I'm going to have to disagree with you there. The Turing Test is basically saying "if a machine can converse well enough to be indistinguishable form a human, then it is intelligent", not the opposite - it is quite possible for a very intelligent program, human or other lifeform (e.g. dolphins) to fail the Turing Test but still be considered intelligent.
See? A positive result from the Turing Test tells us that something is intelligent, but it is vital to realise that a negative result does not mean that something is stupid. Hence your example doesn't mean that the Turing Test is ridiculous - you can have awful language skills but still be intelligent.
Consider that for a machine to successfully pass the Turing Test, it would have to be able to answer any question that you throw at it in a sensible manner, and you are right that the current crop of stimulus-response chatbots are in no way intelligent. However, I would feel quite confident in labelling a program as intelligent if it was not possible to distinguish it from a human; that kind of task requires intelligence.
So in short, you haven't made a convincing case for why you feel the Turing Test is ridiculous. I'd be interested to hear any further arguments you have.
Seriosly though, high-power, unfocused, inefficient and uncompressed radio signals - the sort of thing SETI might be able to detect - are on the way out. Nowadays, signals travel over cables, or bounce from sattelites, and in any case use compression techniques that make the signal totally useless unless you know the protocol spec.
You're right for the normal information-carrying signals we send out, but if you looked back at the Earth from a vast distance the most obvious and high-power signal you would detect would be military radars. They pump out untold kilowatts of power in all directions and would be very obvious to any aliens who were running their own version of Seti@Home.
Therefore, unless the military stop using radars, the Earth will still be detectable for many years to come. My personal guess would be that if the SETI researchers ever find anything it will be the little green men's radars, not their TV shows.
The following line from the article caught my eye:
"Wireless PCI cards using the Prism chipset are supported thanks to an OS4 driver ported over from Linux."
If they did a direct port of the code, surely that would be a GPL violation? The Linux driver would be under the GPL and therefore they would be forced to either take it out again or license OS4 as GPL.
Of course if they just used the Linux driver to reverse-engineer the workings of a Prism card, that would be acceptable - but the article sounds like that wasn't the case.
I'm assuming there must be some point I've missed, so would anybody more knowledgeable about Amigas be able to set the record straight?
You said: Can anyone offer a plausible explanation for how any one of the pieces of a bacterial flagellum would offer that bacterium some sort of advantage?
A nice quote from the conclusion:...the very fact that a step-by-step Darwinian model can be constructed that is plausible and testable significantly weakens the suggestion that extraordinary explanations might be required.
My understanding of the article is that of all the missions that have ever been launched to Mars since the start of space travel in the 50's, around 66% of them have failed.
I don't think it means that there were three specific missions, and two of these have failed - I just think it means the overall ratio of failures has been fairly high, which means that getting to the surface of Mars safely seems to be a hard problem at the moment.
I may be wrong, but my understanding of the NAT (Masquerading) used by most devices is that they change the source address of the packets they send so that it looks like the data all comes from the device itself. Certainly the Masquerading kernel option in the Linux 2.2 series kernels works this way.
Surely then, the ISP would not be able to detect a NAT/Masquerading box? It would look to them as though you still have only one PC, but you're just downloading a lot of stuff simultaneously. So why are people bothered by any terms and conditions that these ISPs write into their agreements? You could easily turn around to them and say that you are just running one PC, and there's no way they could prove you wrong short of getting a search warrant!
So where's the problem? Unless I've missed something fundamental, this seems to be a non-issue.
I would have to disagree with you there. Basically any machine (either CISC or RISC) has to do a few things to switch context - save the old register and MMU state and then restore new state.
The time taken to program the MMU's memory mapping registers should not be affected by whether the processor is CISC or RISC, so we can rule that out of the equation. Saving and loading register contents is wildly different on different architectures though, and this is where the performance would differ.
I once wrote a preemptive multitasking system for the Acorn Archimedes computers, which use ARM2 RISC chips. As they were RISC, they only had 22 instructions - but two of them were LDM and STM, which means "load multiple" and "save multiple" respectively. These instructions allowed you to dump a whole load of registers to or from a location in memory as fast as the memory bus could take it (1 cycle per register, IIRC). Also, register R15 held the program counter, the flags (including interrupt enable/disable), and the CPU mode (i.e. user or kernel).
All this niceness means that you could use one single instruction to do a context switch - with interrupts disabled in Kernel mode, you could simply say "Load multiple R0 to R15" and it would load in all the registers, enable interrupts, switch CPU mode to User and jump into the correct place in the process's code segment. This would take one memory cycle per register (i.e. 16 memory cycles).
My experience with CISC CPUs is that even though they have vastly less registers to save and restore, you need many instructions to do all the separate things that the ARM chip can do in one go. This is why I'd contend that RISC context switches are at least as fast as CISC, if not faster!
I can't find any evidence that a Sparc distribution has been produced by the Patrick and the other Slackware contributors - the URL in the article links to a packaging tool, not a distribution, and there is absolutely no information about a Sparc port currently on the Slackware website.
I'm going to give Hemos the benefit of the doubt, so surely he did some research before blindly posting the story. If so, can anyone find any confirmation of this story? As far as I can see, it's just plain false, but I'd love to be proved wrong!
You mean the X Prize (at http://www.xprize.org), a prize of ten million U.S. dollars for the first private launch of a three-person reusable spacecraft. Several teams have entries in advanced stages of development, including Dick Rutan, who built the first plane to travel all the way around the world non-stop.
What an amazingly biased and sexist statement! Let me get this straight - you're recommending we vote for a candidate purely on the basis of their gender? That is absolutely the worst way of deciding who to appoint. You didn't mention what qualifications this woman has, or why she would be a better choice than any of the other candidates. Have the last hunderd years of the Women's Rights movement not taught you that it should be ability that matters, not gender? How very disappointing.
Govt. to private corporation
on
Mir Lives
·
· Score: 1
AFAIK, Mir is now "privately" run by the RSC Energia corporation, which used to be the Soviet space agency.
So if the Russian government has given RSC Energia a load of cash, the situation is analogous to Congress allocating funds to NASA for the International Space Station. Or, to use your own analogy, it's like your mom allocating lunch money to you:)
I actually think this is a cool idea! The only problem would be that you'd have to put some form of live video broadcast equipment in there (as the station would be destroyed, so you can't retrieve a video tape later).
However, the Van Allen belts means that you get a radio blackout as it comes down, so you'd miss the most interesting part of the broadcast! It's such a shame... if anyone can work out how to do it, ring up the Russians and tell them, because I for one would love to see live on-board video of the thing burning up!
Radioactive? "Find something more envrionmentally friendly"? Give me a break! I know I'm probably responding to an obvious troll, but Mir doesn't have anything radioactive on board. Why do you think the Russians put those big solar panels on it - to make it look cool so the Mir astronauts can cruise for chicks easier?:)
I think it could protentially be quite useful, because this way game manufacturers wouldn't have to produce a "Linux version" or a "Windows version", they would just produce a generic PC version.
They could make a game that you put in the CD drive and boot from - the average user would neither know nor care that there is a Linux kernel booting off the CD drive to run his game. Joe Sixpack who plays such games treats them like putting a CD into a PlayStation and turning it on - he simply puts the game CD in his new whizzy games PC and boots it.
However, this situation would probably require the inclusion of lots of graphics and sound card drivers on the CD, and a completely automatic hardware detection routine that could boot the correct drivers up. But once such things are written, gamers could be using their PC like they use their consoles - just boot off the CD. OS? What OS?:)
I have switched to an optical mouse for purely one reason: I eat so much chocolate, crisps (potato chips) and biscuits (cookies) when I'm using my machine, that traditional ball mice get all gummed up within a few days. With an optical mouse, the crumbs actually help the camera to see the direction the mouse is moving...
I was most intrigued by the descriptiong of the bi-slicing technique that the Swedish team used in their Stage 9 key search. They realised that the DES algorithm could be implemented using only boolean operators (AND, OR, NOT etc.) and so caluclated 32 keys simultaneously by using each bit of a 32-bit word to represent a different key.
So basically they were searching 32 keys at once, which is a very clever use of resources. Does anyone know if the Distributed.Net clients use similar techniques to speed up the RC5 key searches, or is it impossible to implement RC5 using the boolean operators?
Actually, here in the UK domain names are written in exactly the same way as the rest of the world (i.e. I am posting to slashdot.org and not org.slashdot).
What I believe you are referring to is the early days of JANET, the UK's Joint Academic Network, which links all of the universities together (the *.ac.uk domain). Early on, that used the "reverse style" domain names that you were talking about, but they changed over to the correct style sometime around 1991 (if I remember correctly).
This is from memory - the only net access I had when it changed was from Uni, so it may well have been the whole UK (please correct me if I'm wrong!) But the whole point is that the UK has been using "correct" URLs and domain names for nearly yen years now!
Slashdot Mirror
$50? That's nothing! How about paying $1200 for a 6 foot power cord from Dynamic Designs?
Man, Cisco have nothing on those guys...
I'm going to have to disagree with you there. The Turing Test is basically saying "if a machine can converse well enough to be indistinguishable form a human, then it is intelligent", not the opposite - it is quite possible for a very intelligent program, human or other lifeform (e.g. dolphins) to fail the Turing Test but still be considered intelligent.
See? A positive result from the Turing Test tells us that something is intelligent, but it is vital to realise that a negative result does not mean that something is stupid. Hence your example doesn't mean that the Turing Test is ridiculous - you can have awful language skills but still be intelligent.
Consider that for a machine to successfully pass the Turing Test, it would have to be able to answer any question that you throw at it in a sensible manner, and you are right that the current crop of stimulus-response chatbots are in no way intelligent. However, I would feel quite confident in labelling a program as intelligent if it was not possible to distinguish it from a human; that kind of task requires intelligence.
So in short, you haven't made a convincing case for why you feel the Turing Test is ridiculous. I'd be interested to hear any further arguments you have.
Seriosly though, high-power, unfocused, inefficient and uncompressed radio signals - the sort of thing SETI might be able to detect - are on the way out. Nowadays, signals travel over cables, or bounce from sattelites, and in any case use compression techniques that make the signal totally useless unless you know the protocol spec.
You're right for the normal information-carrying signals we send out, but if you looked back at the Earth from a vast distance the most obvious and high-power signal you would detect would be military radars. They pump out untold kilowatts of power in all directions and would be very obvious to any aliens who were running their own version of Seti@Home.
Therefore, unless the military stop using radars, the Earth will still be detectable for many years to come. My personal guess would be that if the SETI researchers ever find anything it will be the little green men's radars, not their TV shows.
The following line from the article caught my eye:
"Wireless PCI cards using the Prism chipset are supported thanks to an OS4 driver ported over from Linux."
If they did a direct port of the code, surely that would be a GPL violation? The Linux driver would be under the GPL and therefore they would be forced to either take it out again or license OS4 as GPL.
Of course if they just used the Linux driver to reverse-engineer the workings of a Prism card, that would be acceptable - but the article sounds like that wasn't the case.
I'm assuming there must be some point I've missed, so would anybody more knowledgeable about Amigas be able to set the record straight?
You said: Can anyone offer a plausible explanation for how any one of the pieces of a bacterial flagellum would offer that bacterium some sort of advantage?
Why yes, I think we can!
A nice quote from the conclusion: ...the very fact that a step-by-step Darwinian model can be constructed that is plausible and testable significantly weakens the suggestion that extraordinary explanations might be required.
Nice try though...
My understanding of the article is that of all the missions that have ever been launched to Mars since the start of space travel in the 50's, around 66% of them have failed.
I don't think it means that there were three specific missions, and two of these have failed - I just think it means the overall ratio of failures has been fairly high, which means that getting to the surface of Mars safely seems to be a hard problem at the moment.
I think you'll find that factoring the product of two huge primes becomes slightly easier when the attacker knows one of the primes!
I may be wrong, but my understanding of the NAT (Masquerading) used by most devices is that they change the source address of the packets they send so that it looks like the data all comes from the device itself. Certainly the Masquerading kernel option in the Linux 2.2 series kernels works this way.
Surely then, the ISP would not be able to detect a NAT/Masquerading box? It would look to them as though you still have only one PC, but you're just downloading a lot of stuff simultaneously. So why are people bothered by any terms and conditions that these ISPs write into their agreements? You could easily turn around to them and say that you are just running one PC, and there's no way they could prove you wrong short of getting a search warrant!
So where's the problem? Unless I've missed something fundamental, this seems to be a non-issue.
Not all female game characters are skinny blondes - check out Lucy from Quake 3 Arena...
Did you know that the original entry in the Jargon File for this was "Eight Megs And Constantly Swapping"? Just goes to show how times change, eh? :)
I would have to disagree with you there. Basically any machine (either CISC or RISC) has to do a few things to switch context - save the old register and MMU state and then restore new state.
The time taken to program the MMU's memory mapping registers should not be affected by whether the processor is CISC or RISC, so we can rule that out of the equation. Saving and loading register contents is wildly different on different architectures though, and this is where the performance would differ.
I once wrote a preemptive multitasking system for the Acorn Archimedes computers, which use ARM2 RISC chips. As they were RISC, they only had 22 instructions - but two of them were LDM and STM, which means "load multiple" and "save multiple" respectively. These instructions allowed you to dump a whole load of registers to or from a location in memory as fast as the memory bus could take it (1 cycle per register, IIRC). Also, register R15 held the program counter, the flags (including interrupt enable/disable), and the CPU mode (i.e. user or kernel).
All this niceness means that you could use one single instruction to do a context switch - with interrupts disabled in Kernel mode, you could simply say "Load multiple R0 to R15" and it would load in all the registers, enable interrupts, switch CPU mode to User and jump into the correct place in the process's code segment. This would take one memory cycle per register (i.e. 16 memory cycles).
My experience with CISC CPUs is that even though they have vastly less registers to save and restore, you need many instructions to do all the separate things that the ARM chip can do in one go. This is why I'd contend that RISC context switches are at least as fast as CISC, if not faster!
I can't find any evidence that a Sparc distribution has been produced by the Patrick and the other Slackware contributors - the URL in the article links to a packaging tool, not a distribution, and there is absolutely no information about a Sparc port currently on the Slackware website.
I'm going to give Hemos the benefit of the doubt, so surely he did some research before blindly posting the story. If so, can anyone find any confirmation of this story? As far as I can see, it's just plain false, but I'd love to be proved wrong!
You mean the X Prize (at http://www.xprize.org), a prize of ten million U.S. dollars for the first private launch of a three-person reusable spacecraft. Several teams have entries in advanced stages of development, including Dick Rutan, who built the first plane to travel all the way around the world non-stop.
What an amazingly biased and sexist statement! Let me get this straight - you're recommending we vote for a candidate purely on the basis of their gender? That is absolutely the worst way of deciding who to appoint. You didn't mention what qualifications this woman has, or why she would be a better choice than any of the other candidates. Have the last hunderd years of the Women's Rights movement not taught you that it should be ability that matters, not gender? How very disappointing.
AFAIK, Mir is now "privately" run by the RSC Energia corporation, which used to be the Soviet space agency.
So if the Russian government has given RSC Energia a load of cash, the situation is analogous to Congress allocating funds to NASA for the International Space Station. Or, to use your own analogy, it's like your mom allocating lunch money to you :)
Measure your people
By the output they produce
All else is inferred
Use of PGP
Like poking sleeping tiger
You will be noticed
Steganography
Ancient art of hiding things
Makes useful ally
I actually think this is a cool idea! The only problem would be that you'd have to put some form of live video broadcast equipment in there (as the station would be destroyed, so you can't retrieve a video tape later).
However, the Van Allen belts means that you get a radio blackout as it comes down, so you'd miss the most interesting part of the broadcast! It's such a shame... if anyone can work out how to do it, ring up the Russians and tell them, because I for one would love to see live on-board video of the thing burning up!
I know, I'm a pyromaniac...
It means that he just has to charter a submarine instead of a rocket...
Radioactive? "Find something more envrionmentally friendly"? Give me a break! I know I'm probably responding to an obvious troll, but Mir doesn't have anything radioactive on board. Why do you think the Russians put those big solar panels on it - to make it look cool so the Mir astronauts can cruise for chicks easier? :)
I think it could protentially be quite useful, because this way game manufacturers wouldn't have to produce a "Linux version" or a "Windows version", they would just produce a generic PC version.
They could make a game that you put in the CD drive and boot from - the average user would neither know nor care that there is a Linux kernel booting off the CD drive to run his game. Joe Sixpack who plays such games treats them like putting a CD into a PlayStation and turning it on - he simply puts the game CD in his new whizzy games PC and boots it.
However, this situation would probably require the inclusion of lots of graphics and sound card drivers on the CD, and a completely automatic hardware detection routine that could boot the correct drivers up. But once such things are written, gamers could be using their PC like they use their consoles - just boot off the CD. OS? What OS? :)
I have switched to an optical mouse for purely one reason: I eat so much chocolate, crisps (potato chips) and biscuits (cookies) when I'm using my machine, that traditional ball mice get all gummed up within a few days. With an optical mouse, the crumbs actually help the camera to see the direction the mouse is moving...
I was most intrigued by the descriptiong of the bi-slicing technique that the Swedish team used in their Stage 9 key search. They realised that the DES algorithm could be implemented using only boolean operators (AND, OR, NOT etc.) and so caluclated 32 keys simultaneously by using each bit of a 32-bit word to represent a different key.
So basically they were searching 32 keys at once, which is a very clever use of resources. Does anyone know if the Distributed.Net clients use similar techniques to speed up the RC5 key searches, or is it impossible to implement RC5 using the boolean operators?
The error there stems from the line (x + y)(x - y) = y(x - y), because (x - y) is zero.
It's a subtle point, but having ax == bx does not imply a == b if x is zero!
Just my pedantic tuppence worth :)
Actually, here in the UK domain names are written in exactly the same way as the rest of the world (i.e. I am posting to slashdot.org and not org.slashdot).
What I believe you are referring to is the early days of JANET, the UK's Joint Academic Network, which links all of the universities together (the *.ac.uk domain). Early on, that used the "reverse style" domain names that you were talking about, but they changed over to the correct style sometime around 1991 (if I remember correctly).
This is from memory - the only net access I had when it changed was from Uni, so it may well have been the whole UK (please correct me if I'm wrong!) But the whole point is that the UK has been using "correct" URLs and domain names for nearly yen years now!