Shouldn't filtering be set up by kids so that their parents don't suffer the trauma of stumbling across naughty bits on the net?
It's only parents that have trouble with alleged pornography. Under-age kids either don't understand what they're seeing, or they react with "Yuk!" or else they think it's hilarious (and you've got to admit, rubbing squishy bits together has to be the funniest thing on the planet), and older kids positively love it.
And since kids understand the technology far better than their parents anyway and can easily bypass any block they like, it's really they that ought to be doing the censoring to protect their parents.
If trademark X is registered in country Y, then use of the name X in non-Y countries will create a conflict whenever items X from non-Y are used in Y.
Is this concept too difficult for judges to comprehend? Why are they even entertaining such legal action, when name clashes are such an obvious result of national trademarking?
In the rest of the world we've always wondered why the US legal system seemed so utterly screwed up, with everyone suing everyone else and only the lawyers winning in every case, but now we see the process in action, even here.
Far from trying to introduce a note of sanity, even the techie community encourages the lawyers in their headless-chicken run with their sue-everyone guns blazing, as in the response to this item. At first glance the replies seem reasonable... but you're just not looking far enough ahead. A worm's eye view won't show you the monster lying in wait just over the hill.
Well, you deserve what you get. Just don't complain later about DMCA, RIAA, MPAA, and every other darn piece of lawyer predation.
The law is not a silver bullet. A pact with the devil may be a closer analogy --- eventually you will pay the price.
In the UK, it was the Conservatives that were trying to get these powers in first, against the expressed policies of Labour.
Then Labour came into power, made a total about-face despite their pre-election statements, and brought in laws that are every bit as bad or worse than those which they had opposed.
Needless to say, the Conservatives are currently berating Labour for it, despite their previous stance. But when the Conservatives finally oust Labour again, they'll slam the iron fist down even harder --- guaranteed.
Because you see, it has nothing to do with the nature of politicians' beliefs, and everything to do with the fact that the machine of government (which is much more than just politicians) demands absolute control if it can get away with it quietly.
Well, in this instance it can get away with it, because in our innocence we've created an easily interceptable transport and storage mechanism to underpin the Internet's mail service. No big deal: let's create another mechanism which isn't susceptible to interception.
Encryption on its own is not enough to deal with the issue of mail interception. The interceptors are giving themselves progressively greater powers to ensure that encrypted mail can either be decrypted by them or used as incriminating evidence in its own right by being encrypted in the first place.
The problem is inherent in the current naive SMTP/POP3/IMAP model of mail service. It served us well for two decades, at a time of network openness and innocence. That time has now gone. DARPA researchers wanted to make their network stand the odd atomic bomb. Now we have a different evil to overcome.
There are five aspects of Internet mail services that have attracted this attention from the interceptors:
1. Mail addressing is in the clear during transport. 2. Mail content is in the clear during transport. 3. Mail storage is in the clear after delivery. 4. Senders send mail to their own ISP's servers. 5. Recipients receive mail from their own ISP's servers.
(This refers to the mail services used by the largest proportion of Internet users, the ordinary Joe Bloggs with a dynamic dialin account with a free or very low-cost high-volume ISP.)
These features together have been instrumental in the current domino effect as more and more countries decide to violate their citizens' privacy. They provide the interceptors with a known fixed point at which to intercept any given person's mail, full knowledge of where his mail is going, full knowledge of the source of incoming mail, and full legibility of the content of correspondance.
If you consider the nature of the people concerned, one might as well have called these mail standards the Please Intercept Me Protocol. We've made it ridiculously easy for them to snoop, so they're doing it. It's our fault. You can't blame them for lack of scruples -- if they had any, they wouldn't have placed themselves in a position where they can wield coercive power over others.
So, let's take our standard catch-all phrase and modify it to suit the new circumstances: The Internet interprets mail interception as a fault and routes around it.
In other words, let's create a mail system with the following attributes:
1. Mail addressing details are not visible except between pairs of delivery points, which see only the from/to information that applies to them at that point in the transport. Multiple layers of encryption keep all other details of addressing invisible. Something along the lines of onion routing seems feasible.
2. Separation of payload from addressing, and the payload is of course always encrypted. Encryption must not be optional, ever, and apart from a strong default, the encryption algorithms used must be arbitrary, multiple, and unidentified.
3. No storage of mailboxes on a customer's ISP's servers. This can be addressed either by using remote servers in safe jurisdictions, or preferably by doing away with the concept of remote mailboxes altogether, ie. keeping mail in transit at various dynamic funnel points until the destinee appears online and signals his presence.
4. No single transit destination for a person's mail. The biggie here is that MX records direct mail to the fixed point of the ISP's choosing, so this whole methodology needs revising to allow the use of a dynamic set of customer-chosen remote funnels instead.
5. Senders should not send mail to their own ISP's relay as smarthost, but bypass it, ie. communicate directly with some remote destination. Unix-type boxes already do direct end-to-end delivery by default anyway, but the new scheme should make that the norm on all platforms.
Well, that doesn't sound like a particularly difficult spec. Let's have a little think about it, rummage around the IETF to see if there's anything already in the works that might do the trick or be a good starting point, and get to it.
Yes, it's desirable, because it'll tie up some of the Microsoft lawyers in fruitless litigation. Every little bit helps.
This won't work though if the project is at the mercy of injunctions. It needs to have a distributed base in several countries where MS hasn't extended strong tendrils into government, plus a floating US presence just to annoy the hell out of them.
And of course it should be faithful to the MS way by embracing and extending MS's own standards.
The problem with both sides of the anime-on-Slashdot argument (a sub-genre of the films-on-Slashdot argument) is that this is not a black and white issue against which an easy value judgement can be made.
It's part of a continuum: films in which technology figures highly, and real life in which the explosive growth of technology is changing all the rules daily. These are not totally separate areas, whatever some hardcore "realists" may think. Today's reality is yesterday's dream, and the day before yesterday it was magic.
The answer as always is to not make any arbitrary decisions, but to give each reader total control over what he or she sees. Well, that shouldn't be a problem -- it's merely technology, not magic, after all.
They've got this all wrong. It wasn't a chance coincidence that the comet erupted when they trained Hubble on it. The act of observing it caused its wave function to collapse, and the cat said "Miaow!".
Sheesh, even scientists are forgetting their physics these days.
Action failed: copyright reg was prerequisite!
on
Napster Ruling Stayed
·
· Score: 5
They didn't need to look hard for reasons to grant a stay: the 9th Circuit judges tore the earlier court's decision to shreds in so many ways that the stay verdict document (pdf) is an absolutely hoot to read. They practically said "You're a load of idiots."
I especially liked this little technicality:
Copyright registration is not a prerequisite to a valid copyright, but it is a prerequisite to a suit based on copyright. [Kodadek v. MTV]
Apparently the plaintiffs had merely identified some 200 songs for which they allegedly claimed copyright without providing proof of copyright registration, and to add to their incompetence, they then tried to extend the claim to millions of songs which they didn't even bother to name, let alone declare under copyright. Apparently this is a legal no-no.
Actually, parts of the legislation may have been weakened by ammendments to the point of being unusable, judging by this excerpt from the discussions.
It seems that because of conflict with a previously enacted law, the new powers will not be allowed to touch content of communications, only traffic/transport-related information/addressing. If so, that was one hell of an ammendment.
On the other hand, the actual details of the law have never stopped those in power from doing what the hell they damn well want, regardless.
It's even worse than you suggest. In presenting such a paper, the author creates the general impression that this problem can be tackled through application of legal measures. That's a totally hilarious proposition.
Talk about trying to hold back the rising tide. How many police departments with mega-brilliant networking specialists are there in the world, and how many lawyers and judges are there with even the remotest comprehension of the issues involved? Sure, there are some... and once those are tied up on major cases, what then?
Estimating the number of active crackers or attackers is difficult, but clearly the number is large even today. With eventual total Internet populations heading into 9 or even 10 figures, the number of cases is going to exceed any possibility of legal enforcement by huge orders of magnitude, if they don't already. In this area, the law is like ants trying to hold back not just the tide but a tidal wave.
Needless to say, the legal profession doesn't like to admit that it is utterly powerless already in any practical sense, let alone that the situation in just a few years time will be massively worse. Academic papers like this (in the worst sense of academic) probably get written by well-intentioned people in the hope that it'll help, but this is a case of massive self-delusion. The numbers just don't add up.
And that's exactly the problem. We anti-spoof aggregated dialpools as I outlined, but when you have hundreds of thousands of addresses in those pools then there is ample opportunity for source address randomization that passes freely through the anti-spoof blocks.
It barely helps at all. Only anti-spoofing individually allocated connections will eliminate this problem entirely, and in a large-ISP environment of multiple dynamic dialpools, that just has to be integrated into the NAS machinery by default. In practice, nothing else is manageable, and nothing else provides the accountability that is needed for abuse control.
There's one thing that large ISPs simply cannot do, and that's to anti-spoof every dialup client connection separately, because that just doesn't scale well as dynamic dialpools grow in size -- that's just too many ACLs for current-day routers.
Consequently, at best they'll anti-spoof aggregated blocks, but even that can be impossible to do when the IP address allocation is fully dynamic as it has to be for really big ISPs, especially those that are virtualized into more than one branded product that must operate in disjoint dialpool spaces but using the same dialin hardware. And even where aggregated anti-spoofing is posssible, it does not provide full accountability --- you can nail the DoS to an ISP, but that's all, which doesn't help when DDoS attacks are synchronized to appear from multiple ISPs simultaneously.
As a result, until the NAS/RAS/HomeGateways etc do their own per-connection anti-spoofing, there will always be opportunity for attackers to randomize their source addresses to some degree. The problem at the moment is that there appears to be virtually no anti-spoofing on dialpools at all in practice, and that's real bad.
I've just had this hilarious vision of psychology researchers across the land in academia probing the nature of the Slashdot community by injecting stimuli like this SETI accelerator announcement and watching how we react. Some vapourware is taken to heart, other vapourware is rejected out of hand -- an excellent subject for study!:-)
I wonder how many doctorates will be awarded for the analysis of such technical community communication and behaviour? More than a few, I bet.
I think I read in their FAQ that it requires a driver for the host, so there is a way in which this could work without the client being ported to their processors. Notice that they don't say that the uploaded linux client actually runs on their processors, which are described as just some sort of vector processing engine, but instead they appear to merely store the client code on the board.
So, here's a possible interpretation: they copy their flash-stored client into a kernel buffer created by their driver. Escape points are poked into the code during the copy in order to occasionally pass control back to their ROM routines for communicating with the vector processors, ie. removing completed work units and loading new ones. So, in effect a dynamic composite client runs in brief bursts in kernel mode, although all the heavy computation is actually in the vector processors and so doesn't take up significant host cycles. And finally, there would have to be a proxy client running in user space to relay communications to and from the in-kernel client.
That would be an odd architecture and very difficult to accomplish (poking a user-space binary to get it to execute in kernel sounds like a nightmare, much more than just ELF loading), but presumably is not impossible. The only part that this doesn't seem to address is client upgrades, since their ROM routines would have to be very client-specific for this to work and so there would be little hope of it working with another client binary.
Although we can only speculate at this stage, it's fun to do. Although it could be a hoax (chips from missiles is wonderfully inventive), it can't be a scam, since they say the item is COD.
Any living organism caught by the shock front from this thing is going to be jelly. A megawatts drive underwater would translate directly into millions or billions of fish dead on each trip, and probably several boats and human lives lost as well for good measure.
It would be nice to think that in the modern UK there will be a similar forward-looking announcement. Unfortunately, the likelihood of that is questionable, because of the multiple conflicts of interest that operate in government here. (We've seen it before: eg. Labour being vociferously against encryption controls before the election, and then mysteriously doing a complete about-face once in government.) Similar problems appear in many other countries of course, but here we have some special long-term legacy baggage which affects broadcasting even more than usual.
In the first place, the good ol' BBC wants to have its cake and eat it too: it strives for editorial independence (allegedly), but it also continues to work through the old boys' network to achieve its goals, especially a continuation of the TV license fee. Well, the demon of payback time always arrives eventually, and so the Beeb finds itself doing what the old boys want, not surprisingly. After decades (generations?) of this, the establishment has acquired an insatiable taste for control of broadcasting. On its own that wouldn't matter too much, but unfortunately there are two other demons lurking in the background.
The second demon takes the form of some extremely powerful global media organizations that operate strongly in the UK, also through the old boys' network but with the extra power of massive financial leverage. Here we're talking about people that make even heads of state shudder, and possibly roll. Well, think about it, if free Internet streaming is perceived as possibly undermining an existing multi-billion dollar media empire founded on controlled broadcasting, are politicians going to legislate freely for the benefit of the people at large? Politicians are not entirely puppets, but once in government they're certainly not free to do the right thing.
And the final demon is the non-independence of Church and State in the UK. As a result, the value judgements of the state religion are never far from the legislators' minds, and censorship is part of the main business for government rather than just a way of appeasing external pressure groups. Well, licensing is the primary means of controlling broadcasters, so the likelihood of Internet broadcasting being given a free pass here seems remote. Fortunately, at the moment streaming broadcasts don't appear to have reached their collective consciousness yet, or maybe they're still frozen in shock like a deer in the headlights, wondering what to do about this horrible flood of freedom.
Be that as it may, the situation is complicated here. One thing that would probably help us in the UK would be for other modern countries to follow the lead of Australia, since politicians don't like to advocate policies that are clearly out of step with the rest of the world. We can but hope.
You write: We have to come up with something better - make spam illegal!
Wow, yes, that's bound to work, just like it has with drugs and speeding.
Sorry to burst your bubble, but laws are powerless when enforcement is powerless, and in an international Internet, passing national anti-spam laws is so utterly pointless that it's just funny.
If huge lists of non-existent addresses are created and then infiltrated into spammers' lists then the proportion of spam that actually gets delivered will go down, the onset of a spam will be easier to detect and to stop, and spammers' efforts will be less effective. Heck, one could even make money from selling them dud data.:-)
Disinformation could be quite an effective measure here. As you say, it has served others well before. All that's required is a complete lack of morality.
If customer data is now deemed to have a direct and independent monetary value rather than merely in association with a product, maybe the coporations that are gathering that data from us ought to be paying for it -- ie. paying us.
After all, everything's a business. We're not a charity either.:-)
It's a whole lot more; it's an attempt to pressure and force Deja to do what "we" want.... You don't like it, fine, complain once,...
Yes, that's what's happening, and no, sorry, it's not going to stop because this is not yesterday's world anymore. The consumer will do what the hell he or she wants, in whatever way they feel like, and not just in the way that some corporate provides for them.
Previously the power resided almost exclusively with big organizations, including government and corporates. Individual empowerment means not only that individuals have more of a say individually, it also means that they can organize and join in the fun of playing the pressure politics game.
Slashdot is one forum where such fun gets expressed and featured a lot, both formally and informally. Trying to limit this new-found freedom and channel it into just those forms that companies like Deja find acceptable is just not going to happen. Sorry.
People are dealing with it: they're providing feedback to Deja on parts of it that they don't like. That's much more constructive than merely ignoring the service as you advocate.
Evidently you don't like direct feedback. Deal with it. The Internet has empowered consumers out of their previous mere passive roles, and the change is here to stay.
Heck,, most TV adverts are more sexually arousing than Tenchi, by a mile. You know, I get the feeling that this can't have anything to do with sex and nudity, as not even US bible-bashing censors could be that brainless. It probably stems from US studio propaganda that Japanese animation is dodgy, and as a result all anime is getting "gently cleansed" with a wire brush and power sander.
Oh well, just ignore the TV and buy the DVDs. Or move to Europe where public nudity is merely "naughty but nice", rather like cream cakes, except that cream cakes are much worse for you.
You've got to be kidding, that would be the most pointless and brainless snip in the history of censorship.
Ryoko is an alien demon (actually, she was manufactured by Washu, the "Greatest Scientist in the Universe", so I suppose that makes her some kind of construct), and she doesn't understand the first thing about human inhibitions about anything, so when she strips off in the pool it's about as innocent as it comes. Cinematographically it's for its effect on Ayeka and Tenchi, and the result is absolutely hilarious.
There is absolutely nothing to censor here. There is no sex (needless to say), nothing remotely provocative, and certainly no sleeze of any kind. There's just some bare (cartoon-style) featureless skin, and that in the context of a demon that is not only non-human but not even really physical in the normal sense (ie. flies right through matter).
Are the censors in the US completely off their collective rockers?
Slashdot Mirror
Shouldn't filtering be set up by kids so that their parents don't suffer the trauma of stumbling across naughty bits on the net?
It's only parents that have trouble with alleged pornography. Under-age kids either don't understand what they're seeing, or they react with "Yuk!" or else they think it's hilarious (and you've got to admit, rubbing squishy bits together has to be the funniest thing on the planet), and older kids positively love it.
And since kids understand the technology far better than their parents anyway and can easily bypass any block they like, it's really they that ought to be doing the censoring to protect their parents.
If trademark X is registered in country Y, then use of the name X in non-Y countries will create a conflict whenever items X from non-Y are used in Y.
Is this concept too difficult for judges to comprehend? Why are they even entertaining such legal action, when name clashes are such an obvious result of national trademarking?
In the rest of the world we've always wondered why the US legal system seemed so utterly screwed up, with everyone suing everyone else and only the lawyers winning in every case, but now we see the process in action, even here.
... but you're just not looking far enough ahead. A worm's eye view won't show you the monster lying in wait just over the hill.
Far from trying to introduce a note of sanity, even the techie community encourages the lawyers in their headless-chicken run with their sue-everyone guns blazing, as in the response to this item. At first glance the replies seem reasonable
Well, you deserve what you get. Just don't complain later about DMCA, RIAA, MPAA, and every other darn piece of lawyer predation.
The law is not a silver bullet. A pact with the devil may be a closer analogy --- eventually you will pay the price.
In the UK, it was the Conservatives that were trying to get these powers in first, against the expressed policies of Labour.
Then Labour came into power, made a total about-face despite their pre-election statements, and brought in laws that are every bit as bad or worse than those which they had opposed.
Needless to say, the Conservatives are currently berating Labour for it, despite their previous stance. But when the Conservatives finally oust Labour again, they'll slam the iron fist down even harder --- guaranteed.
Because you see, it has nothing to do with the nature of politicians' beliefs, and everything to do with the fact that the machine of government (which is much more than just politicians) demands absolute control if it can get away with it quietly.
Well, in this instance it can get away with it, because in our innocence we've created an easily interceptable transport and storage mechanism to underpin the Internet's mail service. No big deal: let's create another mechanism which isn't susceptible to interception.
Encryption on its own is not enough to deal with the issue of mail interception. The interceptors are giving themselves progressively greater powers to ensure that encrypted mail can either be decrypted by them or used as incriminating evidence in its own right by being encrypted in the first place.
The problem is inherent in the current naive SMTP/POP3/IMAP model of mail service. It served us well for two decades, at a time of network openness and innocence. That time has now gone. DARPA researchers wanted to make their network stand the odd atomic bomb. Now we have a different evil to overcome.
Here are the key aspects of the problem, and the areas that a solution would need to address: Time to replace SMTP/POP3/IMAP.
There are five aspects of Internet mail services that have attracted this attention from the interceptors:
1. Mail addressing is in the clear during transport.
2. Mail content is in the clear during transport.
3. Mail storage is in the clear after delivery.
4. Senders send mail to their own ISP's servers.
5. Recipients receive mail from their own ISP's servers.
(This refers to the mail services used by the largest proportion of Internet users, the ordinary Joe Bloggs with a dynamic dialin account with a free or very low-cost high-volume ISP.)
These features together have been instrumental in the current domino effect as more and more countries decide to violate their citizens' privacy. They provide the interceptors with a known fixed point at which to intercept any given person's mail, full knowledge of where his mail is going, full knowledge of the source of incoming mail, and full legibility of the content of correspondance.
If you consider the nature of the people concerned, one might as well have called these mail standards the Please Intercept Me Protocol. We've made it ridiculously easy for them to snoop, so they're doing it. It's our fault. You can't blame them for lack of scruples -- if they had any, they wouldn't have placed themselves in a position where they can wield coercive power over others.
So, let's take our standard catch-all phrase and modify it to suit the new circumstances: The Internet interprets mail interception as a fault and routes around it.
In other words, let's create a mail system with the following attributes:
1. Mail addressing details are not visible except between pairs of delivery points, which see only the from/to information that applies to them at that point in the transport. Multiple layers of encryption keep all other details of addressing invisible. Something along the lines of onion routing seems feasible.
2. Separation of payload from addressing, and the payload is of course always encrypted. Encryption must not be optional, ever, and apart from a strong default, the encryption algorithms used must be arbitrary, multiple, and unidentified.
3. No storage of mailboxes on a customer's ISP's servers. This can be addressed either by using remote servers in safe jurisdictions, or preferably by doing away with the concept of remote mailboxes altogether, ie. keeping mail in transit at various dynamic funnel points until the destinee appears online and signals his presence.
4. No single transit destination for a person's mail. The biggie here is that MX records direct mail to the fixed point of the ISP's choosing, so this whole methodology needs revising to allow the use of a dynamic set of customer-chosen remote funnels instead.
5. Senders should not send mail to their own ISP's relay as smarthost, but bypass it, ie. communicate directly with some remote destination. Unix-type boxes already do direct end-to-end delivery by default anyway, but the new scheme should make that the norm on all platforms.
Well, that doesn't sound like a particularly difficult spec. Let's have a little think about it, rummage around the IETF to see if there's anything already in the works that might do the trick or be a good starting point, and get to it.
Yes, it's desirable, because it'll tie up some of the Microsoft lawyers in fruitless litigation. Every little bit helps.
This won't work though if the project is at the mercy of injunctions. It needs to have a distributed base in several countries where MS hasn't extended strong tendrils into government, plus a floating US presence just to annoy the hell out of them.
And of course it should be faithful to the MS way by embracing and extending MS's own standards.
That should be fun to watch.
The problem with both sides of the anime-on-Slashdot argument (a sub-genre of the films-on-Slashdot argument) is that this is not a black and white issue against which an easy value judgement can be made.
It's part of a continuum: films in which technology figures highly, and real life in which the explosive growth of technology is changing all the rules daily. These are not totally separate areas, whatever some hardcore "realists" may think. Today's reality is yesterday's dream, and the day before yesterday it was magic.
The answer as always is to not make any arbitrary decisions, but to give each reader total control over what he or she sees. Well, that shouldn't be a problem -- it's merely technology, not magic, after all.
They've got this all wrong. It wasn't a chance coincidence that the comet erupted when they trained Hubble on it. The act of observing it caused its wave function to collapse, and the cat said "Miaow!".
Sheesh, even scientists are forgetting their physics these days.
They didn't need to look hard for reasons to grant a stay: the 9th Circuit judges tore the earlier court's decision to shreds in so many ways that the stay verdict document (pdf) is an absolutely hoot to read. They practically said "You're a load of idiots."
...
I especially liked this little technicality:
Copyright registration is not a prerequisite to a valid copyright, but it is a prerequisite to a suit based on copyright. [Kodadek v. MTV]
Apparently the plaintiffs had merely identified some 200 songs for which they allegedly claimed copyright without providing proof of copyright registration, and to add to their incompetence, they then tried to extend the claim to millions of songs which they didn't even bother to name, let alone declare under copyright. Apparently this is a legal no-no.
And lawyers get paid for all this fun. Sigh
Actually, parts of the legislation may have been weakened by ammendments to the point of being unusable, judging by this excerpt from the discussions.
It seems that because of conflict with a previously enacted law, the new powers will not be allowed to touch content of communications, only traffic/transport-related information/addressing. If so, that was one hell of an ammendment.
On the other hand, the actual details of the law have never stopped those in power from doing what the hell they damn well want, regardless.
It's even worse than you suggest. In presenting such a paper, the author creates the general impression that this problem can be tackled through application of legal measures. That's a totally hilarious proposition.
... and once those are tied up on major cases, what then?
Talk about trying to hold back the rising tide. How many police departments with mega-brilliant networking specialists are there in the world, and how many lawyers and judges are there with even the remotest comprehension of the issues involved? Sure, there are some
Estimating the number of active crackers or attackers is difficult, but clearly the number is large even today. With eventual total Internet populations heading into 9 or even 10 figures, the number of cases is going to exceed any possibility of legal enforcement by huge orders of magnitude, if they don't already. In this area, the law is like ants trying to hold back not just the tide but a tidal wave.
Needless to say, the legal profession doesn't like to admit that it is utterly powerless already in any practical sense, let alone that the situation in just a few years time will be massively worse. Academic papers like this (in the worst sense of academic) probably get written by well-intentioned people in the hope that it'll help, but this is a case of massive self-delusion. The numbers just don't add up.
And that's exactly the problem. We anti-spoof aggregated dialpools as I outlined, but when you have hundreds of thousands of addresses in those pools then there is ample opportunity for source address randomization that passes freely through the anti-spoof blocks.
It barely helps at all. Only anti-spoofing individually allocated connections will eliminate this problem entirely, and in a large-ISP environment of multiple dynamic dialpools, that just has to be integrated into the NAS machinery by default. In practice, nothing else is manageable, and nothing else provides the accountability that is needed for abuse control.
There's one thing that large ISPs simply cannot do, and that's to anti-spoof every dialup client connection separately, because that just doesn't scale well as dynamic dialpools grow in size -- that's just too many ACLs for current-day routers.
Consequently, at best they'll anti-spoof aggregated blocks, but even that can be impossible to do when the IP address allocation is fully dynamic as it has to be for really big ISPs, especially those that are virtualized into more than one branded product that must operate in disjoint dialpool spaces but using the same dialin hardware. And even where aggregated anti-spoofing is posssible, it does not provide full accountability --- you can nail the DoS to an ISP, but that's all, which doesn't help when DDoS attacks are synchronized to appear from multiple ISPs simultaneously.
As a result, until the NAS/RAS/HomeGateways etc do their own per-connection anti-spoofing, there will always be opportunity for attackers to randomize their source addresses to some degree. The problem at the moment is that there appears to be virtually no anti-spoofing on dialpools at all in practice, and that's real bad.
I've just had this hilarious vision of psychology researchers across the land in academia probing the nature of the Slashdot community by injecting stimuli like this SETI accelerator announcement and watching how we react. Some vapourware is taken to heart, other vapourware is rejected out of hand -- an excellent subject for study! :-)
I wonder how many doctorates will be awarded for the analysis of such technical community communication and behaviour? More than a few, I bet.
I think I read in their FAQ that it requires a driver for the host, so there is a way in which this could work without the client being ported to their processors. Notice that they don't say that the uploaded linux client actually runs on their processors, which are described as just some sort of vector processing engine, but instead they appear to merely store the client code on the board.
So, here's a possible interpretation: they copy their flash-stored client into a kernel buffer created by their driver. Escape points are poked into the code during the copy in order to occasionally pass control back to their ROM routines for communicating with the vector processors, ie. removing completed work units and loading new ones. So, in effect a dynamic composite client runs in brief bursts in kernel mode, although all the heavy computation is actually in the vector processors and so doesn't take up significant host cycles. And finally, there would have to be a proxy client running in user space to relay communications to and from the in-kernel client.
That would be an odd architecture and very difficult to accomplish (poking a user-space binary to get it to execute in kernel sounds like a nightmare, much more than just ELF loading), but presumably is not impossible. The only part that this doesn't seem to address is client upgrades, since their ROM routines would have to be very client-specific for this to work and so there would be little hope of it working with another client binary.
Although we can only speculate at this stage, it's fun to do. Although it could be a hoax (chips from missiles is wonderfully inventive), it can't be a scam, since they say the item is COD.
Any living organism caught by the shock front from this thing is going to be jelly. A megawatts drive underwater would translate directly into millions or billions of fish dead on each trip, and probably several boats and human lives lost as well for good measure.
Someone's got to be kidding.
It would be nice to think that in the modern UK there will be a similar forward-looking announcement. Unfortunately, the likelihood of that is questionable, because of the multiple conflicts of interest that operate in government here. (We've seen it before: eg. Labour being vociferously against encryption controls before the election, and then mysteriously doing a complete about-face once in government.) Similar problems appear in many other countries of course, but here we have some special long-term legacy baggage which affects broadcasting even more than usual.
In the first place, the good ol' BBC wants to have its cake and eat it too: it strives for editorial independence (allegedly), but it also continues to work through the old boys' network to achieve its goals, especially a continuation of the TV license fee. Well, the demon of payback time always arrives eventually, and so the Beeb finds itself doing what the old boys want, not surprisingly. After decades (generations?) of this, the establishment has acquired an insatiable taste for control of broadcasting. On its own that wouldn't matter too much, but unfortunately there are two other demons lurking in the background.
The second demon takes the form of some extremely powerful global media organizations that operate strongly in the UK, also through the old boys' network but with the extra power of massive financial leverage. Here we're talking about people that make even heads of state shudder, and possibly roll. Well, think about it, if free Internet streaming is perceived as possibly undermining an existing multi-billion dollar media empire founded on controlled broadcasting, are politicians going to legislate freely for the benefit of the people at large? Politicians are not entirely puppets, but once in government they're certainly not free to do the right thing.
And the final demon is the non-independence of Church and State in the UK. As a result, the value judgements of the state religion are never far from the legislators' minds, and censorship is part of the main business for government rather than just a way of appeasing external pressure groups. Well, licensing is the primary means of controlling broadcasters, so the likelihood of Internet broadcasting being given a free pass here seems remote. Fortunately, at the moment streaming broadcasts don't appear to have reached their collective consciousness yet, or maybe they're still frozen in shock like a deer in the headlights, wondering what to do about this horrible flood of freedom.
Be that as it may, the situation is complicated here. One thing that would probably help us in the UK would be for other modern countries to follow the lead of Australia, since politicians don't like to advocate policies that are clearly out of step with the rest of the world. We can but hope.
You write: We have to come up with something better - make spam illegal!
Wow, yes, that's bound to work, just like it has with drugs and speeding.
Sorry to burst your bubble, but laws are powerless when enforcement is powerless, and in an international Internet, passing national anti-spam laws is so utterly pointless that it's just funny.
How about fighting spam with misinformation?
:-)
If huge lists of non-existent addresses are created and then infiltrated into spammers' lists then the proportion of spam that actually gets delivered will go down, the onset of a spam will be easier to detect and to stop, and spammers' efforts will be less effective. Heck, one could even make money from selling them dud data.
Disinformation could be quite an effective measure here. As you say, it has served others well before. All that's required is a complete lack of morality.
If customer data is now deemed to have a direct and independent monetary value rather than merely in association with a product, maybe the coporations that are gathering that data from us ought to be paying for it -- ie. paying us.
:-)
After all, everything's a business. We're not a charity either.
It's a whole lot more; it's an attempt to pressure and force Deja to do what "we" want. ... You don't like it, fine, complain once, ...
Yes, that's what's happening, and no, sorry, it's not going to stop because this is not yesterday's world anymore. The consumer will do what the hell he or she wants, in whatever way they feel like, and not just in the way that some corporate provides for them.
Previously the power resided almost exclusively with big organizations, including government and corporates. Individual empowerment means not only that individuals have more of a say individually, it also means that they can organize and join in the fun of playing the pressure politics game.
Slashdot is one forum where such fun gets expressed and featured a lot, both formally and informally. Trying to limit this new-found freedom and channel it into just those forms that companies like Deja find acceptable is just not going to happen. Sorry.
People are dealing with it: they're providing feedback to Deja on parts of it that they don't like. That's much more constructive than merely ignoring the service as you advocate.
Evidently you don't like direct feedback. Deal with it. The Internet has empowered consumers out of their previous mere passive roles, and the change is here to stay.
Yeah, really hard core. ;-)
Heck,, most TV adverts are more sexually arousing than Tenchi, by a mile. You know, I get the feeling that this can't have anything to do with sex and nudity, as not even US bible-bashing censors could be that brainless. It probably stems from US studio propaganda that Japanese animation is dodgy, and as a result all anime is getting "gently cleansed" with a wire brush and power sander.
Oh well, just ignore the TV and buy the DVDs. Or move to Europe where public nudity is merely "naughty but nice", rather like cream cakes, except that cream cakes are much worse for you.
Huh, the "nudity" in Tenchi Muyo censored?
You've got to be kidding, that would be the most pointless and brainless snip in the history of censorship.
Ryoko is an alien demon (actually, she was manufactured by Washu, the "Greatest Scientist in the Universe", so I suppose that makes her some kind of construct), and she doesn't understand the first thing about human inhibitions about anything, so when she strips off in the pool it's about as innocent as it comes. Cinematographically it's for its effect on Ayeka and Tenchi, and the result is absolutely hilarious.
There is absolutely nothing to censor here. There is no sex (needless to say), nothing remotely provocative, and certainly no sleeze of any kind. There's just some bare (cartoon-style) featureless skin, and that in the context of a demon that is not only non-human but not even really physical in the normal sense (ie. flies right through matter).
Are the censors in the US completely off their collective rockers?