I-Opener's model is based on selling the internet service no on selling hardware so I don't see what's so bad about making people take the hardware as is and not modifying it. This really has nothing to do with OpenSource. I-Opener made a mistake by not requiring buyers of the hardware to sign up and now they're fixing it. Expecting them to keep selling hardware for 99 or whatever the price was would be like expecting Sun to keep selling E10Ks for 3000 after they made a mistake and sold a few at that price. It was good deal while it lasted but to expect to them to keep doing it is ludicrous
I belive you're wrong there. Although beowulf clusters are nice for applications that are computationally intensive but don't require much bandwidth, a lot of problems require a lot of bandwidth. This is where SMP boxes rule since bandwidth between processors on a single motherboard/backplane is a LOT larger than across a network. For these types of problems, the time required is more function of bandwidth rather than cpu so a smp box with fewer processors but more bandwidth between the processors will beat even a large beowulf/clustering solution.
The article talks about how FP on the x86 is stack based. However, I seem to remember intel switching to a stack/register model back in the pentium. I believe all the recent x86 chips have allowed you to address stack locations using F(1), F(0), etc. but you can still treat the registers as a stack if you want. I may be wrong though.
The most important detail that the article leaves out is how the new 64 bit extensions will work. Personally I'm hoping that AMD keeps the current instructions but creates 64 bit extensions that work in a more RISClike fashion. Something like limiting addressing modes to register and register+constant, do the more complex instructions in microcode, and most importantly add another 10 or 20 gp registers to the integer code.
Even at 50% atmosphere, your eyeballs will be popping. Atmospheric pressure is 14.7 psi, 50% atm will mean a difference of 7 lbs per sq inch.
Actually your eyes wouldn't pop even in a zero atmospehere. Our body's structual integrity is strong enough to handle even zero pressure without much more damage than a few bruises. Consider putting a vacuum cleaner hose on your arm, the pressure at the end will at least be 7 psi but you won't get more than a small bruise. Also, divers working at 300ft down work in pressures from 20-30 atm (20-30x normal pressure) without problems.
Also death by decompression is painful, but that's more due to gas bubbles forming in your capillaries giving you microstrokes all over your body NOT from your body exploding.
Before you start criticising Slashdot, note that the Slashdot story cited its (only) source, and provided a link to it.
My problem was with the way the information is presented. Instead of saying that the original story was incorrect but the update to the original was correct, Hemos said the same thing but in a way that leads the reader to think that the original was correct. Given that the headline of that story was about Sony recalling some the memory cards, most people are probably going to remember that.
An analogous example are those sweepstakes that seem to suggest that you need to buy something in order to enter but at the very end say that no purchase is necessary. I just think its sleazy when someone makes a mistake, and then frames the apology/correction in such a way as to suggest that the original mistake was not a mistake after all.
According to this story at Daily Radar, the online stories about the Playstation 2 Memory Recall are false.... So, that means our recent story was true, as updated.
So what this is really saying is that Slashdot originally got the story wrong. A lot of people probably did not see the update and Hemo's little sentence implies that the original story was right. This something you would expect to see out of pr people. Sure its technically correct, but to people who aren't reading the description carefully it appears to say that the original story was correct when it wasn't. It would have been a lot clearer to say something like "So are orignal story was wrong but the update was right".
I don't think you understand that for a lot of the algorithms that break things up into cells, the accuracy comes from the size of the cell NOT the number of particles in the cell. In other words, having 90,000 particles inside the cell is the same as having 1. It simply doesn't make a difference.
If you can overclock a 366 celeron to 550, as was demonstrated on slashdot a while back, without flaws in calculations, then it would only need to survive 2/3 the time to get the same amount of computation out of it.
The problem is that you won't be sure if it screwed up on a calculation or not, especially since you don't know before hand what the results are supposed to look like. Suppose the result was off by a factor of.001%. Although it may look harmless, the error propagated through a 1000 iteration may result in a large final error. So the software says that a certain configuration of plutonium won't explode when it really will or the simulation predicts the wrong trajectory for a hurricane. In other words, most people would prefer spending an extra million or waiting a few more hours/days to get the correct result.
The problem is that each particle's position needs to be sent to all the other nodes for them to do their calculations. So taking your assumption that each particle takes 34 bytes and assuming the simulation has a million particles in it, you need to send at least 34 MB of data through the network after each calculation. So assuming the calculation takes a while, say 10,000 clock cycles (on a 500Mhz machine) then you need to do 50,000 updates a second. This comes out to about 1,700,000 MB/s of bandwidth. Even if the calculation takes 1,000,000 clocks, you still need at least 17GB/s bandwidth. A beowulf cluster may have a 1 Gb/s bandwidth over gigabit ether but that leaves you short by a factor of 136.
Using a beowulf for the example above means you can only do 30 calculations per second even though you could theoretically do 500 to 50,000 depending on the amount of clocks the calculation takes.
Now suppose you had a supercomputer that had a bandwidth of 10GB/s (gigabyte/s). Then you could do about 300 calculations per second even if your processors were running at 300MHZ instead of 500MHz. In cases like this, the I/O bandwidth determines how fast the simulation will go and a beowulf cluster would not be better.
OGRs have application to data communications, cryptography and lithography. I would say that it has a lot of use since it may lead to faster/better encryption/data transfers as well as better/cheaper chip fabs and indirectly cheaper cpu and microprocessors. A lot more useful than pretty pictures,
Nitrogen oxides? You may have a lot of these computer science geeks believing you that an internal combustion of hydrogen gives you nitrogen oxides. I don't see in the equation where nitrogen comes in.
Well you need oxygen for the combustion to occur. Since the engine won't store oxygen and will presumably use air from the atmosphere to obtain the oxygen, you get nitrogen also. Couple this with the heat the combustion produces and you'll get N combining with O to produce NO_x.
Sure, steaks and cookies are great for dinner parties, but while studying the night before a midterm or for a quick midnight snack, cold pizza or a bowl of ramen generally holds me over until I can get my hands on real food.
Well, that's whatleftovers are for. If you make a stew or chili, you usually have a lot of it left over and if you keep it in tupperware or some other storage container it'll be available for a late night snack. If you have a potato, pop it in the microwave and in less than 6 minutes you have a baked potato with chili or stew.
Maybe its just me but I never understood why ramen is so popular. If price is a factor then sphagetti or pasta is available for slightly more. It takes about the same time and tastes a lot better. Hell, baked potatoes cook in the same time as ramen and taste much better.
If cost isn't a consideration, something like lamb stew, chili, quiche, etc. take about 10-15 minutes to prepare (they take about 30-60 min to cook, but you can do something else while they're cooking).
I'm not sure where this thing with geeks eating only pizza and ramen comes from. I'm a math/cs major and most of the other physics/math/cs I know don't eat that stuff on a regular basis. At our parties, you're more likely to see stuff like salmon steaks, grilled portabella mushrooms, fresh baked cookies than pizza or ramen.
Don't use xhost +. That leaves your system open to anyone. Someone could start a hidden X program that looked at the keystrokes you were entering and snag your passwords. That's one of the reasons xterm has a secure keyboard mode.
I'm sorry in advance if your the guy who posted this to bugtraq. But the exact same message appeared with the domain names changed about 3 weeks ago in bugtraq. Next time be a little more creative.
This problem has been discussed a fair bit in bugtraq. The consensus was that DNS wasn't really secure using the crypt and signed message may help to prevent this but in general were not that great since netsol sometimes ignore crypt-pw and their pgp signed mail thing is often broken. Essentially if someone can forge their header so that it looks like its coming from the technical contact, it probably go through.
t this point I think arguments along the lines of "well, Linux/Open Source loses because they have really geeky UI's" is more FUD than accurate. It will become even less accurate over a somewhat short time scale! One metric to testing this assertion it to see how many "normal" admin UNIX tasks (sysadmin or user adminstrative tasks) are being pushed over to programs with UI's than are done on the shell.
I think the point of the article was that good UI doesn't necessarily equate to having a GUI interface. It's more on how programs have a shared consistent interface. For example, the shortcut to quit in Gnome programs is ctrl-q, in netscape and some others its atl-q, xemacs and emacs set it to ctrl-x-ctrl-c. If I want to quit a program, I don't really want to have to stop a second and figure out what the shortcut is, or end up pressing two or three shortcuts in other to quit. The MacOS and WinXX systems are a lot better in this regard since certain common functions are mapped to the same shortcut in all programs. If I want to quit a program while using those oses I can just hit alt-f4 or apple-q and know that it'll work.
Human brains kick-ass at undecideble problems. Almost anyting that pertain to instruction scheduling is NP complete. Human brains kick-ass at NP-complete problems
People aren't really kick ass at undecidable problems. People aren't able to decide whether certain programs halt or not. If the program is trivial then a computer can decide whether it halts or not probably with the same effectiveness as a person. The undecidability problem just says that there is no algorithm that can figure out whether all programs halt or not. An algorithm may be able to decide whether a certain subclasses of programs halt or not.
In addition, people most certainly do not kick ass at NP-complete problems. To solve any problem, we need an algorithm to do it and all such algorithms are nonpolynomial. If anything people would be slower at solving these problems since we can't compute things as quickly as computers.
Given their target of small mobile devices, like webpads and the like, its low power consumption and sleep mode, I don't think it's intended to be "turned off"
Regardless, it will have to be turned off even if it is for battery changes. Also people may drain the battery after using it for a while, requiring a power down. These points aside, even low power devices today are turned off (i.e. laptops, palm pilots, etc.) since even standby mode drains too much power. The crusoe systems will probably drain more power than a palm pilot so you'll probably need to turn it off.
Seriously. If this 'switch' is patched onto genes of a fetus before development, it could switch off any known genetic errors resulting in birth defects, deformities, etc. This could make genetic disorders rare.
Actually switches like this are are already present. They let developing embyros regulate their own growth so that things like a brain develop only after a skull has formed. More importantly they let cells in the embyro figure out where they are so that they can become the right sort of cells. Errors in these are responsible for a lot of birth defects.
The switches(promoters & repressors) are also called operons. Their existence and functions have been known since at least the mid 80s and possibly even earlier. The most well known system is the Lac operon in E. Coli which triggers the transcription and generation of a protein that allows the utilization of lactose when there is no glucose present and a sizable lactose concentration. It's been pretty well studied and understood for a while. Although the article doesn't mention, the researchers were probably able to reliably splice a gene into a system like this and probably did not create the operon from scratch.
I have some concerns about the performance that the Cruose processors will actually have. The article mentions that translated instructions will be cached and then be reused if the CodeMorph software sees it again. However, it seems like the CodeMorph's state information will not be mantained between runs. If you power off the computer, the software loses the cached information and has to start from scratch again. In addition, the cache's size or location isn't given. Is it a small cache on die or is it located in system memory? The cache is probably on die for speed reasons but this would limit the size of the cache. This could be a performance hit since the cache is also used as a data cache and instruction cache.
Another question concerns the way the instructions are being cached. For example suppose the following instructions were given
ADD AX, BX SUB CX, AX JNZ Cx
Would the translation for each instruction be cached, or is the sequence cached? The article implies that the sequence is cached since the CodeMorph software can optimize the speed on subsequent passes. However, this seems to limit the benefit gained from caching to relatively tight loops or common sequences of code depending on the cache size.
On a side note, the article implies that the CodeMorph software lightyears beyond anything else. However, some of its highly touted features appeared in other software before. For example, DEC's FX!32 would initially just translate code but would also observe the application behaviour and then optimize the code based on that after the application finished executing. It could do this optimization several times, optimizing more aggressively on each pass. Also Apple's 680x0 emulator was also based in rom that would start up initially so what the MacOS could boot. The CodeMorph software has some new features if it really does OO scheduling and optimization on the fly but that seems like a pretty big hit on performance.
If future server/desktop oriented processors implement large parts of the CodeMorph software in hardware, how will that be any different than AMD or Intel's processors since they'll all be implementing a hardware instruction translation unit besides the Transmeta core being VLIW. Plus the transistor count and power consumption will also sky rocket along with that.
NSF ocean sciences: $220 Million- which is not to say Oceanography does not benefit from NASA, the military, etc, but I think the point is clear)
I think that oceanographers tend to get more than that figure in the form of free time on equipment. A couple of people I worked with were working on installing some instruments for DUMAND in Hawaii were getting a a few weeks every year from the Navy on their submersible. From what I understand, the submersible time was around 100k a day if they had to buy it (they were working at about 3k down).
However, I don't doubt that the geosciences are underfunded and I think its a shame.
There seems to be a trend in high energy physics towards larger and larger collabrations due to the increasinly high costs to conduct experiments. Do you believe that the physics community should focus on these increasingly expensive experimenmts or should the focus shift to other areas such as nonlinear dynamics, fluid mechanics and interdisciplinary fields such as biophysics?
Also, I worked with a group that was collabrating on the Superkamiokande experiment which found evidence that neutrinos oscillate and thus have mass. How would these results affect the current standard model and the various attempts to formulate an unified theory.
Slashdot Mirror
I-Opener's model is based on selling the internet service no on selling hardware so I don't see what's so bad about making people take the hardware as is and not modifying it. This really has nothing to do with OpenSource. I-Opener made a mistake by not requiring buyers of the hardware to sign up and now they're fixing it. Expecting them to keep selling hardware for 99 or whatever the price was would be like expecting Sun to keep selling E10Ks for 3000 after they made a mistake and sold a few at that price. It was good deal while it lasted but to expect to them to keep doing it is ludicrous
I belive you're wrong there. Although beowulf clusters are nice for applications that are computationally intensive but don't require much bandwidth, a lot of problems require a lot of bandwidth. This is where SMP boxes rule since bandwidth between processors on a single motherboard/backplane is a LOT larger than across a network. For these types of problems, the time required is more function of bandwidth rather than cpu so a smp box with fewer processors but more bandwidth between the processors will beat even a large beowulf/clustering solution.
The article talks about how FP on the x86 is stack based. However, I seem to remember intel switching to a stack/register model back in the pentium. I believe all the recent x86 chips have allowed you to address stack locations using F(1), F(0), etc. but you can still treat the registers as a stack if you want. I may be wrong though.
The most important detail that the article leaves out is how the new 64 bit extensions will work. Personally I'm hoping that AMD keeps the current instructions but creates 64 bit extensions that work in a more RISClike fashion. Something like limiting addressing modes to register and register+constant, do the more complex instructions in microcode, and most importantly add another 10 or 20 gp registers to the integer code.
Actually your eyes wouldn't pop even in a zero atmospehere. Our body's structual integrity is strong enough to handle even zero pressure without much more damage than a few bruises. Consider putting a vacuum cleaner hose on your arm, the pressure at the end will at least be 7 psi but you won't get more than a small bruise. Also, divers working at 300ft down work in pressures from 20-30 atm (20-30x normal pressure) without problems.
Also death by decompression is painful, but that's more due to gas bubbles forming in your capillaries giving you microstrokes all over your body NOT from your body exploding.
My problem was with the way the information is presented. Instead of saying that the original story was incorrect but the update to the original was correct, Hemos said the same thing but in a way that leads the reader to think that the original was correct. Given that the headline of that story was about Sony recalling some the memory cards, most people are probably going to remember that.
An analogous example are those sweepstakes that seem to suggest that you need to buy something in order to enter but at the very end say that no purchase is necessary. I just think its sleazy when someone makes a mistake, and then frames the apology/correction in such a way as to suggest that the original mistake was not a mistake after all.
So what this is really saying is that Slashdot originally got the story wrong. A lot of people probably did not see the update and Hemo's little sentence implies that the original story was right. This something you would expect to see out of pr people. Sure its technically correct, but to people who aren't reading the description carefully it appears to say that the original story was correct when it wasn't. It would have been a lot clearer to say something like "So are orignal story was wrong but the update was right".
I don't think you understand that for a lot of the algorithms that break things up into cells, the accuracy comes from the size of the cell NOT the number of particles in the cell. In other words, having 90,000 particles inside the cell is the same as having 1. It simply doesn't make a difference.
The web page says nothing about nuclear simulations. Its a few molecular dynamics/astrophysics/pde applications.
The problem is that you won't be sure if it screwed up on a calculation or not, especially since you don't know before hand what the results are supposed to look like. Suppose the result was off by a factor of .001%. Although it may look harmless, the error propagated through a 1000 iteration may result in a large final error. So the software says that a certain configuration of plutonium won't explode when it really will or the simulation predicts the wrong trajectory for a hurricane. In other words, most people would prefer spending an extra million or waiting a few more hours/days to get the correct result.
The problem is that each particle's position needs to be sent to all the other nodes for them to do their calculations. So taking your assumption that each particle takes 34 bytes and assuming the simulation has a million particles in it, you need to send at least 34 MB of data through the network after each calculation. So assuming the calculation takes a while, say 10,000 clock cycles (on a 500Mhz machine) then you need to do 50,000 updates a second. This comes out to about 1,700,000 MB/s of bandwidth. Even if the calculation takes 1,000,000 clocks, you still need at least 17GB/s bandwidth. A beowulf cluster may have a 1 Gb/s bandwidth over gigabit ether but that leaves you short by a factor of 136.
Using a beowulf for the example above means you can only do 30 calculations per second even though you could theoretically do 500 to 50,000 depending on the amount of clocks the calculation takes.
Now suppose you had a supercomputer that had a bandwidth of 10GB/s (gigabyte/s). Then you could do about 300 calculations per second even if your processors were running at 300MHZ instead of 500MHz. In cases like this, the I/O bandwidth determines how fast the simulation will go and a beowulf cluster would not be better.
OGRs have application to data communications, cryptography and lithography. I would say that it has a lot of use since it may lead to faster/better encryption/data transfers as well as better/cheaper chip fabs and indirectly cheaper cpu and microprocessors. A lot more useful than pretty pictures,
Well you need oxygen for the combustion to occur. Since the engine won't store oxygen and will presumably use air from the atmosphere to obtain the oxygen, you get nitrogen also. Couple this with the heat the combustion produces and you'll get N combining with O to produce NO_x.
Well, that's whatleftovers are for. If you make a stew or chili, you usually have a lot of it left over and if you keep it in tupperware or some other storage container it'll be available for a late night snack. If you have a potato, pop it in the microwave and in less than 6 minutes you have a baked potato with chili or stew.
Maybe its just me but I never understood why ramen is so popular. If price is a factor then sphagetti or pasta is available for slightly more. It takes about the same time and tastes a lot better. Hell, baked potatoes cook in the same time as ramen and taste much better.
If cost isn't a consideration, something like lamb stew, chili, quiche, etc. take about 10-15 minutes to prepare (they take about 30-60 min to cook, but you can do something else while they're cooking).
I'm not sure where this thing with geeks eating only pizza and ramen comes from. I'm a math/cs major and most of the other physics/math/cs I know don't eat that stuff on a regular basis. At our parties, you're more likely to see stuff like salmon steaks, grilled portabella mushrooms, fresh baked cookies than pizza or ramen.
Don't use xhost +. That leaves your system open to anyone. Someone could start a hidden X program that looked at the keystrokes you were entering and snag your passwords. That's one of the reasons xterm has a secure keyboard mode.
I'm sorry in advance if your the guy who posted this to bugtraq. But the exact same message appeared with the domain names changed about 3 weeks ago in bugtraq. Next time be a little more creative.
This problem has been discussed a fair bit in bugtraq. The consensus was that DNS wasn't really secure using the crypt and signed message may help to prevent this but in general were not that great since netsol sometimes ignore crypt-pw and their pgp signed mail thing is often broken. Essentially if someone can forge their header so that it looks like its coming from the technical contact, it probably go through.
I think the point of the article was that good UI doesn't necessarily equate to having a GUI interface. It's more on how programs have a shared consistent interface. For example, the shortcut to quit in Gnome programs is ctrl-q, in netscape and some others its atl-q, xemacs and emacs set it to ctrl-x-ctrl-c. If I want to quit a program, I don't really want to have to stop a second and figure out what the shortcut is, or end up pressing two or three shortcuts in other to quit. The MacOS and WinXX systems are a lot better in this regard since certain common functions are mapped to the same shortcut in all programs. If I want to quit a program while using those oses I can just hit alt-f4 or apple-q and know that it'll work.
People aren't really kick ass at undecidable problems. People aren't able to decide whether certain programs halt or not. If the program is trivial then a computer can decide whether it halts or not probably with the same effectiveness as a person. The undecidability problem just says that there is no algorithm that can figure out whether all programs halt or not. An algorithm may be able to decide whether a certain subclasses of programs halt or not.
In addition, people most certainly do not kick ass at NP-complete problems. To solve any problem, we need an algorithm to do it and all such algorithms are nonpolynomial. If anything people would be slower at solving these problems since we can't compute things as quickly as computers.
Regardless, it will have to be turned off even if it is for battery changes. Also people may drain the battery after using it for a while, requiring a power down. These points aside, even low power devices today are turned off (i.e. laptops, palm pilots, etc.) since even standby mode drains too much power. The crusoe systems will probably drain more power than a palm pilot so you'll probably need to turn it off.
Actually switches like this are are already present. They let developing embyros regulate their own growth so that things like a brain develop only after a skull has formed. More importantly they let cells in the embyro figure out where they are so that they can become the right sort of cells. Errors in these are responsible for a lot of birth defects.
The switches(promoters & repressors) are also called operons. Their existence and functions have been known since at least the mid 80s and possibly even earlier. The most well known system is the Lac operon in E. Coli which triggers the transcription and generation of a protein that allows the utilization of lactose when there is no glucose present and a sizable lactose concentration. It's been pretty well studied and understood for a while. Although the article doesn't mention, the researchers were probably able to reliably splice a gene into a system like this and probably did not create the operon from scratch.
I have some concerns about the performance that the Cruose processors will actually have. The article mentions that translated instructions will be cached and then be reused if the CodeMorph software sees it again. However, it seems like the CodeMorph's state information will not be mantained between runs. If you power off the computer, the software loses the cached information and has to start from scratch again. In addition, the cache's size or location isn't given. Is it a small cache on die or is it located in system memory? The cache is probably on die for speed reasons but this would limit the size of the cache. This could be a performance hit since the cache is also used as a data cache and instruction cache.
Another question concerns the way the instructions are being cached. For example suppose the following instructions were given
ADD AX, BX
SUB CX, AX
JNZ Cx
Would the translation for each instruction be cached, or is the sequence cached? The article implies that the sequence is cached since the CodeMorph software can optimize the speed on subsequent passes. However, this seems to limit the benefit gained from caching to relatively tight loops or common sequences of code depending on the cache size.
On a side note, the article implies that the CodeMorph software lightyears beyond anything else. However, some of its highly touted features appeared in other software before. For example, DEC's FX!32 would initially just translate code but would also observe the application behaviour and then optimize the code based on that after the application finished executing. It could do this optimization several times, optimizing more aggressively on each pass. Also Apple's 680x0 emulator was also based in rom that would start up initially so what the MacOS could boot. The CodeMorph software has some new features if it really does OO scheduling and optimization on the fly but that seems like a pretty big hit on performance.
If future server/desktop oriented processors implement large parts of the CodeMorph software in hardware, how will that be any different than AMD or Intel's processors since they'll all be implementing a hardware instruction translation unit besides the Transmeta core being VLIW. Plus the transistor count and power consumption will also sky rocket along with that.
which is not to say Oceanography does not benefit from NASA, the military, etc, but I think the point is clear)
I think that oceanographers tend to get more than that figure in the form of free time on equipment. A couple of people I worked with were working on installing some instruments for DUMAND in Hawaii were getting a a few weeks every year from the Navy on their submersible. From what I understand, the submersible time was around 100k a day if they had to buy it (they were working at about 3k down).
However, I don't doubt that the geosciences are underfunded and I think its a shame.
There seems to be a trend in high energy physics towards larger and larger collabrations due to the increasinly high costs to conduct experiments. Do you believe that the physics community should focus on these increasingly expensive experimenmts or should the focus shift to other areas such as nonlinear dynamics, fluid mechanics and interdisciplinary fields such as biophysics?
Also, I worked with a group that was collabrating on the Superkamiokande experiment which found evidence that neutrinos oscillate and thus have mass. How would these results affect the current standard model and the various attempts to formulate an unified theory.