Those are peculiar choices, but I think they may be only partly intended to squat on possible "sucks" domains. Remember that the way they apparently suckered some of their VC backers was by convincing them that their barcode reader would eventually replace the mouse (that's also the point of making it look like a cat, which is otherwise pretty inexplicable). This is pretty funny because it shows how those same large company investors view web users - as consumer drones only interested in looking up product info on the net.
The really interesting thing about DC will be in watching how quickly they manage to burn through their 200 megabucks and show up on fuckedcompany.com.
This leads rapidly into a nasty, divisive nature/nurture argument. I personally do believe that the sum total of violent imagery in our society does lead to more violent action (by adults as well as kids - adults are more impressionable than they think, they just have more tools to rationalise away their actions). Nonetheless I was impressed by this essay, which takes a more moderated view (if not the opposite one):
Actually that would make it a Linux port of a UNIX command, though it's really a GNU port, particularly since it can be built for other Unixy platforms as well. There's a long history of replacing the half-broken tools that came with one's Unix distribution with the reasonably standard GNU ports of them, though I personally began to lose interest in that when they started consistently using those --extremely --long --option --strings. Not that dd can be helped in that regard.
I'd probably be willing to call something a "Linux command" if it were specifically written for Linux without portability in mind. cdparanoia would be a current example, though portability is in the works there. More likely, I just wouldn't use the term at all. But then I'm probably just another pedantic idiot.;)
And how do you know which ones to ignore? The whole point of SYN cookies is to be able to make that distinction (in a way that avoids the resource load of maintaining server side connection state before the traditional SYN-ACK).
If you just randomly drop a certain percentage of packets, the DOS attack has simply succeeded. Indeed this will happen on its own when the server's TCP stack and/or interceding equipment is overloaded - the point of the attack.
Most software that's worth a crap _won't_ make you change or patch your system- if it tries, get rid of it, run Macsbug and if it installs stuff and says 'Reboot now (OK)' and there's no 'no, that's not OK!' button, cackle and hit 'cmd-power' and drop into the low level debugger and go 'es' (escape to shell) and get rid of whatever got added.
Just to note, this sequence of actions is pretty much exactly equivalent to a regular force-quit (command-option-escape, without Macsbug). Not that having Macsbug around isn't fun and useful in its own right.;)
Of COURSE a modern true-multitasking OS is going to require more hardware than a 2 decade old non-multitasking OS.
While I agree with the gist of your comment, there's no real reason to believe the above - nothing about preemptive multitasking per se increases the hardware requirement over a singletasking system (the overhead is generally miniscule). Lots of people have written multitasking OSes for the 6502, for instance.
Real memory protection, a different issue, does require some sort of executive memory management unit within or external to the CPU. It'd probably also like a CPU that has at least a supervisor and user mode. But I suspect by "hardware" you meant more memory, CPU speed, and I/O bandwidth, and those aren't required.
Not to detract from the sheer awesomeness of the hack, but is it really "woodgrain' when he appears to have used actual wood? I mean, I thought woodgrain implied a certain level of cheezy 1970's fakeness.
It does just make the thing look better (than a real 2600) of course, and he did do a really nice job on the shroud.
Hmm, that hex encoded ASCII "encryption" technique looks familiar...rather like the way slashdot stores userids and passwords in a login cookie to be sent over the net in near-cleartext with every damn HTTP request. And about as useful for preventing interception.
%2532%2532%2535%2532%2535%2534%253a%253a etc...
If you're going to pass the text through some automatic munger anyway, why not use strong encryption? And in the case of slashdot user cookies, it should be a one-way hash (an account can still be temporarily hijacked, but the password wouldn't be compromised and thus couldn't be changed).
I think you underestimate the state of the art in OCR. If human eyes can make sense of it, computer ones probably can too. The only way to really avoid this is to make it into a visual rebus sort of puzzle - require the human to exercise language comprehension as well as character (or word or even sentence) recognition. This has the advantage of making terrorism fun, but since these sorts of games generally play with ambiguities in language you also open up the possibility of misinterpretation by the intended recipient (I didn't mean to bomb *that* kindergarten!).
However, I think the original poster was referring to digital steganography, where the message is distributed through the (digital encoding of the) image as ordered noise - digital watermarks use this. This one is strictly digital though.
Or in other words, the API abstracts away a great deal of actual code, which takes us full circle to the original post - OpenGL *is* doing a lot of work (more than half of it, according to the post above). The fact that it only takes 2% of your application code is why OpenGL is so cool - the people who designed and implemented it have written upwards of 90% of your game for you.
Or, they're spammers - that's what we call other people who flood publicly-accessible networks with advertising.
Note that I don't actually think BNL is spamming, both because of the typically cute and funny way they've gone about this (listen to the track), and because you can't really "flood" Napster - it's fundamentally a receiver-driven system, and the most you can do is put your own versions up on servers you control and hope people (a) download them, and (b) include them in their own shared store so they'll propagate. In fact that's probably why they made the "advertising" funny and only mildly intrusive - what these really are are *remixes* that people will find interesting in their own right. I'm not actually a fan of Napster - I think it's a lame, Windows-designed protocol and it just works poorly - but it's at least not subject to being flooded or spammed automatically.
Three person-years to come up with a BASIC for the Altair (let alone that Gates and Allen had to hire someone else even to get started - hmm). Shortly afterward, Steve Wozniak wrote Integer BASIC for the Apple ][ in his head (after first designing the computer to run it on, natch).
It seems that paying through the nose to fund Gates's and Micro-Soft's incompetence in software development is not a new thing - and neither is the availability of better alternatives.
It's not just acid rain - my understanding is that the average pH of a landfill is quite low. The other stuff that's thrown out with the lead (circuit boards et al) probably has a lot to do with this, acidifying the water that naturally flows through so that nasty heavy metals are liberated much more quickly.
As for mercury, the poor old Mad Hatter got it much worse than you might think - they not only worked with the mercury, but licked the brims of felt hats to stiffen them. Mercury and lead are very rapidly absorbed across a mucous membrane (note to slashdotters - don't masturbate immediately after handling lead solder - or for you libertarians, lead ammo;).
Queen Victoria's death is also attributed to the heavy metals (cadmium?) in her Virgin-white makeup.
The really funny thing here is that this would duplicate the joke in the original post for blind people. That is, it takes the existing 3D sensory cues they have about the world and masks it with an inferior synthesised version.
Possibly you could come up with a useful echolocation device, but you'd want to do it without interfering with the detailed auditory sense of the world a blind person already receives (or a sighted person, we just tend to ignore it, especially those of us who use computers too much...).
Beyond the bizarre idea that "sucking up bandwidth", or even copyright violation is grounds for search and seizure - sure, the University owns the connection. It'd be perfectly reasonable for them to pull the connection and not return it unless the student provided some evidence that he wasn't going to engage in this behaviour any more (probably involving voluntary access to his equipment). Seizing it goes way over the line, and it's particularly galling that it was performed by campus rent-a-cops instead of the real police. That kind of delegation of police authority never bodes well IMO, especially when the real police are already acting as industry schills in this case.
On the one hand, you say "he was doing something illegal, good thing they got him". On the other hand you say that you do legally the exact same thing. The fact that the mode of distribution differs in your eyes is an immensely hypocritical rationalisation. The law doesn't care and wouldn't make that distinction - you (and I) practice the same copyright infringement as he does (he wasn't selling this stuff for profit). That's why you should be concerned.
Don't confuse legality with morality (and don't assume anyone else will agree with your morality anyway).
Pretty cocky for someone who got it wrong.;) Ninth grade level test, perhaps.
Shipping driver source with the converted binaries isn't the loophole Perens is referring to, and it's not the problem Becker has with their action. It is a mildly interesting issue, since the drivers are converted from binaries built from GPL'd source (rather than directly from the source itself). I doubt there's much question that the original source should be made available, but it's not actually stated anywhere that they failed to do this. They don't need to include the source of the tool that makes the binary conversion, though, any more than they would need to for their compiler.
The issue is whether they can incorporate (or encourage people to incorporate) GPL'd device drivers into a non-GPL'd kernel. That's why the manner of linking is such an issue. This is what Becker is up in arms about, and what Perens regrets in the current GPL. It's not actually clear to me that this is something that really needs fixing, though.
It's also interesting to note that in a proper microkernel (or even "minikernel") this wouldn't be an issue, since a driver on that sort of system isn't part of the kernel at all, being basically a user process. Which is probably the way things ought to be anyway, for other reasons.
Psychological stress is a factor in nearly every illness, and it's particualrly unsurprising in RSI. I'd go further and venture that the people who've suffered the worst RSI from desk clerk and data entry positions are those who feel the least control over their job - they're not just at the scant mercy of their superiors, but of the balky computing machinery they're forced to use. This would explain why people who generally feel more control over the computers they use (programmers and other relatively technosavvy types) tend to be less susceptible to RSI, and why women have generally been worse struck by it (because they're still given the shitty powerless jobs in way disproportionate numbers, and thus their job stress is higher).
I have long felt that worsening other people's job stress is a form of assault. When you do it en masse, it ought to be prosecuted as a crime. So when Microsoft (convenient example) makes everyone's life more difficult with unreliable software and lousy user interfaces, they are physically as well as psychologically harming them. Class-action assault.
Hmm. You make a valid point, but the intent of parts of the DMCA is still to restrict certain kinds of expression (as well as activities) that were formerly not illegal, and perhaps even protected in the US. Most laws aren't really fleshed out until there's a legal precedent, and Kaplan couldn't have reached his decision in this civil case without the existence of the DMCA. So I dunno; since the Act, as part of its intent, puts a chill on this form of expression, distinguishing it from political censorship seems like six of one half-dozen of the other. It is however certainly a new form of censorship-style restriction, and that's what's so worrisome.
More likely mp3.com only has a certain number of categories they fit unnacceptable tracks into - offensive lyrics, copyright violation, New Country, that sort of thing. Offensive lyrics was probably the closest they could come on this one, in terms of their user agreement.
If nongovernment organisations are forced to restrict third-party expression because it's illegal for them to do otherwise (as with the DMCA), it's censorship - legislative government censorship. If they chose not to post the song because the word "void" were semantically equivalent to "fuck" and would offend their listenership, that'd be a different matter. In this case they removed the track because of the legal chill created by the DMCA and the recent DeCSS decision. Censorship doesn't have to be as direct as burning books (otherwise we'd have nothing to fear of net censorship, by definition).
Public companies can't make as many of their own rules as privately-held ones - their fiduciary obligation to their shareholders restricts them in many ways (enforceable through civil law) and their are SEC restrictions as well (under criminal law). For instance, one reason Apple is charging for that Mac OS X beta is probably that their lawyers told them they have to - if there's the possibility of making a profit by selling it, they can't legally give it away.
I get that the x86 is a piece of crap and all that, but why did Yodaiken (and someone else implicitly in the parent post) comment that the best-case latencies for the PPC and Alpha are so much better than for the x86 - what's the technical reason for this?
Slashdot Mirror
Those are peculiar choices, but I think they may be only partly intended to squat on possible "sucks" domains. Remember that the way they apparently suckered some of their VC backers was by convincing them that their barcode reader would eventually replace the mouse (that's also the point of making it look like a cat, which is otherwise pretty inexplicable). This is pretty funny because it shows how those same large company investors view web users - as consumer drones only interested in looking up product info on the net.
The really interesting thing about DC will be in watching how quickly they manage to burn through their 200 megabucks and show up on fuckedcompany.com.
This leads rapidly into a nasty, divisive nature/nurture argument. I personally do believe that the sum total of violent imagery in our society does lead to more violent action (by adults as well as kids - adults are more impressionable than they think, they just have more tools to rationalise away their actions). Nonetheless I was impressed by this essay, which takes a more moderated view (if not the opposite one):
http://www.motherjon es. com/reality_check/violent_media.html
Skins are hollow, two-dimensional shells with no actual substance inside. Elected officials are...wait.
Actually that would make it a Linux port of a UNIX command, though it's really a GNU port, particularly since it can be built for other Unixy platforms as well. There's a long history of replacing the half-broken tools that came with one's Unix distribution with the reasonably standard GNU ports of them, though I personally began to lose interest in that when they started consistently using those --extremely --long --option --strings. Not that dd can be helped in that regard.
;)
I'd probably be willing to call something a "Linux command" if it were specifically written for Linux without portability in mind. cdparanoia would be a current example, though portability is in the works there. More likely, I just wouldn't use the term at all. But then I'm probably just another pedantic idiot.
And how do you know which ones to ignore? The whole point of SYN cookies is to be able to make that distinction (in a way that avoids the resource load of maintaining server side connection state before the traditional SYN-ACK).
If you just randomly drop a certain percentage of packets, the DOS attack has simply succeeded. Indeed this will happen on its own when the server's TCP stack and/or interceding equipment is overloaded - the point of the attack.
Just to note, this sequence of actions is pretty much exactly equivalent to a regular force-quit (command-option-escape, without Macsbug). Not that having Macsbug around isn't fun and useful in its own right. ;)
While I agree with the gist of your comment, there's no real reason to believe the above - nothing about preemptive multitasking per se increases the hardware requirement over a singletasking system (the overhead is generally miniscule). Lots of people have written multitasking OSes for the 6502, for instance.
Real memory protection, a different issue, does require some sort of executive memory management unit within or external to the CPU. It'd probably also like a CPU that has at least a supervisor and user mode. But I suspect by "hardware" you meant more memory, CPU speed, and I/O bandwidth, and those aren't required.
Not to detract from the sheer awesomeness of the hack, but is it really "woodgrain' when he appears to have used actual wood? I mean, I thought woodgrain implied a certain level of cheezy 1970's fakeness.
It does just make the thing look better (than a real 2600) of course, and he did do a really nice job on the shroud.
Hmm, that hex encoded ASCII "encryption" technique looks familiar...rather like the way slashdot stores userids and passwords in a login cookie to be sent over the net in near-cleartext with every damn HTTP request. And about as useful for preventing interception.
%2532%2532%2535%2532%2535%2534%253a%253a etc...
If you're going to pass the text through some automatic munger anyway, why not use strong encryption? And in the case of slashdot user cookies, it should be a one-way hash (an account can still be temporarily hijacked, but the password wouldn't be compromised and thus couldn't be changed).
I think you underestimate the state of the art in OCR. If human eyes can make sense of it, computer ones probably can too. The only way to really avoid this is to make it into a visual rebus sort of puzzle - require the human to exercise language comprehension as well as character (or word or even sentence) recognition. This has the advantage of making terrorism fun, but since these sorts of games generally play with ambiguities in language you also open up the possibility of misinterpretation by the intended recipient (I didn't mean to bomb *that* kindergarten!).
However, I think the original poster was referring to digital steganography, where the message is distributed through the (digital encoding of the) image as ordered noise - digital watermarks use this. This one is strictly digital though.
Or in other words, the API abstracts away a great deal of actual code, which takes us full circle to the original post - OpenGL *is* doing a lot of work (more than half of it, according to the post above). The fact that it only takes 2% of your application code is why OpenGL is so cool - the people who designed and implemented it have written upwards of 90% of your game for you.
Or, they're spammers - that's what we call other people who flood publicly-accessible networks with advertising.
Note that I don't actually think BNL is spamming, both because of the typically cute and funny way they've gone about this (listen to the track), and because you can't really "flood" Napster - it's fundamentally a receiver-driven system, and the most you can do is put your own versions up on servers you control and hope people (a) download them, and (b) include them in their own shared store so they'll propagate. In fact that's probably why they made the "advertising" funny and only mildly intrusive - what these really are are *remixes* that people will find interesting in their own right. I'm not actually a fan of Napster - I think it's a lame, Windows-designed protocol and it just works poorly - but it's at least not subject to being flooded or spammed automatically.
It seems that paying through the nose to fund Gates's and Micro-Soft's incompetence in software development is not a new thing - and neither is the availability of better alternatives.
It's not just acid rain - my understanding is that the average pH of a landfill is quite low. The other stuff that's thrown out with the lead (circuit boards et al) probably has a lot to do with this, acidifying the water that naturally flows through so that nasty heavy metals are liberated much more quickly.
;).
As for mercury, the poor old Mad Hatter got it much worse than you might think - they not only worked with the mercury, but licked the brims of felt hats to stiffen them. Mercury and lead are very rapidly absorbed across a mucous membrane (note to slashdotters - don't masturbate immediately after handling lead solder - or for you libertarians, lead ammo
Queen Victoria's death is also attributed to the heavy metals (cadmium?) in her Virgin-white makeup.
The really funny thing here is that this would duplicate the joke in the original post for blind people. That is, it takes the existing 3D sensory cues they have about the world and masks it with an inferior synthesised version.
Possibly you could come up with a useful echolocation device, but you'd want to do it without interfering with the detailed auditory sense of the world a blind person already receives (or a sighted person, we just tend to ignore it, especially those of us who use computers too much...).
Beyond the bizarre idea that "sucking up bandwidth", or even copyright violation is grounds for search and seizure - sure, the University owns the connection. It'd be perfectly reasonable for them to pull the connection and not return it unless the student provided some evidence that he wasn't going to engage in this behaviour any more (probably involving voluntary access to his equipment). Seizing it goes way over the line, and it's particularly galling that it was performed by campus rent-a-cops instead of the real police. That kind of delegation of police authority never bodes well IMO, especially when the real police are already acting as industry schills in this case.
On the one hand, you say "he was doing something illegal, good thing they got him". On the other hand you say that you do legally the exact same thing. The fact that the mode of distribution differs in your eyes is an immensely hypocritical rationalisation. The law doesn't care and wouldn't make that distinction - you (and I) practice the same copyright infringement as he does (he wasn't selling this stuff for profit). That's why you should be concerned.
Don't confuse legality with morality (and don't assume anyone else will agree with your morality anyway).
Pretty cocky for someone who got it wrong. ;) Ninth grade level test, perhaps.
Shipping driver source with the converted binaries isn't the loophole Perens is referring to, and it's not the problem Becker has with their action. It is a mildly interesting issue, since the drivers are converted from binaries built from GPL'd source (rather than directly from the source itself). I doubt there's much question that the original source should be made available, but it's not actually stated anywhere that they failed to do this. They don't need to include the source of the tool that makes the binary conversion, though, any more than they would need to for their compiler.
The issue is whether they can incorporate (or encourage people to incorporate) GPL'd device drivers into a non-GPL'd kernel. That's why the manner of linking is such an issue. This is what Becker is up in arms about, and what Perens regrets in the current GPL. It's not actually clear to me that this is something that really needs fixing, though.
It's also interesting to note that in a proper microkernel (or even "minikernel") this wouldn't be an issue, since a driver on that sort of system isn't part of the kernel at all, being basically a user process. Which is probably the way things ought to be anyway, for other reasons.
Psychological stress is a factor in nearly every illness, and it's particualrly unsurprising in RSI. I'd go further and venture that the people who've suffered the worst RSI from desk clerk and data entry positions are those who feel the least control over their job - they're not just at the scant mercy of their superiors, but of the balky computing machinery they're forced to use. This would explain why people who generally feel more control over the computers they use (programmers and other relatively technosavvy types) tend to be less susceptible to RSI, and why women have generally been worse struck by it (because they're still given the shitty powerless jobs in way disproportionate numbers, and thus their job stress is higher).
I have long felt that worsening other people's job stress is a form of assault. When you do it en masse, it ought to be prosecuted as a crime. So when Microsoft (convenient example) makes everyone's life more difficult with unreliable software and lousy user interfaces, they are physically as well as psychologically harming them. Class-action assault.
Hmm. You make a valid point, but the intent of parts of the DMCA is still to restrict certain kinds of expression (as well as activities) that were formerly not illegal, and perhaps even protected in the US. Most laws aren't really fleshed out until there's a legal precedent, and Kaplan couldn't have reached his decision in this civil case without the existence of the DMCA. So I dunno; since the Act, as part of its intent, puts a chill on this form of expression, distinguishing it from political censorship seems like six of one half-dozen of the other. It is however certainly a new form of censorship-style restriction, and that's what's so worrisome.
More likely mp3.com only has a certain number of categories they fit unnacceptable tracks into - offensive lyrics, copyright violation, New Country, that sort of thing. Offensive lyrics was probably the closest they could come on this one, in terms of their user agreement.
Could you transcribe it, for those of us who refuse to support DVD by buying players or discs?
If nongovernment organisations are forced to restrict third-party expression because it's illegal for them to do otherwise (as with the DMCA), it's censorship - legislative government censorship. If they chose not to post the song because the word "void" were semantically equivalent to "fuck" and would offend their listenership, that'd be a different matter. In this case they removed the track because of the legal chill created by the DMCA and the recent DeCSS decision. Censorship doesn't have to be as direct as burning books (otherwise we'd have nothing to fear of net censorship, by definition).
Public companies can't make as many of their own rules as privately-held ones - their fiduciary obligation to their shareholders restricts them in many ways (enforceable through civil law) and their are SEC restrictions as well (under criminal law). For instance, one reason Apple is charging for that Mac OS X beta is probably that their lawyers told them they have to - if there's the possibility of making a profit by selling it, they can't legally give it away.
I get that the x86 is a piece of crap and all that, but why did Yodaiken (and someone else implicitly in the parent post) comment that the best-case latencies for the PPC and Alpha are so much better than for the x86 - what's the technical reason for this?