The main thing for which the the Shuttle is a "proven delivery system" is the transportation of breathtaking amounts of taxpayer money to a cabal of well-connected aerospace contractors.
Aside from that one feature, the design, capabilities and risk profile of the Shuttle launch system make almost zero sense.
Microsoft was very careful not to infringe on others' IP on this point. They made certain that every blank line in their system was different from any blank line in Unix. I applaud them for that.
Hmmm... I'd say that you can't be completely sure about what really happens to those parts unless you hitch a ride on the back of one of the delivery trucks like Charlton Heston in Soylent Green.
You mean to tell me that Alpha Decay is rare in the universe? I simply don't buy the argument.
Alpha decay generally happens to elements heavier than lead. Those elements are only created as a small side reaction in supernova explosions. Only a fraction of matter is in stars, and only a fraction of stars become supernovas, and only a small fraction of the matter in a supernova becomes heavy elements. Relative to the total matter in the universe, alpha decay is in the parts-per-billion category. In particular, the abundance of helium is NOT due to alpha decay.
The Earth is not so special that we can have zero Helium.
It is pretty special. The only thing that holds helium over the long term is gravity, and earth just doesn't have enough of it. The only place we can get at abundant helium is gas giant planets, and we don't have any technology in the foreseeable future to lift anything out of those gravity wells.
Arguing that helium is abundant in the universe therefore there must be plenty on earth is silly. Using that logic, I could say that a much, if not most, of the planetary mass in our solar system is in the form of metallic hydrogen. Therefore metallic hydrogen is available for us to use here.
I understand where people are coming from when they warn of this kind of stuff, but LONG term this stuff resupplies at a pretty decent rate. Hence the reason He is the second most abundant element in the universe.
The actual reason He is the second most abundant element in the universe is that huge amounts of it were formed in the first moments of the Big Bang. A little more has been formed since then by fusion in stars. Unfortunately, essentially none of the helium from either of those sources has stayed put on earth. It all floated away long ago.
Helium created by decay of heavy elements in incredibly rare in the universe, and it's rare on the earth as well, but it's the only helium we can get at. It forms at a rate that's way too low and too diluted for us to use. It has accumulated over millions of years in the same geological structures that capture natural gas, but those special traps certainly aren't being replenished fast enough for our needs.
*All* processor architectures are putting lipstick on pigs. Every so often a new "clean" RISC architecture comes out where the instruction set is supposed to look like the actual hardware. Then within a couple of CPU generations, the hardware landscape shifts and they slap ugly layers of abstraction between their backwards compatible instruction set and the actual new hardware. (Many years ago I listened to a pitch from a guy at MIPS, who was touting their chip named for "Microprocessor without Interlocked Pipeline Stages". Guess what they added to the next version of the processor: Interlocked pipeline stages!)
Intel themselves tried the hardest to get out of this cycle by making the hardware extremely visible to the instruction set with Itanium, trusting compiler technology to handle the resulting morass. Result of this experiment: Epic fail.
At the end of the day, X86-derivatives run at speeds in the same ballpark as any other CPU architecture that can be programmed with real-world tools by real-world coders, and they do it at a fraction of the cost. Why break all of the code out there if there's no big payback?
In the low performance efficiency market, ARM currently has advantages over X86. But by the time they bloat up ARM with a few more generations of "innovation" to get into the X86 performance range, it probably won't have that much of an advantage in size or power.
that I asked myself when I read the GPL. why the FUCK doesn't Stallman communicate directly and get away from the obsufcated communication style that he uses.
If the GPL causes you so much distress, I suggest you go try to read a typical proprietary EULA. Then maybe that will cause your head to explode and we'll all be better off.
MMX and related instructions are handled by a coprocessor.
Look at modern CPU architecture diagrams. The media units are no more of a coprocessor than any other functional unit. In particular, the instructions are decoded and run in the same pipelines as everything else. But I suppose if you choose to define any data path that doesn't match the pointer size as a "coprocessor", then we're both right.
String handling works twice as fast on 64 bit architectures as on 32 bit ones,
I think that assertion is totally false. One of the biggest bottlenecks on performance of modern CPUs on this kind of workload is accurate branch prediction, which bit size does almost nothing to help. Just block moving strings around is not an issue, and even if it were, newer 32-bit X86 processors have special tricks using a specially crafted REP MOVSD operations that can move entire cache lines in one clock.
There is no sense to having 32 bit components to a 64 bit OS, unless they are necessary for the support of 32 bit applications.
But you're not looking at it from the standpoint of a software vendor, who are the ones who actually provide the components. They have to release the 32-bit code anyway for the foreseeable future, and managing two separate releases is a lot more work for them in return for the single-digit performance increases from lower register pressure. Users most likely don't perceive such small performance increments, so a 64-bit release gets dropped in favor of more crucial work.
Whatever. If you can find a popular "64-bit architecture" that only supports 32-bit pointers, feel free to point it out.
My point was: browsers do not need either 64-bit address pointers nor 64-bit integers for performance. They can already operate on 128 bits of data (split into several smaller numbers) per clock cycle on today's commodity 32-bit processors to do multimedia processing and graphics. The rest of what they do is mainly string processing and compiling, both of which run just fine in 32-bit mode.
Actually, the "bit size" of a CPU generally refers to its address pointer size. The 32-bit x86 has been able to deal with 64-bit chunks of integer data since the MMX was introduced in 1996, and now it does 128 bits at a time with SSE. Multimedia operations that are be done in a browser process don't usually benefit from working on big numbers, but they do benefit from crunching lots of smaller numbers at one time. That's what these SIMD extensions do, and a "64-bit" architecture isn't necessary for that.
I'm shocked that it's been almost 2 hours since you commented, and NOBODY has butted in to spout some nonsense about how the manned program is unnecessary because it would have just been cheaper to launch a new Hubble.
It's not nonsense. It would have indeed been cheaper (and safer) to launch new Hubbles for each and every time it got "fixed". It's called assembly line production: Each additional Hubble gets cheaper. (This is also why disposable rockets are cheaper than doing a custom overhaul of the space shuttle after each launch.)
Notice that the National Reconnaissance Office, which has a lot more experience operating Hubble equivalents in orbit than NASA, doesn't waste time trying to send humans up to fix things.
Had we gone the more sensible route, we also wouldn't have been subjected to endless Discovery Channel documentaries about how Story Musgrave and his friends had to spend 8-hour shifts trying to twist stuck screws in zero-G.
Ok, but also can't ignore the energy cost to produce and transport the gas, coal and/or uranium to the power plant to generate the electricity. It probably comes out similar to that of delivering gas to your house, so it's not really an issue specific to gas heat.
What I was originally talking about was the excess energy needed to produce electricity due to the laws of thermodynamics as they relate to heat engines. That assumes that the fuel is already available at the power plant. Factoring in additional energy needed to get the fuel would just make electricity look even worse.
My plasma and electronic gizmos heat is not wasted one bit.
On average, for every watt of electric power delivered to your house, 2 more watts goes up the smokestacks and cooling towers of your power plant. Those 2 watts are still wasted. (That's why the only cost effective way to heat with electricity is a heat pump, which delivers multiple watts of heating for every watt of electricity consumed.) If, OTOH, you used a modern gas furnace to get that same heating, you'd waste less than 10% of the energy in the fuel source.
Running power hungry appliances to create heat is wasteful and does not make economic sense.
You can only drink so much undistilled beer before evacuating
OTOH, given the typical price of these esoteric products, you can only drink so much "distilled beer" before your wallet is evacuated. That will tend to limit your ability to over-consume any toxins.
There's no fundamental reason why you couldn't orbit within the atmosphere at Mach 22, with the occupants experiencing weightlessness. We don't currently have the materials and engines to withstand such conditions, nor would it make any sense to try, but it could be done in theory.
Anyway back to the OP's point, since kinetic energy is proportional to the square of the velocity, Mach 6 is only 7% of the energy needed to reach orbit. IMO, scramjets are just a complex diversion if the goal is to go into orbit. (And you need to first reach supersonic speeds by yet some other means before you can even switch them on.) Reaching Mach 6 with an ordinary booster rocket stage is child's play in comparison.
From what I've read about PLATO (I was born quite a bit after PLATO's heyday) it seemed to be in stark contrast with today's methods of teaching computers. It seemed like PLATO actually encouraged students to explore computers. Today though, teachers are too paranoid, thinking that the command prompt will "break" the computer and other stupidities.
When I was a student back in the days of PLATO, I had a part time job as the human tutor in one of the PLATO terminal rooms. I don't remember it being focused at all on exploring computers. The system was all about the pre-canned apps. In fact, my memory is a little rusty, but I don't recall that they really had a command prompt at all, at least as far as end users were concerned. I think it was all a hierarchical full screen menu-driven system. (I assume that some CS majors were taught how to write software for PLATO, but that would be a small minority of the users.)
One problem with the course that I worked with was that the software was a bit too linear and inflexible. For example, students weren't allowed to go on to the next problem until they correctly answered the current one, and the range of acceptable answers was usually very constrained. The software basically kept repeating: "Wrong. Try again.", and you were stuck at a dead end.
Unfortunately, back in those days this was often the first exposure the users had to a computer system of any kind. They had never experienced anything as exacting and unforgiving as a computer, and it didn't help to heap that on top of the inherent stress of a "weed-out" engineering class. That's why they needed me to be in there as a backup; I think that some of the people would have eventually gone postal on the terminals if they didn't have access to someone who could see how and why they were stuck, and dole out helpful hints.
In a nutshell: It was a preview of most of the features of the Internet (analogs of web 2.0, email, usenet, etc), except it was done on dumb terminals hooked to a central mainframe. Many PLATO systems were hosted on school campuses and used mainly for computer-based education.
They somehow managed to support hundreds of simultaneous interactive user sessions hosted on a single CPU with horsepower comparable to that of an 80286. The graphics-capable terminals used a cool 500x500 plasma display that took advantage of the fact that a grid of plasma dots can act as a memory array, so no frame buffer was required.
Re:Religious Viewers= $
on
Lost Ends
·
· Score: 4, Funny
Me, too. Another show that infurated me was I Dream of Jeannie. I mean, under the laws of physics and rational human reason, there's just *no way* that Barbara Eden could fit into that tiny little bottle. The only explanation possible was supernatural mumbo jumbo, which was an insult to Larry Hagman and the rest of the scientific community at NASA.
I once had a friend who said had a similar outlook when he would practice his habit of driving like an idiot: "The faster we go, the less time we spend in danger!"
The FCC gets to control broadcast language because the government retains ultimate ownership of the electromagnetic spectrum and issues licenses to TV and radio stations. They get to make the rules in that case since they own the medium. Notice that they only choose to regulate certain transmissions. Satellite airwaves have no shortage of profanity, for example.
The government does *not* own sound waves, so their ability to arbitrarily censor audio communications are much more limited. They have to prove that the sound created is causing actual harm to some victim in order to put constraints on your natural right to make sounds. Running a jackhammer in the middle of the night so that nobody can sleep is an example of actual harm; someone overstepping your overly delicate sense of prudishness is not.
It keeps thousands of aerospace engineers, scientists, and technicians productively employed.
You can spin it any way you want, but it doesn't change the fact that when the government gets into the business of designing and micromanaging rocket programs, it's socialism.
Slashdot Mirror
The main thing for which the the Shuttle is a "proven delivery system" is the transportation of breathtaking amounts of taxpayer money to a cabal of well-connected aerospace contractors.
Aside from that one feature, the design, capabilities and risk profile of the Shuttle launch system make almost zero sense.
Where would 1970s progressive rock have been if Roger Waters or Roger Hodgson had never had a bad teacher?
How dare they copy/paste those blank lines!
Microsoft was very careful not to infringe on others' IP on this point. They made certain that every blank line in their system was different from any blank line in Unix. I applaud them for that.
which is now (somewhat disputably) owned by SCO.
It was just proved yet again to *not* be owned by SCO at all, despite their claims to the contrary.
Basically, SCO tried to plagiarize then entire Unix operating system. I would hope that your are "completely, absolutely opposed" to that behavior.
Hmmm... I'd say that you can't be completely sure about what really happens to those parts unless you hitch a ride on the back of one of the delivery trucks like Charlton Heston in Soylent Green.
You mean to tell me that Alpha Decay is rare in the universe? I simply don't buy the argument.
Alpha decay generally happens to elements heavier than lead. Those elements are only created as a small side reaction in supernova explosions. Only a fraction of matter is in stars, and only a fraction of stars become supernovas, and only a small fraction of the matter in a supernova becomes heavy elements. Relative to the total matter in the universe, alpha decay is in the parts-per-billion category. In particular, the abundance of helium is NOT due to alpha decay.
The Earth is not so special that we can have zero Helium.
It is pretty special. The only thing that holds helium over the long term is gravity, and earth just doesn't have enough of it. The only place we can get at abundant helium is gas giant planets, and we don't have any technology in the foreseeable future to lift anything out of those gravity wells.
Arguing that helium is abundant in the universe therefore there must be plenty on earth is silly. Using that logic, I could say that a much, if not most, of the planetary mass in our solar system is in the form of metallic hydrogen. Therefore metallic hydrogen is available for us to use here.
I understand where people are coming from when they warn of this kind of stuff, but LONG term this stuff resupplies at a pretty decent rate. Hence the reason He is the second most abundant element in the universe.
The actual reason He is the second most abundant element in the universe is that huge amounts of it were formed in the first moments of the Big Bang. A little more has been formed since then by fusion in stars. Unfortunately, essentially none of the helium from either of those sources has stayed put on earth. It all floated away long ago.
Helium created by decay of heavy elements in incredibly rare in the universe, and it's rare on the earth as well, but it's the only helium we can get at. It forms at a rate that's way too low and too diluted for us to use. It has accumulated over millions of years in the same geological structures that capture natural gas, but those special traps certainly aren't being replenished fast enough for our needs.
*All* processor architectures are putting lipstick on pigs. Every so often a new "clean" RISC architecture comes out where the instruction set is supposed to look like the actual hardware. Then within a couple of CPU generations, the hardware landscape shifts and they slap ugly layers of abstraction between their backwards compatible instruction set and the actual new hardware. (Many years ago I listened to a pitch from a guy at MIPS, who was touting their chip named for "Microprocessor without Interlocked Pipeline Stages". Guess what they added to the next version of the processor: Interlocked pipeline stages!)
Intel themselves tried the hardest to get out of this cycle by making the hardware extremely visible to the instruction set with Itanium, trusting compiler technology to handle the resulting morass. Result of this experiment: Epic fail.
At the end of the day, X86-derivatives run at speeds in the same ballpark as any other CPU architecture that can be programmed with real-world tools by real-world coders, and they do it at a fraction of the cost. Why break all of the code out there if there's no big payback?
In the low performance efficiency market, ARM currently has advantages over X86. But by the time they bloat up ARM with a few more generations of "innovation" to get into the X86 performance range, it probably won't have that much of an advantage in size or power.
And the operational history of the Shuttle program shows that "manrated" = meaningless.
that I asked myself when I read the GPL. why the FUCK doesn't Stallman communicate directly and get away from the obsufcated communication style that he uses.
If the GPL causes you so much distress, I suggest you go try to read a typical proprietary EULA. Then maybe that will cause your head to explode and we'll all be better off.
MMX and related instructions are handled by a coprocessor.
Look at modern CPU architecture diagrams. The media units are no more of a coprocessor than any other functional unit. In particular, the instructions are decoded and run in the same pipelines as everything else. But I suppose if you choose to define any data path that doesn't match the pointer size as a "coprocessor", then we're both right.
String handling works twice as fast on 64 bit architectures as on 32 bit ones,
I think that assertion is totally false. One of the biggest bottlenecks on performance of modern CPUs on this kind of workload is accurate branch prediction, which bit size does almost nothing to help. Just block moving strings around is not an issue, and even if it were, newer 32-bit X86 processors have special tricks using a specially crafted REP MOVSD operations that can move entire cache lines in one clock.
There is no sense to having 32 bit components to a 64 bit OS, unless they are necessary for the support of 32 bit applications.
But you're not looking at it from the standpoint of a software vendor, who are the ones who actually provide the components. They have to release the 32-bit code anyway for the foreseeable future, and managing two separate releases is a lot more work for them in return for the single-digit performance increases from lower register pressure. Users most likely don't perceive such small performance increments, so a 64-bit release gets dropped in favor of more crucial work.
Whatever. If you can find a popular "64-bit architecture" that only supports 32-bit pointers, feel free to point it out.
My point was: browsers do not need either 64-bit address pointers nor 64-bit integers for performance. They can already operate on 128 bits of data (split into several smaller numbers) per clock cycle on today's commodity 32-bit processors to do multimedia processing and graphics. The rest of what they do is mainly string processing and compiling, both of which run just fine in 32-bit mode.
Actually, the "bit size" of a CPU generally refers to its address pointer size. The 32-bit x86 has been able to deal with 64-bit chunks of integer data since the MMX was introduced in 1996, and now it does 128 bits at a time with SSE. Multimedia operations that are be done in a browser process don't usually benefit from working on big numbers, but they do benefit from crunching lots of smaller numbers at one time. That's what these SIMD extensions do, and a "64-bit" architecture isn't necessary for that.
Another reason: You don't need a 64-bit address space to run a web browser.
I'm shocked that it's been almost 2 hours since you commented, and NOBODY has butted in to spout some nonsense about how the manned program is unnecessary because it would have just been cheaper to launch a new Hubble.
It's not nonsense. It would have indeed been cheaper (and safer) to launch new Hubbles for each and every time it got "fixed". It's called assembly line production: Each additional Hubble gets cheaper. (This is also why disposable rockets are cheaper than doing a custom overhaul of the space shuttle after each launch.)
Notice that the National Reconnaissance Office, which has a lot more experience operating Hubble equivalents in orbit than NASA, doesn't waste time trying to send humans up to fix things.
Had we gone the more sensible route, we also wouldn't have been subjected to endless Discovery Channel documentaries about how Story Musgrave and his friends had to spend 8-hour shifts trying to twist stuck screws in zero-G.
Ok, but also can't ignore the energy cost to produce and transport the gas, coal and/or uranium to the power plant to generate the electricity. It probably comes out similar to that of delivering gas to your house, so it's not really an issue specific to gas heat.
What I was originally talking about was the excess energy needed to produce electricity due to the laws of thermodynamics as they relate to heat engines. That assumes that the fuel is already available at the power plant. Factoring in additional energy needed to get the fuel would just make electricity look even worse.
My plasma and electronic gizmos heat is not wasted one bit.
On average, for every watt of electric power delivered to your house, 2 more watts goes up the smokestacks and cooling towers of your power plant. Those 2 watts are still wasted. (That's why the only cost effective way to heat with electricity is a heat pump, which delivers multiple watts of heating for every watt of electricity consumed.) If, OTOH, you used a modern gas furnace to get that same heating, you'd waste less than 10% of the energy in the fuel source.
Running power hungry appliances to create heat is wasteful and does not make economic sense.
You can only drink so much undistilled beer before evacuating
OTOH, given the typical price of these esoteric products, you can only drink so much "distilled beer" before your wallet is evacuated. That will tend to limit your ability to over-consume any toxins.
As there is no atmosphere in orbit
There's no fundamental reason why you couldn't orbit within the atmosphere at Mach 22, with the occupants experiencing weightlessness. We don't currently have the materials and engines to withstand such conditions, nor would it make any sense to try, but it could be done in theory.
Anyway back to the OP's point, since kinetic energy is proportional to the square of the velocity, Mach 6 is only 7% of the energy needed to reach orbit. IMO, scramjets are just a complex diversion if the goal is to go into orbit. (And you need to first reach supersonic speeds by yet some other means before you can even switch them on.) Reaching Mach 6 with an ordinary booster rocket stage is child's play in comparison.
From what I've read about PLATO (I was born quite a bit after PLATO's heyday) it seemed to be in stark contrast with today's methods of teaching computers. It seemed like PLATO actually encouraged students to explore computers. Today though, teachers are too paranoid, thinking that the command prompt will "break" the computer and other stupidities.
When I was a student back in the days of PLATO, I had a part time job as the human tutor in one of the PLATO terminal rooms. I don't remember it being focused at all on exploring computers. The system was all about the pre-canned apps. In fact, my memory is a little rusty, but I don't recall that they really had a command prompt at all, at least as far as end users were concerned. I think it was all a hierarchical full screen menu-driven system. (I assume that some CS majors were taught how to write software for PLATO, but that would be a small minority of the users.)
One problem with the course that I worked with was that the software was a bit too linear and inflexible. For example, students weren't allowed to go on to the next problem until they correctly answered the current one, and the range of acceptable answers was usually very constrained. The software basically kept repeating: "Wrong. Try again.", and you were stuck at a dead end.
Unfortunately, back in those days this was often the first exposure the users had to a computer system of any kind. They had never experienced anything as exacting and unforgiving as a computer, and it didn't help to heap that on top of the inherent stress of a "weed-out" engineering class. That's why they needed me to be in there as a backup; I think that some of the people would have eventually gone postal on the terminals if they didn't have access to someone who could see how and why they were stuck, and dole out helpful hints.
In a nutshell: It was a preview of most of the features of the Internet (analogs of web 2.0, email, usenet, etc), except it was done on dumb terminals hooked to a central mainframe. Many PLATO systems were hosted on school campuses and used mainly for computer-based education.
They somehow managed to support hundreds of simultaneous interactive user sessions hosted on a single CPU with horsepower comparable to that of an 80286. The graphics-capable terminals used a cool 500x500 plasma display that took advantage of the fact that a grid of plasma dots can act as a memory array, so no frame buffer was required.
Me, too. Another show that infurated me was I Dream of Jeannie. I mean, under the laws of physics and rational human reason, there's just *no way* that Barbara Eden could fit into that tiny little bottle. The only explanation possible was supernatural mumbo jumbo, which was an insult to Larry Hagman and the rest of the scientific community at NASA.
The faster you go, the less time you spend going.
I once had a friend who said had a similar outlook when he would practice his habit of driving like an idiot: "The faster we go, the less time we spend in danger!"
The FCC gets to control broadcast language because the government retains ultimate ownership of the electromagnetic spectrum and issues licenses to TV and radio stations. They get to make the rules in that case since they own the medium. Notice that they only choose to regulate certain transmissions. Satellite airwaves have no shortage of profanity, for example.
The government does *not* own sound waves, so their ability to arbitrarily censor audio communications are much more limited. They have to prove that the sound created is causing actual harm to some victim in order to put constraints on your natural right to make sounds. Running a jackhammer in the middle of the night so that nobody can sleep is an example of actual harm; someone overstepping your overly delicate sense of prudishness is not.
It keeps thousands of aerospace engineers, scientists, and technicians productively employed.
You can spin it any way you want, but it doesn't change the fact that when the government gets into the business of designing and micromanaging rocket programs, it's socialism.