I distinctly remember a bunch of hoopla about how Win2K wasn't going to require reboots, either. Stories were fed to the press prior to release about how Bill Gates sent down an edict that there would be no more gratuitous rebooting, or heads were going to roll in Redmond. Here we are, six years later, and nothing ever came of it.
The latest RISC chips do all sorts of similar slicing and dicing on the instruction stream to try to find low-level parallelism anyway, so the X86 isn't really at a disadvantage. The more compact instruction coding means that it's munging less data when it does this, too.
Instruction decoding used to be a significant portion of the die real estate back in the Pentium Pro days, but the decode logic hasn't grown that much compared to other areas, especially the on-chip cache which currently dominates the die area. Compact instructions save on that cache area as well.
The only major non-embedded chip that's really trying to get away from those hardware tricks is the Itanium, and we've all seen just how well it has blown away the competition (not).
More detailed analysis of the radar image indicates that the shape of the flat region actually appears to be almost perfectly rectangular, with an aspect ratio of 4:9. Nobody is quite sure what to make of that.
Blah blah blah. As I said, the defense system probably doesn't work anyway. There is little evidence that the tests done so far haven't been rigged, and many of those haven't worked. All it would take is the tiniest change to make ensure that it doesn't work at all. Few of your assertions about handling the countermeasures have been or ever will be tested in the real world.
Much of your objection is "it would make the payload heavier". So what? They'll build bigger missiles or lighter warheads. You assume a static situation; that's not the case. Our potential adversaries will adapt as fast as we can sink money into this giant black hole.
This really is nothing more than the latest in a long line of US technological bluffs that included the SAGE air defense network and the original SDI. Each has been extremely expensive and would never work in actual combat conditions, but just by letting the world see us spending so much money, we hope that they're bamboozled into thinking that we're building something that actually does something.
This is much easier to do if the attacker knows that a significant portion of the attacking missiles won't make it to their targets, and that he cannot prevent a certain, devastating retaliation.
We were talking about China and other small countries here. They're already going to get certain, devastating retaliation if they try to launch any missiles. They already know, as you've admitted, that at least some fraction of the missiles are going to get through. The best way for them to increase that number is to just build more nukes, and that's exactly what's going to happen each time we increase defensive deployments. And we won't even feel any safer, because no matter how much handwaving you do, we can never be really certain that they haven't put effective countermeasures into their missiles.
1. Launch a couple of extra nukes and explode them in the ionosphere just ahead of the real missiles. There won't be time to intercept them before they go off. It will be kind of hard to operate radars with all the plasma disruption. Handling this scenario can't be tested for, and any good engineer knows that something that's not tested isn't likely to work.
2. Give the warhead heat shields a good liquid nitrogen wash just before launch. They'll be a just a little hard to spot with infrared sensors. By the time they hit the atmosphere and warm up again, you're going to have just a minute or two before it reaches the target. Little impact on launch system other than a little plumbing.
3. Toss a few dozen flares out with the warhead. More problems with infrared sensors. Flares weigh almost nothing.
4. Cold gas thrusters constantly jostle the warhead back and forth randomly by a few yards. Kinetic interceptors wouldn't have much chance. It's a little bit of engineering, but it's standard spacecraft technology.
5. Release a bright cloud of vapor or dust around the warhead. It will be hard to pinpoint in the glare. Not much harder to implement than strapping an aerosol can onto the payload.
6. Spin the warhead around a small counterweight on a long tether. Once again, hard to hit. Harder to implement, but it's been done with satellites.
7. Inflatable decoys. Simple to implement.
8. Real decoy missiles with inert warheads, with identical mass and external materials as actual warheads. Relatively expensive, but much cheaper than adding more interceptors and command structure on the defense side.
9. Strobelights and radar transmitters on the warheads broadcast noise signals. Likely mess up delicate tracking systems designed to look for feeble signals.
10. Lots of cheap decoy rockets that simulate only the boost phase of real rockets. Defense systems would have a lot of sorting out to do before trying to locate the actual targets.
I could go on and on. How about you prove that not a single one of these or any combination of them would impact the effectiveness of the defense systems?
This raises the stakes considerably -- a knock-out first strike becomes impossible
China doesn't currently have a first strike capability anyway. Why drive them towards that direction? With their current posture, all they need is for some percentage to get through to deter us from doing an Iraq on them.
Right, don't mention that, because it isn't true.
It is true. There are literally hundreds of possible easy-to-implement countermeasures, and we have no idea what combination they might try. These missile defenses don't properly work today even under the laboratory conditions they're tested under, much less in the real world with evasive features tacked on.
The cold-war defeat of the Soviet Union was exactly for this reason: they couldn't afford the spending to maintain their ICBM force at their desired threat level in the face of the various defensive strategies.
The USSR couldn't afford their own missile shield, which would have been ridiculously expensive. We couldn't have afforded it either. China is not going to be making a missile shield for the same reasons I stated above. All they have to do is crank out some more missiles, which are relatively cheap. The current situation is not like the old cold war.
That the long term threat will be China, and your close launch system will not apply.
Trying to build a missile defense system to counter China is futile. China has the resources to build enough ICBMs to overwhelm any possible missile defense, and we're sending them $Billions more resources every quarter in exchange for trinkets.
Given that it costs orders of magnitude more to defend against an ICBM as it does to build one, they would ultimately win any ICBM defense arms race. Not to mention that countermeasures against the defense system are trivial, it's questionable if these systems even work properly without countermeasures, and it would only take a handful of intercept failures to cause devastating damage to the US. All such a deployment would do is prompt China to crank out more nukes. We would not be one iota safer than we are today after that happens.
Well, there's your problem. He was talking about automatic updates. Automatic updates aren't controlled by the "install PostGreSQL" command under any OS.
IMO, the single biggest timesaver in bash is the Ctl-R history recall search feature. (It was quite a while before I found out about it, and I wish I had found it sooner.)
If you crank up your history list to a few thousand entries and set it to forget dupes, you can recall any command you've issued in the last couple of months with just a couple of keystrokes.
At one point we needed to fly to Colorado to conduct some testing - it would have cost several $million to test locally, and several $thousand to test in CO.
Wow... with those kind of discounts available, maybe Massachusetts should have built their huge highway tunnel in Colorado, too.
The vast majority of "investment" funds by record labels goes into promotion and antiquated distribution mechanisms. But their expensive forms of promotion are driven by the needs their particular business model.
Without the labels as middlemen who shift 90% of the overall revenue into overhead costs, artists would need little if any outside "investment". Now that publishing music no longer technically requires centralized control, the industry should move to a system where artists do direct online sales through independent low-overhead sites. Instead of signing their copyrights over to middlemen, they would hire them as needed. In this new efficient market, the artists will make far more money, and the music will be much cheaper for consumers to buy.
A jar of Cheez Whiz is sterile when you open it, too. Now try spreading it all over your toilet, and keep it nice and moist. See what happens after a few days.
And what about the OTHER diseases that mosquitos spread?
Most of them aren't serious enough to justify the cost of placing millions of mosquito magnets in every inhabited tropical area of the world, connecting them to the AC power grid, and then constantly replenishing their propane tanks. Once again, developing vaccines or treatments would be far more practical and effective.
Oh, and calm down. You don't need to tack insults on to the end of every post you make in this thread. People might think that you're fucking prick or something.
The appropriate solution to your concerns is finishing development of a malaria vaccine. It will be cheaper, more effective and less harmful than trying to wipe out a whole chunk of the food chain.
nice, he makes a big ostentatious show of covering up his RFID strip with foil so "they" can't get at him, and of course all that happens is "they" make a big show of harassing him.
This is exactly what he intended. If they hadn't harassed him, then the story wouldn't be in the news, and nobody would know about it. However, he knew that this would most likely cause some kind of stink, which would generated a news story that gets people talking about the issue. Now we're all here thinking about RFD, just as RMS wanted.
RMS played the UN security like puppets on strings just the same way as terrorists play the administration and congress: they know what the knee-jerk reaction will be and they use it to their advantage.
FAT is no longer used for serious hard drive storage, so it has nothing to do with ext3, etc. It is now a common interchange standard for peripheral devices like cameras and thumb drives. By attempting to institute a tax on a previously-considered-free standard at this late date, they are impeding progress in the peripheral hardware area.
BTW, the utility of this patent has to do with backwards compatibility with OSes that only understand the 8.3 file format, which nobody gives a shit about anymore. However, the particular way that long filenames are kludged into VFAT are now cast in concrete, and any implementation is stuck infringing the patent claims regardless of whether anybody will ever access the 8.3 filenames. In other words, the patent no longer has any valid technical use other than creating market barriers and collecting licensing revenue.
I've since been desoldering leds, and using X10 modules to turn off VCR clocks (I have both a watch and a cellphone - but thanks for the valueadd of a clock on my microwave, coffee maker, vcr, phone, scale, etc.)
The problem isn't the clock in the device. The clock logic and LED display use up a tiny fraction of one watt. The problem is the power supply.
Take the microwave for example: people expect to be able to walk up and start punching in a cooking time without first having to push a huge mechanical power switch. (The manufacturer doesn't want to design in a costly extra power switch either.)
This means that the electronics need to be powered on at all times. That wouldn't be a problem, but most appliances use a simple transformer to drive their power supplies. Inexpensive transformers are leaky even when they are supplying no current to the secondary, so the microwave's transformer is probably wasting a couple of watts at all times. The solution to the problem is a better power supply, not omitting the clock or desoldering LEDs.
Some recent wall warts and power bricks that I've got weigh almost nothing and don't seem to get hot. I presume that they've put in switching power circuits and eliminated the 60Hz transformer altogether. Putting that kind of power supply in every appliance would go a long way towards solving this problem.
I recently hooked an ammeter to a bunch of gadgets to find out what they draw. My current cable box is fairly toasty, but not as hot as some that I've seen. When "on", it draws 17W. When "off", it draws 16W. That wasn't surprising, since the audio still comes out of the box when it's off, even on digital channels. All they seem to do is shut off the video out circuit and the "power" LED.
I still can't believe he told us, a group of college students, that we weren't smart enough to control a power switch.
Hey, guess what: People aren't smart enough to turn off their equipment.
At one place I worked at a few years back, almost all of the CRTs in the offices would sit drawing > 100W each showing screen savers every night and weekend. I usually would at least swing by the lab on my way out each night and turn off about 1000W worth of monitors left on by others.
Al Gore was exactly right. The 2W drawn by a sleeping monitor would come out as a win if even only 10% of the people left their monitors on. And from what I've seen, far more than 10% don't bother to shut off their monitors.
Great. But I hope you live within walking distance of that store, because if you're driving there, you're probably going to be burning a bit more than 10 cents' worth of gas. Or paying more than 10 cents in postage, if you go mail order. You need to be printing a LOT of pictures before you start saving money by paying someone else to print them.
Of course, if you plop down $60 on a new set of inkjet cartridges and then don't use them much, you'll be lucky to get 6 months before they dry out. In that case your ink will be drying up at the rate of 10 cents every 8 hours, 24x7. So if you print only a few pictures, you also save money by sending them out. That's why I junked my last inkjet and got a cheap laser printer.
Ever since then, I've wondered why they don't use this kind of memory in multi-processor systems
IIRC, the video port of a VRAM wasn't really random access. It would load an entire row of the RAM array into an output buffer, and then you had to shift out the columns serially. This avoided conflicts on the address bus, and it works fine for refreshing a video display, but it isn't at all suitable for general multiprocessor use.
I seem to remember that you could buy true dual-ported static ram at one time, with 2 address buses and 2 data buses, but it was hideously expensive.
The folks over at Diebold are happy to hear this--now they can charge a whole bunch extra for printers...
Actually, the printers will be provided at no extra charge. However, the consumables will be a different story. Diebold predicts that by 2009, ink and paper refills will generate 87% of their revenue and over 94% of their total profits. The remainder of the profits will be generated largely by sales of extended warranty plans.
If they could clone an IBM, why can't people clone nintendo?
There were several reasons:
Software patents were extremely rare in the early 1980s. Nobody would have bothered patenting something as boring as the BIOS.
The IBM hardware was little more than an Intel demo board reference design plus a few assorted kludges thrown together by some recent college hires. Back then, people tended to think that you needed to actually do something clever to get a patent, so IBM didn't really patent much of the PC hardware.
Much of IBM's top management thought that the PC was a toy in the first couple of years. They just didn't care that much at first.
IBM at the time was operating under a restrictive terms from their decades-long anti-trust case. One effect of this was that they didn't really push too hard on any patents.
In the mean time, Compaq collected some valuable patents on enhancements they added to PCs, such as multisync monitors that could handle both graphics and text mode. By the time IBM woke up and came looking for license revenue, Compaq had enough collateral to get reasonable terms on a comprehensive patent cross license.
If you look at recent motherboards, much of the real-estate is taken up by DC-to-DC power circuits that convert the 3.3V from the power supply to the various voltages actually used by the CPU, memory and maybe other miscellaneous chips.
It seems to me that it would be ultimately more efficient to just supply AC power to the motherboard and let it create the various voltages that it needs. The same goes for disk drives, etc. In fact, maybe all computing equipment ought to have been standardized on a single "safe" AC voltage, like 12VAC. Then one simple transformer could supply all computing needs without having to figure out how to distribute "clean" DC power. I'm sure it's far too late to make a change like that now, though.
Slashdot Mirror
I distinctly remember a bunch of hoopla about how Win2K wasn't going to require reboots, either. Stories were fed to the press prior to release about how Bill Gates sent down an edict that there would be no more gratuitous rebooting, or heads were going to roll in Redmond. Here we are, six years later, and nothing ever came of it.
Instruction decoding used to be a significant portion of the die real estate back in the Pentium Pro days, but the decode logic hasn't grown that much compared to other areas, especially the on-chip cache which currently dominates the die area. Compact instructions save on that cache area as well.
The only major non-embedded chip that's really trying to get away from those hardware tricks is the Itanium, and we've all seen just how well it has blown away the competition (not).
More detailed analysis of the radar image indicates that the shape of the flat region actually appears to be almost perfectly rectangular, with an aspect ratio of 4:9. Nobody is quite sure what to make of that.
Much of your objection is "it would make the payload heavier". So what? They'll build bigger missiles or lighter warheads. You assume a static situation; that's not the case. Our potential adversaries will adapt as fast as we can sink money into this giant black hole.
This really is nothing more than the latest in a long line of US technological bluffs that included the SAGE air defense network and the original SDI. Each has been extremely expensive and would never work in actual combat conditions, but just by letting the world see us spending so much money, we hope that they're bamboozled into thinking that we're building something that actually does something.
This is much easier to do if the attacker knows that a significant portion of the attacking missiles won't make it to their targets, and that he cannot prevent a certain, devastating retaliation.
We were talking about China and other small countries here. They're already going to get certain, devastating retaliation if they try to launch any missiles. They already know, as you've admitted, that at least some fraction of the missiles are going to get through. The best way for them to increase that number is to just build more nukes, and that's exactly what's going to happen each time we increase defensive deployments. And we won't even feel any safer, because no matter how much handwaving you do, we can never be really certain that they haven't put effective countermeasures into their missiles.
2. Give the warhead heat shields a good liquid nitrogen wash just before launch. They'll be a just a little hard to spot with infrared sensors. By the time they hit the atmosphere and warm up again, you're going to have just a minute or two before it reaches the target. Little impact on launch system other than a little plumbing.
3. Toss a few dozen flares out with the warhead. More problems with infrared sensors. Flares weigh almost nothing.
4. Cold gas thrusters constantly jostle the warhead back and forth randomly by a few yards. Kinetic interceptors wouldn't have much chance. It's a little bit of engineering, but it's standard spacecraft technology.
5. Release a bright cloud of vapor or dust around the warhead. It will be hard to pinpoint in the glare. Not much harder to implement than strapping an aerosol can onto the payload.
6. Spin the warhead around a small counterweight on a long tether. Once again, hard to hit. Harder to implement, but it's been done with satellites.
7. Inflatable decoys. Simple to implement.
8. Real decoy missiles with inert warheads, with identical mass and external materials as actual warheads. Relatively expensive, but much cheaper than adding more interceptors and command structure on the defense side.
9. Strobelights and radar transmitters on the warheads broadcast noise signals. Likely mess up delicate tracking systems designed to look for feeble signals.
10. Lots of cheap decoy rockets that simulate only the boost phase of real rockets. Defense systems would have a lot of sorting out to do before trying to locate the actual targets.
I could go on and on. How about you prove that not a single one of these or any combination of them would impact the effectiveness of the defense systems?
China doesn't currently have a first strike capability anyway. Why drive them towards that direction? With their current posture, all they need is for some percentage to get through to deter us from doing an Iraq on them.
Right, don't mention that, because it isn't true.
It is true. There are literally hundreds of possible easy-to-implement countermeasures, and we have no idea what combination they might try. These missile defenses don't properly work today even under the laboratory conditions they're tested under, much less in the real world with evasive features tacked on.
The cold-war defeat of the Soviet Union was exactly for this reason: they couldn't afford the spending to maintain their ICBM force at their desired threat level in the face of the various defensive strategies.
The USSR couldn't afford their own missile shield, which would have been ridiculously expensive. We couldn't have afforded it either. China is not going to be making a missile shield for the same reasons I stated above. All they have to do is crank out some more missiles, which are relatively cheap. The current situation is not like the old cold war.
Trying to build a missile defense system to counter China is futile. China has the resources to build enough ICBMs to overwhelm any possible missile defense, and we're sending them $Billions more resources every quarter in exchange for trinkets.
Given that it costs orders of magnitude more to defend against an ICBM as it does to build one, they would ultimately win any ICBM defense arms race. Not to mention that countermeasures against the defense system are trivial, it's questionable if these systems even work properly without countermeasures, and it would only take a handful of intercept failures to cause devastating damage to the US. All such a deployment would do is prompt China to crank out more nukes. We would not be one iota safer than we are today after that happens.
Well, there's your problem. He was talking about automatic updates. Automatic updates aren't controlled by the "install PostGreSQL" command under any OS.
If you crank up your history list to a few thousand entries and set it to forget dupes, you can recall any command you've issued in the last couple of months with just a couple of keystrokes.
Wow... with those kind of discounts available, maybe Massachusetts should have built their huge highway tunnel in Colorado, too.
What leads you to the conclusion that krray made no revenue on that project?
Without the labels as middlemen who shift 90% of the overall revenue into overhead costs, artists would need little if any outside "investment". Now that publishing music no longer technically requires centralized control, the industry should move to a system where artists do direct online sales through independent low-overhead sites. Instead of signing their copyrights over to middlemen, they would hire them as needed. In this new efficient market, the artists will make far more money, and the music will be much cheaper for consumers to buy.
A jar of Cheez Whiz is sterile when you open it, too. Now try spreading it all over your toilet, and keep it nice and moist. See what happens after a few days.
Most of them aren't serious enough to justify the cost of placing millions of mosquito magnets in every inhabited tropical area of the world, connecting them to the AC power grid, and then constantly replenishing their propane tanks. Once again, developing vaccines or treatments would be far more practical and effective.
Oh, and calm down. You don't need to tack insults on to the end of every post you make in this thread. People might think that you're fucking prick or something.
The appropriate solution to your concerns is finishing development of a malaria vaccine. It will be cheaper, more effective and less harmful than trying to wipe out a whole chunk of the food chain.
This is exactly what he intended. If they hadn't harassed him, then the story wouldn't be in the news, and nobody would know about it. However, he knew that this would most likely cause some kind of stink, which would generated a news story that gets people talking about the issue. Now we're all here thinking about RFD, just as RMS wanted.
RMS played the UN security like puppets on strings just the same way as terrorists play the administration and congress: they know what the knee-jerk reaction will be and they use it to their advantage.
BTW, the utility of this patent has to do with backwards compatibility with OSes that only understand the 8.3 file format, which nobody gives a shit about anymore. However, the particular way that long filenames are kludged into VFAT are now cast in concrete, and any implementation is stuck infringing the patent claims regardless of whether anybody will ever access the 8.3 filenames. In other words, the patent no longer has any valid technical use other than creating market barriers and collecting licensing revenue.
The problem isn't the clock in the device. The clock logic and LED display use up a tiny fraction of one watt. The problem is the power supply.
Take the microwave for example: people expect to be able to walk up and start punching in a cooking time without first having to push a huge mechanical power switch. (The manufacturer doesn't want to design in a costly extra power switch either.)
This means that the electronics need to be powered on at all times. That wouldn't be a problem, but most appliances use a simple transformer to drive their power supplies. Inexpensive transformers are leaky even when they are supplying no current to the secondary, so the microwave's transformer is probably wasting a couple of watts at all times. The solution to the problem is a better power supply, not omitting the clock or desoldering LEDs.
Some recent wall warts and power bricks that I've got weigh almost nothing and don't seem to get hot. I presume that they've put in switching power circuits and eliminated the 60Hz transformer altogether. Putting that kind of power supply in every appliance would go a long way towards solving this problem.
I recently hooked an ammeter to a bunch of gadgets to find out what they draw. My current cable box is fairly toasty, but not as hot as some that I've seen. When "on", it draws 17W. When "off", it draws 16W. That wasn't surprising, since the audio still comes out of the box when it's off, even on digital channels. All they seem to do is shut off the video out circuit and the "power" LED.
Hey, guess what: People aren't smart enough to turn off their equipment.
At one place I worked at a few years back, almost all of the CRTs in the offices would sit drawing > 100W each showing screen savers every night and weekend. I usually would at least swing by the lab on my way out each night and turn off about 1000W worth of monitors left on by others.
Al Gore was exactly right. The 2W drawn by a sleeping monitor would come out as a win if even only 10% of the people left their monitors on. And from what I've seen, far more than 10% don't bother to shut off their monitors.
Of course, if you plop down $60 on a new set of inkjet cartridges and then don't use them much, you'll be lucky to get 6 months before they dry out. In that case your ink will be drying up at the rate of 10 cents every 8 hours, 24x7. So if you print only a few pictures, you also save money by sending them out. That's why I junked my last inkjet and got a cheap laser printer.
IIRC, the video port of a VRAM wasn't really random access. It would load an entire row of the RAM array into an output buffer, and then you had to shift out the columns serially. This avoided conflicts on the address bus, and it works fine for refreshing a video display, but it isn't at all suitable for general multiprocessor use.
I seem to remember that you could buy true dual-ported static ram at one time, with 2 address buses and 2 data buses, but it was hideously expensive.
Actually, the printers will be provided at no extra charge. However, the consumables will be a different story. Diebold predicts that by 2009, ink and paper refills will generate 87% of their revenue and over 94% of their total profits. The remainder of the profits will be generated largely by sales of extended warranty plans.
There were several reasons:
It seems to me that it would be ultimately more efficient to just supply AC power to the motherboard and let it create the various voltages that it needs. The same goes for disk drives, etc. In fact, maybe all computing equipment ought to have been standardized on a single "safe" AC voltage, like 12VAC. Then one simple transformer could supply all computing needs without having to figure out how to distribute "clean" DC power. I'm sure it's far too late to make a change like that now, though.