum wiring you could run two 16bit DACs in parallel, one handling the top 16 bits at twice the voltage, the other handling the low 16.
You mean the top 16 bits at 65536x the voltage, and the other handling the low 16. Else you've just produced a 17 bit DAC.
And again you can't make up your mind... Are humans robots or are they valuable troubleshooters?
I'd rather write off the 20% of experiments on which the automation fails-- to launch them again later.. than launch a lousy troubleshooting robot that weighs 100kg and consumes 40kg of materials per day. You can fit a lot of interesting experiments in that 100kg.
That was true. But the EXPRESS system for ISS, as well as outfitting it with general purpose lab benches rather than special purpose instruments is changing that.
I don't know how I feel about ISS. I'm pretty sure right now it's worthless with the degree of overhead. Maybe when completed it will be scientifically useful. Unfortunately, the orbit it is in is poorly suited for launching missions to escape velocity.
True, but confused. The Apollo astronauts were not really doing analysis, but rather the vastly more bounded problem of sample collection. The real geology,...
Sure, it took till Apollo 17 for a real geologist to fly. I just feel, based on how lousy autonomous robots are on earth at picking routes and doing things in known, mapped terrain.. That it would be impossible to build a robot to show the adaptability I've seen in the Apollo EVA transcripts. And with fundamental limitations on the number of pictures that robots can return governed by communications limitations.. (High path loss, high noise, small communications windows, long speed of light delay)... there's a very small number of things that can be studied in detail per mission. No current robotic probe is going to decide on its own "Hey, that rock I just passed, it was kinda shiny and looked like glass!" like the Apollo astronauts did.
I'm all for the robotic technology getting better. At the same time, I want us to solve the engineering problems of building things in space and living other places like the moon for a long time. The science will come cheaply as a byproduct after that. And I have an idea by that point we'll be wiser as to what to study.
The speed of light delay to LEO is trivial, and the problems of telepresence for a particular scientific experiemnt are much better defined. In short, telepresence/robotic systems are likely to work a lot better.
It is my belief that with the experiment load carried upon the Shuttle science missions, the people are acting pretty much like robots. Sure, I've seen cases when shuttle/space station astronauts have successfully troubleshooted experiments and have gotten them to run. But despite the often-heard premise from space science advocates on the value of spontaneous human observations, I have not seen a case of an experiment being adjusted based on an unusual observation on a first run and a significant scientific find occuring (perhaps you can inform me). And indeed, because of the expensive cost of launching payload to orbit the amount of instrumentation and raw materials for launching a spontaneous follow-up experiment is rather limited.
On the other hand, it's pretty clear with Apollo that the bulk of the science happened with manned lunar missions-- in number of geologically signfiicant photographs, quantity and quality of sample return, etc etc etc.
The fact is, running a particular experiment (watch the spiders in a cage-- what kind of webs do they weave?) is an infinitely more bounded problem than performing geology on another world (let's figure out what here is like earth, and what is different, and try and figure out what processes caused things to be this way).
Look at how well telepresence has worked with the Hubble Space Telescope. There's little functional difference in getting your experiment approved to run on a large telescope on Earth and receiving the data back, and the same process on Hubble. The speed of light, path loss, and orbital mechanics ensure that no matter how good your instrumentation is on another world, you won't have real telepresence.
It's a sad state of affairs when people actually start to believe that robots could ever replace human explorers.
I agree. We need to go to Mars and learn about other worlds.
The sad thing is, all that we've been spending on "space science" all this time in LEO... IMO, a large amount of it could be done better and cheaper by robots. All of these "what patterns of ash does fire make on this glass" and "how does this kind of crystal grow" could be cheaply put into space by Delta IV or Ariane and observed for long periods. To space science in LEO, I say "good riddance". It's too bad Hubble will be a casualty but tradeoffs need to be made for the job to get done. IMO, what we need to do in LEO is focus on one thing-- human longevity and health in micro/low gravity.
But when it comes to exploring other worlds, humans are definitely best. Even though the instrumentation/actuation for telepresence has improved to perhaps be on the level with humans, the speed of light and limited communications bandwidth mean that human intelligence can't act on many observations in a mission.
Each shuttle flight costs about $450 million. The Columbia Accident Investigation Board has also set special requirements for flights that don't allow an ISS "lifeboat" option-- so a special one-time-use tile repair kit would have to be built and certified to comply with the CAIB. And of course, there's the other $40 million in instrumentation development/certification for the servicing mission.
With the focus on retiring the shuttle to permit construction of the Crew Exploration Vehicle, this really doesn't make sense. The James Webb telescope will be up soon enough, and it is ilkely that the life of Hubble can be extended with special 2-gyro + reference star stabilization software (albeit, with slightly degraded image quality / pointing ability).
The MERs use X-Band for high data rate communications back to earth-- which has a wavelength of 3cm, making high gain antennas considerably smaller and more practical.
It's my understanding that the high gain antenna on MER is a compact phased array design. Even parabolic antennas could be practical at the 3cm wavelength, though they wouldn't be flat (which was obviously preferable for footprint issues).
No, it means it has 8 cylinders, 4 per bank, in a V arrangement. They're not in a long straight line, like cars with inline-4's, and they're not horizontally opposed (boxers) like some Porsche or aircraft engines.
Yes. This is very common. You don't want to make things that are direct threats to litigation before a suit is filed; else it's likely that the other party can bring a suit for declaratory judgment (DJ) in the venue of their choice.
That would be 129,600,000,000. The missing zero is why you've caused a shitstorm. Check your math next time.
(And does it really sound logical to you that you can only download 1.5GB in a day with a 1.5mbit connection? Takes me a little over a couple hours...)
Do you know what a Mexican drug dealer will do? Well, they will get some white guy to move the drugs across the airport. Instead of the Mexican carrying the drugs, he will pay some white person to carry it across customs. Once through, he/she will pick it up themselves. If the police does spend a ton of resources profiling the Mexicans, they will likely let the white trafficker through. That is why profiling does not work.
Brilliant. The terrorists need to pay me so I can conduct a suicide attack or two. Al qaeda has millions, right? Just think what I could do with that cash.
Oh wait, that doesn't work-- because I'd be fucking dead. Maybe they'll have trouble finding someone to provide such services.
Agreed on all points. Yes, digital reception falls off much quicker than analog. But by the time digital drops, AMPS would have been unusable for a long time. Not to mention the benefits of adaptive transmit power on battery life, etc etc etc.
Phone codecs have gotten a lot better at rejecting background noise and sending just speech. But yes, that needs to improve, too. There are problems with the user experience-- one of my brothers always talks very loudly into his cellular phone... causing clipping and all kinds of harmonics to go to the codec, making him unintelligible even when there's tons of signal strength. So things like that need to get better. But all in all, a cellphone isn't a device to send music.. it's intended to send speech from one or two people in conversational tones.
The fact is, with our limits in battery density, spectrum, and antenna technology... you can't have it all. Perhaps intelligent cell sites with beam steering phased arrays will mitigate some of these trade-offs, but it's not happening anytime soon.
Yes, you can tune in plain FM stations with a HD-FM receiver.
BTW, digital cellular is popular with the carriers not only because of spectrum efficiency, but because of superior link budgeting with lower output power. The range is actually better on digital cellular protocols (whether TDMA or CDMA) than FDMA/AMPS. The reason why your user experience is better with analog is that there is so much more analog stuff deployed. This is likely to change (not exactly a ton of AMPS equipment is still getting deployed).
Data compression reduces signal bandwidth. And reduced bandwidth means less noise in the band where the signal is, and also means that the signal, since it is less wide, is stronger. This translates to better S/N and thus better link budget. Also, there are things like coding gain which you can't make use of with analog transmissions.
I don't know how the HD FM divides output power/spectrum to the subcarriers. But it is likely that you can still get a perfectly clear digital signal when the analog FM station would be unlistenable.
No shit sherlock. That's why I said on most airliners.
And you do realize that there are multiple systems, running in parallel, over multiple communications buses, with fault-isolation models, right? FBW planes are generally acknowledged to have a higher control-system reliability than conventional control-system couplings.
Sure, the software has to be reliable. But the certification requirements are intense, with all of the code getting exercised and intensely inspected.
I doubt this. I've seen plenty of evidence for EMI causing difficulty with reception of navaids in flight.
But given that on most airliners the primary actuators for control surfaces are hydraulic and hooked up directly to control cables run from the cockpit... it's hard to see how this would occur. The actuators used for autopilots are also purposely weak and generally can't accomplish full control surface deflections in flight.
Note that "full rudder deflection in cruise" would most likely mean "tail departing from the airplane and subsequent in flight breakup". Demonstrating that a full control surface deflection can be tolerated in cruise is not part of certification requirements. It's quite possible to pull the wings or tail off almost any airplane.
And by definition-- a failure of control systems will not result in CFIT.
And look at how now you've changed from defending your point to calling me a troll-- when it's obvious that you didn't know what you were talking about. This is the best evidence of someone with a weak argument.
Hopefully you do realize that the total amount of delta V necessary to execute a transfer ellipse maneuver decreases with the time over which the maneuver is executed.
Also, gravitational transfer maneuvers can create high perceived accelerations on a spacecraft.
Higher total delta V's generally also indicate a higher degree of precision required in executing burns. Hopefully you understand/can reason out why.
But all of this is mostly irrelevent. The hard bit of Mars is entry. Sure, Venus's atmosphere makes probes tend not to last long. But Mar's thin atmosphere requires exotic systems for a lander beyond simple drag parachutes in all realistic entry scenarios.
I Think You Don't Know What You're Talking About (tm).
Shuttle peak acceleration during launch is 3G. Both Soyuz and Apollo are typically under 6G.
There is so much more wrong with your post that I prefer not to comment any further. Please claim a clue before acting authoritatively on these topics again.
When you adjust for the time value of money, though, it's not nearly so bad, and the fact that the mortgage is tax deductible..
Say he's paying 6% on that home loan, and he's in a 35% total tax bracket. That makes his effective interest rate 3.9%. After inflation, that makes the amount of interest he's paying around 2% effectively. Over the life of the loan, that results in about a 33% premium that he pays the bank over the value of the home. Not a bad deal, all in all, when you also get the house to live in, in the meantime (eliminating the need to pay rent). All in all, mortgages right now are an especially good deal. The numbers I used are conservative, too. Many people pay more in total tax than 35%, and we'll be lucky if inflation stays around 2% for the next 30 years.
He's talking about a system that uses a riser so that the PCI/AGP card mounts parallel to the motherboard. These risers usually end up putting the card opposite-side-up than they'd be in a tower case.
That's a 25% gain at best in performance.. and real world results will be less.. Most of the benchmarks I've seen give the P4 3GHz a 10-20% edge over the P4 2.4.
Maybe overclocking and hassling with things makes that kind of gain worthwhile. It isn't for me. It's not like back in the day where with the Celeron 300A you could get 50% performance boosts.
In my haste to reply I phrased my reply incorrectly.
The only time you will see -more duplicates than you'd expect on a typical hash algorithm- (e.g. not all man values would be duplicated, for 2^56 inputs and 2^56 outputs) is because of truncation effects.
My point was to say it's not like if you had 2^56 inputs and 2^48 outputs, where there would be a little more than 256 duplicates per input value.
Slashdot Mirror
Impossible? Impossible why? I don't see why this would be the case. In fact, I imagine that with minimum wiring you could run two 16bit DACs in parallel, one handling the top 16 bits at twice the voltage, the other handling the low 16.
You mean the top 16 bits at 65536x the voltage, and the other handling the low 16. Else you've just produced a 17 bit DAC.
And again you can't make up your mind... Are humans robots or are they valuable troubleshooters?
I'd rather write off the 20% of experiments on which the automation fails-- to launch them again later.. than launch a lousy troubleshooting robot that weighs 100kg and consumes 40kg of materials per day. You can fit a lot of interesting experiments in that 100kg.
That was true. But the EXPRESS system for ISS, as well as outfitting it with general purpose lab benches rather than special purpose instruments is changing that.
I don't know how I feel about ISS. I'm pretty sure right now it's worthless with the degree of overhead. Maybe when completed it will be scientifically useful. Unfortunately, the orbit it is in is poorly suited for launching missions to escape velocity.
True, but confused. The Apollo astronauts were not really doing analysis, but rather the vastly more bounded problem of sample collection. The real geology,...
Sure, it took till Apollo 17 for a real geologist to fly. I just feel, based on how lousy autonomous robots are on earth at picking routes and doing things in known, mapped terrain.. That it would be impossible to build a robot to show the adaptability I've seen in the Apollo EVA transcripts. And with fundamental limitations on the number of pictures that robots can return governed by communications limitations.. (High path loss, high noise, small communications windows, long speed of light delay)... there's a very small number of things that can be studied in detail per mission. No current robotic probe is going to decide on its own "Hey, that rock I just passed, it was kinda shiny and looked like glass!" like the Apollo astronauts did.
I'm all for the robotic technology getting better. At the same time, I want us to solve the engineering problems of building things in space and living other places like the moon for a long time. The science will come cheaply as a byproduct after that. And I have an idea by that point we'll be wiser as to what to study.
The speed of light delay to LEO is trivial, and the problems of telepresence for a particular scientific experiemnt are much better defined. In short, telepresence/robotic systems are likely to work a lot better.
It is my belief that with the experiment load carried upon the Shuttle science missions, the people are acting pretty much like robots. Sure, I've seen cases when shuttle/space station astronauts have successfully troubleshooted experiments and have gotten them to run. But despite the often-heard premise from space science advocates on the value of spontaneous human observations, I have not seen a case of an experiment being adjusted based on an unusual observation on a first run and a significant scientific find occuring (perhaps you can inform me). And indeed, because of the expensive cost of launching payload to orbit the amount of instrumentation and raw materials for launching a spontaneous follow-up experiment is rather limited.
On the other hand, it's pretty clear with Apollo that the bulk of the science happened with manned lunar missions-- in number of geologically signfiicant photographs, quantity and quality of sample return, etc etc etc.
The fact is, running a particular experiment (watch the spiders in a cage-- what kind of webs do they weave?) is an infinitely more bounded problem than performing geology on another world (let's figure out what here is like earth, and what is different, and try and figure out what processes caused things to be this way).
Look at how well telepresence has worked with the Hubble Space Telescope. There's little functional difference in getting your experiment approved to run on a large telescope on Earth and receiving the data back, and the same process on Hubble. The speed of light, path loss, and orbital mechanics ensure that no matter how good your instrumentation is on another world, you won't have real telepresence.
It's a sad state of affairs when people actually start to believe that robots could ever replace human explorers.
I agree. We need to go to Mars and learn about other worlds.
The sad thing is, all that we've been spending on "space science" all this time in LEO... IMO, a large amount of it could be done better and cheaper by robots. All of these "what patterns of ash does fire make on this glass" and "how does this kind of crystal grow" could be cheaply put into space by Delta IV or Ariane and observed for long periods. To space science in LEO, I say "good riddance". It's too bad Hubble will be a casualty but tradeoffs need to be made for the job to get done. IMO, what we need to do in LEO is focus on one thing-- human longevity and health in micro/low gravity.
But when it comes to exploring other worlds, humans are definitely best. Even though the instrumentation/actuation for telepresence has improved to perhaps be on the level with humans, the speed of light and limited communications bandwidth mean that human intelligence can't act on many observations in a mission.
Each shuttle flight costs about $450 million. The Columbia Accident Investigation Board has also set special requirements for flights that don't allow an ISS "lifeboat" option-- so a special one-time-use tile repair kit would have to be built and certified to comply with the CAIB. And of course, there's the other $40 million in instrumentation development/certification for the servicing mission.
With the focus on retiring the shuttle to permit construction of the Crew Exploration Vehicle, this really doesn't make sense. The James Webb telescope will be up soon enough, and it is ilkely that the life of Hubble can be extended with special 2-gyro + reference star stabilization software (albeit, with slightly degraded image quality / pointing ability).
Yes, there are plenty of advanced technology phased array antennas where delay lines and elements are dynamically swapped in and out.
But it is possible to build a statically aimed phased array-- and this is what most patch antennas are in practice.
The MERs use X-Band for high data rate communications back to earth-- which has a wavelength of 3cm, making high gain antennas considerably smaller and more practical.
It's my understanding that the high gain antenna on MER is a compact phased array design. Even parabolic antennas could be practical at the 3cm wavelength, though they wouldn't be flat (which was obviously preferable for footprint issues).
Not to be pedantic.. but..
But V8 means it has 8 valves..
No, it means it has 8 cylinders, 4 per bank, in a V arrangement. They're not in a long straight line, like cars with inline-4's, and they're not horizontally opposed (boxers) like some Porsche or aircraft engines.
Yes. This is very common. You don't want to make things that are direct threats to litigation before a suit is filed; else it's likely that the other party can bring a suit for declaratory judgment (DJ) in the venue of their choice.
[iguana:~] mlyle% bc
1.5*60*60*24
129600.0
That would be 129,600,000,000. The missing zero is why you've caused a shitstorm. Check your math next time.
(And does it really sound logical to you that you can only download 1.5GB in a day with a 1.5mbit connection? Takes me a little over a couple hours...)
Do you know what a Mexican drug dealer will do? Well, they will get some white guy to move the drugs across the airport. Instead of the Mexican carrying the drugs, he will pay some white person to carry it across customs. Once through, he/she will pick it up themselves. If the police does spend a ton of resources profiling the Mexicans, they will likely let the white trafficker through. That is why profiling does not work.
Brilliant. The terrorists need to pay me so I can conduct a suicide attack or two. Al qaeda has millions, right? Just think what I could do with that cash.
Oh wait, that doesn't work-- because I'd be fucking dead. Maybe they'll have trouble finding someone to provide such services.
Straw.. man... argument.
Agreed on all points. Yes, digital reception falls off much quicker than analog. But by the time digital drops, AMPS would have been unusable for a long time. Not to mention the benefits of adaptive transmit power on battery life, etc etc etc.
Phone codecs have gotten a lot better at rejecting background noise and sending just speech. But yes, that needs to improve, too. There are problems with the user experience-- one of my brothers always talks very loudly into his cellular phone... causing clipping and all kinds of harmonics to go to the codec, making him unintelligible even when there's tons of signal strength. So things like that need to get better. But all in all, a cellphone isn't a device to send music.. it's intended to send speech from one or two people in conversational tones.
The fact is, with our limits in battery density, spectrum, and antenna technology... you can't have it all. Perhaps intelligent cell sites with beam steering phased arrays will mitigate some of these trade-offs, but it's not happening anytime soon.
Yes, you can tune in plain FM stations with a HD-FM receiver.
BTW, digital cellular is popular with the carriers not only because of spectrum efficiency, but because of superior link budgeting with lower output power. The range is actually better on digital cellular protocols (whether TDMA or CDMA) than FDMA/AMPS. The reason why your user experience is better with analog is that there is so much more analog stuff deployed. This is likely to change (not exactly a ton of AMPS equipment is still getting deployed).
Data compression reduces signal bandwidth. And reduced bandwidth means less noise in the band where the signal is, and also means that the signal, since it is less wide, is stronger. This translates to better S/N and thus better link budget. Also, there are things like coding gain which you can't make use of with analog transmissions.
I don't know how the HD FM divides output power/spectrum to the subcarriers. But it is likely that you can still get a perfectly clear digital signal when the analog FM station would be unlistenable.
Especially since we need to figure out what country you're in, to know what to convert from.
Would adding a UKP or pounds on the end be so difficult? This is an international site, it's not like it's obvious where you're from.
No shit sherlock. That's why I said on most airliners.
And you do realize that there are multiple systems, running in parallel, over multiple communications buses, with fault-isolation models, right? FBW planes are generally acknowledged to have a higher control-system reliability than conventional control-system couplings.
Sure, the software has to be reliable. But the certification requirements are intense, with all of the code getting exercised and intensely inspected.
I doubt this. I've seen plenty of evidence for EMI causing difficulty with reception of navaids in flight.
But given that on most airliners the primary actuators for control surfaces are hydraulic and hooked up directly to control cables run from the cockpit... it's hard to see how this would occur. The actuators used for autopilots are also purposely weak and generally can't accomplish full control surface deflections in flight.
Note that "full rudder deflection in cruise" would most likely mean "tail departing from the airplane and subsequent in flight breakup". Demonstrating that a full control surface deflection can be tolerated in cruise is not part of certification requirements. It's quite possible to pull the wings or tail off almost any airplane.
And by definition-- a failure of control systems will not result in CFIT.
no one listens to a self-righteous asshole.
:)
You obviously do.
And look at how now you've changed from defending your point to calling me a troll-- when it's obvious that you didn't know what you were talking about. This is the best evidence of someone with a weak argument.
Nah, I'm just the kind of person who doesn't like people making authoritative statements about stuff they know nothing about.
Because then those authoritative statements will get repeated again, and again, yielding the ignorant masses.
My point on your "delta v" argument remains.
Not my argument. Look at the post history.
Hopefully you do realize that the total amount of delta V necessary to execute a transfer ellipse maneuver decreases with the time over which the maneuver is executed.
Also, gravitational transfer maneuvers can create high perceived accelerations on a spacecraft.
Higher total delta V's generally also indicate a higher degree of precision required in executing burns. Hopefully you understand/can reason out why.
But all of this is mostly irrelevent. The hard bit of Mars is entry. Sure, Venus's atmosphere makes probes tend not to last long. But Mar's thin atmosphere requires exotic systems for a lander beyond simple drag parachutes in all realistic entry scenarios.
I Think You Don't Know What You're Talking About (tm).
Shuttle peak acceleration during launch is 3G. Both Soyuz and Apollo are typically under 6G.
There is so much more wrong with your post that I prefer not to comment any further. Please claim a clue before acting authoritatively on these topics again.
When you adjust for the time value of money, though, it's not nearly so bad, and the fact that the mortgage is tax deductible..
Say he's paying 6% on that home loan, and he's in a 35% total tax bracket. That makes his effective interest rate 3.9%. After inflation, that makes the amount of interest he's paying around 2% effectively. Over the life of the loan, that results in about a 33% premium that he pays the bank over the value of the home. Not a bad deal, all in all, when you also get the house to live in, in the meantime (eliminating the need to pay rent). All in all, mortgages right now are an especially good deal. The numbers I used are conservative, too. Many people pay more in total tax than 35%, and we'll be lucky if inflation stays around 2% for the next 30 years.
Uhmm, it'll be chip up in almost any tower. ;P
;P
He's talking about a system that uses a riser so that the PCI/AGP card mounts parallel to the motherboard. These risers usually end up putting the card opposite-side-up than they'd be in a tower case.
Think before you speak, next time.
I don't think he's out of touch.
That's a 25% gain at best in performance.. and real world results will be less.. Most of the benchmarks I've seen give the P4 3GHz a 10-20% edge over the P4 2.4.
Maybe overclocking and hassling with things makes that kind of gain worthwhile. It isn't for me. It's not like back in the day where with the Celeron 300A you could get 50% performance boosts.
In my haste to reply I phrased my reply incorrectly.
The only time you will see -more duplicates than you'd expect on a typical hash algorithm- (e.g. not all man values would be duplicated, for 2^56 inputs and 2^56 outputs) is because of truncation effects.
My point was to say it's not like if you had 2^56 inputs and 2^48 outputs, where there would be a little more than 256 duplicates per input value.
Most of the time DES crypt will indeed be 1-to-1.
RTFA. This is a contract dispute. It is spelled out in their contract, and a few operators have refused to sign it.
The driver is free to turn off the GPS tracking on the cell phone when they're not on duty.