In the late 1950s, psychologies found that the human brain can only keep about 7 "things" in mind at any one time. This varies depending on various factors, but not by much. It was because of this that telephone numbers, for example, were designed to have seven digits.
The definition of "things", of course, depends on which level you're looking at. If you've got a highly complex system, with dozens of elements, it will be impossible to understand it all at once, unless the elements are subsumed into some sort of hierarchical structure.
Xt and Motif's fatal flaw is that it violates this, failing to provide enough abstraction to allow the programmer to easily get their head around it. As such, Xt/Motif programmers are forced to flip through reference books. (Think of the Xt/Motif API/model as the cognitive equivalent of a horribly bloated program, far too large to fit in memory, and not designed well enough to minimise swapping.)
Aye to that... For a long time, XView was my favourite GUI toolkit. It's the only one which can be programmed in C without line upon line of function calls and huge structs with pointers to functions. Not to mention that the UI looked much better than anything else on UNIX/X; it looked as if they actually had some professional designers do it.
Btw, has anyone appropriated the OPEN LOOK look and feel for Gtk or Qt?
I can't comment on what it's like using Xt from C++, because C++ is an abomination that I would never use. But Xt makes it pretty straightforward to code C in an object-oriented style. Straightforward in a rather painful way. If you don't mind defining lots of functions and stringing them together with big ugly tables, all the while paging back and forth through volumes of documentation. Then again, I found Gtk less than perfect as well. IMHO, object-oriented programming in C is a masochistic exercise. Even C++ (as horribly malformed as it is) is better. For a recent project, I started looking at Gtk, but went over to Qt, and ended up writing it in C++ using STL as well. -- acb [Where Python is not an option, C++ will do...]
Are you sure about this? The Commodore seems like a pretty poor environment for a virus to spread (outside of a testbed environment); it has no operating system proper, no relocatable executables, no hijackable system services (everything's ROM) and very primitive memory management. And in a "10 SYS 2061"-type program, there's nowhere a virus could insert itself.
Unless you refer to the trick by which the cartridge signature and start vector were written into the RAM under where the cartridge ROM would go (it only had 64K address space, so things had to double up, or even triple up), so that on reset the machine would think it had a cartridge and jump to the address specified. A lot of games did this.
I believe they used static RAM rather than dynamic RAM. Static RAM doesn't need to be refreshed, just supplied with electricity. It's more expensive than dynamic RAM, which is why it went out of fashion for main storage once RAM capacities went up. The VIC-20 used static RAM (all 5K of it, in small-denomination chips).
Also, the slow clock rate and smaller RAM size may have meant that the RAM chips used more current per bit, and thus taken longer to drain.
They can communicate by extremely low-power transmissions, or by modulating the body's electrical field. One designated nanobot can take the role of transmitter, sending the data back.
Alternatively, in-brain communication could be done by having messenger nanobots do the rounds of the ones at neurons, relaying information.
The Chinese Room is a classic case of argument by anthropocentric chauvinism. The key assumption is that the guy in the room is the only thing that can embody consciousness, and that the room as a totality cannot be counted as conscious. Basically, it is a mobilisation of unconscious prejudice about "what it means to be human" as a refutation of AI.
If you restrict the marketing to hardcore geeks who spend all their waking time sitting in darkened rooms at a terminal, this wouldn't happen. The flow of digital data (what actually counts) would be much better than in RL, and as long as video and audio are on the standard of a high-res monitor and a good set of speakers respectively, it'd be good enough.:-)
Or like the Ndoli Device ("Jewel") in Greg Egan's Learning To Be Me.
(It's a neural implant inserted at birth which learns to mimic the brain it's connected to. By the time brain development finishes and before deterioration begins, control is switched from the brain to the Jewel. That's just a brief summary; for more info, read the story (in Axiomatic).
It could be that human insight produces results which are computable, only not by means we know enough about to consciously model.
Penrose's theories on how quantum effects may affect/produce consciousness are vacuous at best; they consist primarily of fanciful speculation about gravity in microtubules in the brain, something which has no scientific evidence to lend it weight.
It is simpler and more elegant to imagine that consciousness is the product of a massively complex computational system with self-reference, and that there is no need for a 'soul', or any vestige of Cartesian dualism. Neurology has shown (and is showing) that many aspects of the brain and higher-level human behaviour are mechanistic. Speculation about quantum effects responsible for consciousness is likely to be unnecessary, and sounds too much like Cartesian dualism.
From what I understand, Dixie Flatline wasn't a true copy of a mind, but more like an expert system trained to mimic the mind at the time of programming; i.e., it couldn't learn new things or change. I'd thus dispute whether it's an instance of uploading per se.
Funny you should mention Avid. Didn't they stop supporting Apple's QuickTime format and/or MacOS, thus making their video editing systems Windows-specific, for fear of retaliation from MS?
I seem to recall that Microsoft spun off Softimage (which it had never really assimilated, and which had remained based in Montreal) some months back.
Basically, the only effect of MS's acquisition of Softimage had been ports of its software to NT, which then started to make inroads into graphics/rendering/animation. Even MS didn't have the clout to abandon SGI support.
BMG are evil. More so than the other three big record companies. This is exactly the sort of shit one would expect them to pull.
The thing to know about BMG is that they are control freaks. They play hardball. They aggressively prosecute the copying of long-deleted albums once released on their label, to an extent that the others don't. They're not called the Big Mean German for nothing.
Fortunately they don't seem to release much worth getting (other than David Bowie's single releases). Unless they own lots of "independent" labels and keep quiet about it.
Printing large quantities of counterfeit currency in order to destabilise a country's economy (as Iran and the USSR were said to have done) is considered "economic terrorism". From that, it would be possible for a lawyer to make a link to making counterfeit DVDs, and then to cracking protection.
Even if he ultimately wins, and after a few years of protracted struggle gets his computers back, he will have been inconvenienced massively by this action, and any compensation granted will be insignificant. The reasoning goes, such a spectacle will serve to deter others who might mock the authority of the copyright barons.
If you knew you could end up losing your computers, all your files and possibly your freedom, would you publically release something like DeCSS? Probably not, unless you've a yen for martyrdom.
If they made it so anybody found bringing a member of the opposite sex into their dorm room would face instant expulsion from the dorm and suspension or expulsion from college, and that any staff wilfully collaborating in the breaking of this law (i.e., not enforcing it) would face disciplinary actions, it could be made workable, in a fascistic sort of way. All they'd need to do is to make an example of one or two offenders and the rest would fall in line pretty quickly.
Apparently BMG were on the verge of buying either EMI or Sony Music.
Now that EMI seems to be taken, they may have to talk to Sony. Though if Microsoft is split up, I imagine Sony may be interested in buying one of the Windows companies; as such they may be interested in offloading part or all of their music division to an eager BMG.
EMI was the last of the big independent labels, and it will be sad to see it subsumed into AOL/TW. However, it would have been worse to see it devoured by BMG.
They call BMG the Big Mean German, and not without reason. They're nasty. They have a history of aggressively suing people for copying long-deleted records to an extent that the others don't. And then there is that Bertelsmann Foundation's think-tank looking for means of regulating Internet content.
Also, in this deal, EMI does not merge with AOL/TW outright; no shares change hands. From what I understand, EMI's music division and AOL/TW's pool their asset portfolios and create a jointly-held (though TW/AOL-controlled) company to manage them. It's a far cry from, say, the way PolyGram was subsumed into Universal (i.e., its assets were basically stripped and transferred to the new owner); had BMG bought EMI, we would probably have seen a recurrence of this.
There was once a porn site named linuxchick.com. It had no obvious Linux content, unless one counts the word "Linux" in its title banners. Seems like some pornmonger decided to repackage their content and tie it to a trendy buzzword.
Slashdot Mirror
In the late 1950s, psychologies found that the human brain can only keep about 7 "things" in mind at any one time. This varies depending on various factors, but not by much. It was because of this that telephone numbers, for example, were designed to have seven digits.
The definition of "things", of course, depends on which level you're looking at. If you've got a highly complex system, with dozens of elements, it will be impossible to understand it all at once, unless the elements are subsumed into some sort of hierarchical structure.
Xt and Motif's fatal flaw is that it violates this, failing to provide enough abstraction to allow the programmer to easily get their head around it. As such, Xt/Motif programmers are forced to flip through reference books. (Think of the Xt/Motif API/model as the cognitive equivalent of a horribly bloated program, far too large to fit in memory, and not designed well enough to minimise swapping.)
Aye to that... For a long time, XView was my favourite GUI toolkit. It's the only one which can be programmed in C without line upon line of function calls and huge structs with pointers to functions. Not to mention that the UI looked much better than anything else on UNIX/X; it looked as if they actually had some professional designers do it.
Btw, has anyone appropriated the OPEN LOOK look and feel for Gtk or Qt?
I can't comment on what it's like using Xt from C++, because C++ is an abomination that I would never use. But Xt makes it pretty straightforward to code C in an object-oriented style. Straightforward in a rather painful way. If you don't mind defining lots of functions and stringing them together with big ugly tables, all the while paging back and forth through volumes of documentation. Then again, I found Gtk less than perfect as well. IMHO, object-oriented programming in C is a masochistic exercise. Even C++ (as horribly malformed as it is) is better. For a recent project, I started looking at Gtk, but went over to Qt, and ended up writing it in C++ using STL as well. -- acb [Where Python is not an option, C++ will do...]
Are you sure about this? The Commodore seems like a pretty poor environment for a virus to spread (outside of a testbed environment); it has no operating system proper, no relocatable executables, no hijackable system services (everything's ROM) and very primitive memory management. And in a "10 SYS 2061"-type program, there's nowhere a virus could insert itself.
Unless you refer to the trick by which the cartridge signature and start vector were written into the RAM under where the cartridge ROM would go (it only had 64K address space, so things had to double up, or even triple up), so that on reset the machine would think it had a cartridge and jump to the address specified. A lot of games did this.
I believe they used static RAM rather than dynamic RAM. Static RAM doesn't need to be refreshed, just supplied with electricity. It's more expensive than dynamic RAM, which is why it went out of fashion for main storage once RAM capacities went up. The VIC-20 used static RAM (all 5K of it, in small-denomination chips).
Also, the slow clock rate and smaller RAM size may have meant that the RAM chips used more current per bit, and thus taken longer to drain.
The philosophy seems to be the same one that has made Linux good: It'll be done when it's done.
But will it be much better than MSIE 5.0? Other than it running on Linux, what advantages will it have over IE?
-- acb, running Netscape 3.0 but tempted to buy a more powerful machine to run vmware and IE5
They can communicate by extremely low-power transmissions, or by modulating the body's electrical field. One designated nanobot can take the role of transmitter, sending the data back.
Alternatively, in-brain communication could be done by having messenger nanobots do the rounds of the ones at neurons, relaying information.
The Chinese Room is a classic case of argument by anthropocentric chauvinism. The key assumption is that the guy in the room is the only thing that can embody consciousness, and that the room as a totality cannot be counted as conscious. Basically, it is a mobilisation of unconscious prejudice about "what it means to be human" as a refutation of AI.
If you restrict the marketing to hardcore geeks who spend all their waking time sitting in darkened rooms at a terminal, this wouldn't happen. The flow of digital data (what actually counts) would be much better than in RL, and as long as video and audio are on the standard of a high-res monitor and a good set of speakers respectively, it'd be good enough. :-)
Or like the Ndoli Device ("Jewel") in Greg Egan's Learning To Be Me.
(It's a neural implant inserted at birth which learns to mimic the brain it's connected to. By the time brain development finishes and before deterioration begins, control is switched from the brain to the Jewel. That's just a brief summary; for more info, read the story (in Axiomatic).
It could be that human insight produces results which are computable, only not by means we know enough about to consciously model.
Penrose's theories on how quantum effects may affect/produce consciousness are vacuous at best; they consist primarily of fanciful speculation about gravity in microtubules in the brain, something which has no scientific evidence to lend it weight.
It is simpler and more elegant to imagine that consciousness is the product of a massively complex computational system with self-reference, and that there is no need for a 'soul', or any vestige of Cartesian dualism. Neurology has shown (and is showing) that many aspects of the brain and higher-level human behaviour are mechanistic. Speculation about quantum effects responsible for consciousness is likely to be unnecessary, and sounds too much like Cartesian dualism.
From what I understand, Dixie Flatline wasn't a true copy of a mind, but more like an expert system trained to mimic the mind at the time of programming; i.e., it couldn't learn new things or change. I'd thus dispute whether it's an instance of uploading per se.
Funny you should mention Avid. Didn't they stop supporting Apple's QuickTime format and/or MacOS, thus making their video editing systems Windows-specific, for fear of retaliation from MS?
I seem to recall that Microsoft spun off Softimage (which it had never really assimilated, and which had remained based in Montreal) some months back.
Basically, the only effect of MS's acquisition of Softimage had been ports of its software to NT, which then started to make inroads into graphics/rendering/animation. Even MS didn't have the clout to abandon SGI support.
Didn't extradition come up in the Analyser case (i.e., that of that Israeli cracker from a few years ago)? If so, it could be precedent.
BMG are evil. More so than the other three big record companies. This is exactly the sort of shit one would expect them to pull.
The thing to know about BMG is that they are control freaks. They play hardball. They aggressively prosecute the copying of long-deleted albums once released on their label, to an extent that the others don't. They're not called the Big Mean German for nothing.
Fortunately they don't seem to release much worth getting (other than David Bowie's single releases). Unless they own lots of "independent" labels and keep quiet about it.
If the MPAA wanted to prosecute Johansen, would it have a case for extraditing him to face trial in the US?
Printing large quantities of counterfeit currency in order to destabilise a country's economy (as Iran and the USSR were said to have done) is considered "economic terrorism". From that, it would be possible for a lawyer to make a link to making counterfeit DVDs, and then to cracking protection.
Wonder whether the next wave of web-site cracks will read "FREE JON"...
Even if he ultimately wins, and after a few years of protracted struggle gets his computers back, he will have been inconvenienced massively by this action, and any compensation granted will be insignificant. The reasoning goes, such a spectacle will serve to deter others who might mock the authority of the copyright barons.
If you knew you could end up losing your computers, all your files and possibly your freedom, would you publically release something like DeCSS? Probably not, unless you've a yen for martyrdom.
If they made it so anybody found bringing a member of the opposite sex into their dorm room would face instant expulsion from the dorm and suspension or expulsion from college, and that any staff wilfully collaborating in the breaking of this law (i.e., not enforcing it) would face disciplinary actions, it could be made workable, in a fascistic sort of way. All they'd need to do is to make an example of one or two offenders and the rest would fall in line pretty quickly.
Apparently BMG were
on the verge of buying either EMI or Sony Music.
Now that EMI seems to be taken, they may have to talk to Sony. Though if Microsoft is split up, I imagine Sony may be interested in buying one of the Windows companies; as such they may be interested in offloading part or all of their music division to an eager BMG.
EMI was the last of the big independent labels, and it will be sad to see it subsumed into AOL/TW. However, it would have been worse to see it devoured by BMG.
They call BMG the Big Mean German, and not without reason. They're nasty. They have a history of aggressively suing people for copying long-deleted records to an extent that the others don't. And then there is that Bertelsmann Foundation's think-tank looking for means of regulating Internet content.
Also, in this deal, EMI does not merge with AOL/TW outright; no shares change hands. From what I understand, EMI's music division and AOL/TW's pool their asset portfolios and create a jointly-held (though TW/AOL-controlled) company to manage them. It's a far cry from, say, the way PolyGram was subsumed into Universal (i.e., its assets were basically stripped and transferred to the new owner); had BMG bought EMI, we would probably have seen a recurrence of this.
Not only did he write a C compiler, he's also dating Winona Ryder. The man is a living god.
There was once a porn site named linuxchick.com.
It had no obvious Linux content, unless one counts the word "Linux" in its title banners. Seems like some pornmonger decided to repackage their content and tie it to a trendy buzzword.
It was mentioned in NtK sometime last year.