That was sort of my point - small programs chained together in simple ways used to produce complex results IS the Unix Way. Object-oriented techniques allow this at the programmer level, making systems like KDE possible, but don't do anything for the user-level. So what we need is a way to support the Unix Way for a graphical environment.
Thinking about this, he seems to be accurate on one point - there hasn't been much UI innovation in the open-source community. (And after all, everyone knows that's all that matters!) There has been a lot of innovation in other areas, mostly places the user doesn't see but which improve the overall experience. Things like operating system internals, file-distribution protocols (BitTorrent), server architecture (look at Apache, and all the stuff they do!), build tools, programming languages, software packaging/installation, software frameworks, compression algorithms, file formats, system administration tools... And that's just off the top of my head.
There's definitely room for improvement. Look at the noises coming out from Microsoft about their next-generation database filesystem. Coders who are interested in filesystems should be looking at that and thinking "how can this be done better?" Or.Net - instead of marching to Microsoft's drum, "we" should be asking "how can we do this better?" And there's always the UI and graphics infrastructure issue...
One problem is that a lot of OSS projects (UI ones, mostly) have moved away from the Unix philosophy: small, simple, dedicated programs that do a job well and can be connected with simple tools to perform complex tasks. Sure, you can feed data from one program into another with modern GUIs, but it typically requires a lot of user intervention and the programs are usually monolithic blobs of functionality. Find a way to escape from that limitation, and develop a graphical equivalent to pipes and I/O redirection, and you'll have some real innovation.
Oh yeah, and there's one little open-source innovation he seems to be forgetting about. Its this minor, inconsequential technology that no-one cares about or uses, called "the Internet".
You know, this "unlimited local calls" thing sounds suspiciously like the "unlimited always-on internet access" bait-and-switch that the broadband companies are pulling now.
This IS assuming that all of the parties aren't colluding based on (say) corporate interests pulling the strings behind the scenes. Even with two parties colluding, they could concievably "force" two wasted rounds with exactly the same sort of tactics the Republicans and Democrats have been using to marginalize third parties for decades. Specifically, the "wasted vote" rhetoric: if you don't like our candidate but vote for one of the other right/left candidates instead of NOTA, you've helped the other guy win! By this, they can "force" a draw until the final round, where they can employ traditional wasted vote rhetoric.
Ranked ballots or elimination rounds seem to minimize this by giving me more than one vote, though combining one with a NOTA system also sounds effective. I can "risk" voting for a third party and still be confident that my vote won't be "helping" the left/right candidate I don't want in office win. I think that a modern electoral system not only has to fairly report the will of the people and prevent fraud, but also blunt propaganda attacks aimed at spreading disinformation about the system itself. Oregon-style voter's informational books sound like a promising way to do this...
I'm not directly comparing against the last election, where massive quantities of fraud seem to have occured on both sides. (Making it a biased sample in the first place) Especially not since the US Federal electoral system is currently inherently biased against third parties - they don't get funding, they're barred from televised debates, etc.
Re:DRM will be *needed* by linux
on
Linus on DRM
·
· Score: 1
Ditto. Having a conversation with someone who is capable of presenting a sane argument on Slashdot has been novel.
I think the problem occurs because we're using DRM to mean different things. You're right, a good encryption algorithm (for example) is effective even if its not closed, because all the secrets are outside the algorithm itself. (The keys, usually) The algorithm's just a way of producing suitable keys and applying them to data. And the algorithms are usually really simple, relying on properties of (usually prime) numbers to do all the heavy lifting. But when I refer to DRM, I'm referring to something like CSS or watermarking, used to prevent someone without a license from accessing digital content in a fashion that the publisher disapproves of.
Hmm... To be accurate, you're right. The algorithms don't have to be secret. However, I still hold that it defeats the purpose of the device for them not to be. Because then the only secret is the key, which the device must somehow be able to access to decode the data. And if the device can access it, I can plop a couple of high-grade probes onto the board and yank the bits out of its nice, secure hardware. And I think that only one person has to do this - once its been done and posted on the 'net, the parrot's out of the bag. Having a secret algorithm adds another step to the reverse-engineering. Not an insurmountable one, but another step all the same. (I no longer know where to look for the key, or how to apply it to the data.)
(I'm assuming, again, DRM like what what the RIAA wants. DRM whose purpose is to technologically prevent you from copying/watching things you haven't paid for. This requires some sort of encryption algorithm, which requires a key for decryption... And which still fails to protect against a bit-for-bit copy, requiring hardware protection of dubious value, such as Palladium.)
Now, if we want to broaden the meaning of DRM, there are other uses for it. Signing programs and content, for example, could be used to verify the source of a movie or program. Assuming a secure method of public key distribution (which is by no means easy), this could work. However, it only garuntees that no changes were made in transit. It doesn't stop me from giving a copy to all one million of my buddies-down-the-street on Kazaa.;) It also doesn't help the 48-hour rental thing. You could have an expiring key, yes, but there's still nothing to prevent me from capturing the key in-transit and applying it for my own nefarious purposes. Or just saving the decrypted data, or any number of other simple attacks.
DRM in general seems to run afoul of the man-in-the-middle attack. (Not that most crypto doesn't - its a very hard puzzle to solve!) But the added problem seems to be that the intended recipient (the viewer) is also the man-in-the-middle.
The 48-hour problem has piqued my curiousity, and I can't really see a solution to it with my (admittedly not universal) knowledge of crypto techniques and other technologies. I'm curious - short of sealed "black box" hardware, what sort of technological solutions do you see to it?
(I specify technological because it is my longstanding belief that illegal copying is a social problem, produced by overpriced content and default treatment of content customers as criminals. And as we all know, there is no way to craft a technological solution to a social problem.)
Actually, capping them at three sounds like a really bad idea. That sounds exploitable - for example, fielding horribly objectionable candidates the first two rounds and then putting out a slightly-less-objectionable but still horrible one (eg, Bush or Gore) for the last, when the people can't reject them. This is the problem with this sort of cyclic voting.
The anti-voting proposed in another post sounds good (instead of voting for candidates, you vote against them). Ranking systems also seem to work well, or at least, give third parties a chance.
Re:DRM will be *needed* by linux
on
Linus on DRM
·
· Score: 1
Right, and you know what? Its looking more and more like music isn't something that anyone has any business selling. The market seems to have been manufactured by technological limitations. Just as most markets are manufactured by resource limitations - what happens to them when technology provides near-limitless resources? They - naturally - die off in favour of technology markets. And its looking like technology will do this to music. Not only make it easy to distribute, but make it easy to make, requiring very little investment.
As for CDs, the transition was slow. I still remember tapes being common in the mid-90s, with most record stores split between tapes and CDs. It was only after burners and CD-ROMs took off that tapes finally died completely.
As for technology and government, those are both invalid examples. Why? The very nature of DRM is a closed, secret system. It has to be. Opening it up and letting people see how it works makes it trivially circumventable, and destroys the monopoly on playback devices it tries to manufacture. (Even keeping just the keys secret isn't enough, because the devices have to have the keys in them somewhere, or access them...) Technology and government, or at least good technology and government, is open and transparent. Democracy, open protocols, open source, etc. You can see what's going on inside, and the process that created it, and if you don't like it, you can change it or replace it. (Or for a government, ideally, move somewhere with one you like.)
Linux is a great example of this. Linus is really only in control because he makes decisions the majority of the community seems to like. (Those that don't are a minority and tend to move to other projects, like the BSDs or HURD) Anyone who doesn't like what's going on can fork the project. If he creates enough dissatisfaction, the community could just move to one of these forks. Of course, this would take a pretty big, unpopular, and hard-to-work-around decision, but it could happen.
Re:DRM will be *needed* by linux
on
Linus on DRM
·
· Score: 1
Here's why DRM won't work: Joe Sixpack is used to being able to record a show off TV and do whatever he wants with the tape. Any device that gives him less freedom than that will be ignored. (Ref: DivX, which many business interests tried to ram down the market's throat) "Business" can't force changes on the market if the market doesn't want the changes - the market will dry up. (And if "business" can force the market to accept a new product, and pay money for it, even if the market does not want that product, then there's something wrong. Very probably criminal conduct.)
And your rental point defeats itself. That's an example of where DRM is useless because its redundant. There's no mass copying of rentals, so why do you need to protect against it? If, OTOH, you want digital transmission and storage of "rentals", then you're going to have to accept that people are going to copy it. There is no technological way around this once your data's entered the user's hands in an easy-to-copy format.
I'm not saying we don't seem to fight DRM, but that we SHOULD fight it, not put on the tight leather outfit, bend over, and resign ourselves to the inevitable. Like prohibition, it can be defeated because there's more money in free copying than in restricted copying. (How? Well, people need devices to play things on. They need bandwidth. They need indexing services and reviews. They make small donations to artists/works they enjoy, if its convenient to. Take a look at TV - viewers don't pay for the content, and they can copy it freely... And networks STILL make money.)
I'm not saying bandwidth caps are a bad thing. Rate limiting is good, as long as we're talking a sane rate - 150-250 kbps, max. Most net connections aren't going to be anywhere near that, even with smart swarming technology like BitTorrent. Reasonable volume limits are also good - 30 gigs/month, for example, is well outside of standard usage, even for people trading video files. (Especially with new codecs, like Xvid)
What's unreasonable is:
Tiny rate (15 kbps or less) or volume (1 gig/month or less) limits. This isn't dialup!
Limits on usage. Local DSL ISPs here (Halifax, NS) have clauses in their TOS banning automated email checkers!
Insane penalties, like being rate-limited forever after exceeding volume quotas or having to pay huge fines. Rate-limit for a fixed, advertised period (say, a month?) and then go back to normal, or make users pay business per-usage rates for excess usage. (Which usually amount to cents per mbps of rate/gigabyte transferred)
I agree that Gnutella/Kazaa/etc protocols suck, which was one of my points - that suckage is why P2P is percieved as being such a bandwidth hog. Hopefully, we'll see more net-friendly, well-designed software like BitTorrent in the near future. Hopefully we'll also see broadband companies with infrastructure that actually reflects the well-documented usage patterns of broadband users!
Yes, that's what I mean. "Bandwidth costs" are often artificial, imposed by poor end-user technology, business decisions ("User usage patterns will be the same as with dialup, so we can assume that they won't ever use more than the capacity of a 56K modem!"), and telco monopolies that see cheap bandwidth as a threat to their "core business". Sure, there's problems when traffic exceeds your infrastructure, which just means that you have to plan for it - we've known for 5 years plus that people like filesharing, so why not account for that when designing your network?
(Answer: Carefully rationed shortages provide an excellent, continuing excuse for a power grab.)
Re:what ?
on
Linus on DRM
·
· Score: 4, Insightful
What he actually seems to be supporting is the stated (note stated, not actual) purpose of Palladium/TCPA. Signing code and verification of signed code to ensure that programs are who/what they say they are. The nature of GPL'd software makes "DRM" impossible - if your GPL'd program does X to verify that its allowed to access a file, I can write a program that does X, accesses the file, and then writes the file in a form that I can access without doing X.
DRM relies on secrets buried in the binary and removing elements of the system from the user's control. (As if the user had control over the verification bits, the system would be useless) The GPL is designed to place all elements of the system in the user's control.
Re:DRM will be *needed* by linux
on
Linus on DRM
·
· Score: 4, Interesting
Sorry, you haven't convinced me DRM is inevitable. Saying Linux must support DRM because DRM is inevitable because in order to support their current business plans, companies will require it doesn't work. Business plans change, attitudes change. Especially in response to customer feedback. You give Joe Average a set-top box that he can record his favorite shows on, but only play them back for 48 hours after recording, and Joe will flip you off and go back to using his VCR. Which doesn't try to tell him how he's allowed to use it.
I'd argue the opposite. That the eventual elimination of DRM is inevitable, as customers refuse technology that employs it, and companies see the technology they purchased for billions from some "security company" defeated in 15 seconds by a grad student. Read Bruce Schnieder sometime - "encryption" and "signing" are not the answer.
Though the problems from Kazaa etc. are often because most modern P2P applications are very wasteful of bandwidth. IIRC, Kazaa does a bunch of "tricks" to try to detect and circumvent attempts to throttle its bandwidth, restrict outgoing connections, etc. The end result is a lot of garbage traffic that is not, strictly speaking, necessary for the protocol.
I believe (though haven't checked personally) that newer open-source P2P software like BitTorrent is more responsible about this.
Also, note that you're drastically overstating the cost of bandwidth to the ISP. Bandwidth is naturally cheap, the only time its really expensive is when a telco monopoly controls it. (As cheap bandwidth is against their percieved best interest) The real problem is often upstream bandwidth - many broadband ISPs seem to have assumed that usage patterns would be unchanged from dialup. The idea that an always-on, high-speed connection might lead to people uploading/hosting more never seems to have occurred to them. (Nor that this is desirable, as it creates value for their other customers at no real cost to them)
Yup. This has been pointed out on Slashdot before, and is just as true now as it was then. Either Microsoft is guilty of terrorism, treason, and espionage, or Jim Allchin is guilty of perjury. Of course, no charges will ever be filed. Hitting a student who made the mistake of putting a few hundred MP3s on an SMB share with a $96 billion lawsuit is a much better use of the Justice Department's time and effort.
In fact, Britain's RIP act is an open invitation to throw anyone who uses a computer in jail for an indefinite period of time. Why? Well, all the act says is that you must turn over encryption keys to your data to the cops upon request, or be tossed in jail without trial. No-where does it require the cops to produce any evidence that the data is encrypted. And since you can get thrown in jail for telling your lawyer that they asked...
Bushcroft still has a lot to learn from Blair about how to exploit a terrorist attack into a civil rights feeding frenzy.
Show me where, on the outside of the box in prominent writing, it tells me that this is a license, not a sale. The text must be clearly visible to me before I hand over my money - any terms they try to add after that are not only unethical, but blatantly illegal.
Then again, we've never had a White House softer on real crime than the one we have now, so it'll be at least another year before anyone will get slapped for this.
I'd disagree about Eon, though that's most likely me getting fed up with cold-war-era sci-fi at about the time I read it. Some interesting ideas, but I think the writing quality was poor. I agree about avoiding Anvil of Stars - avoid, avoid, avoid! Slant is also pretty bad, it reads like an attempt to jump onto the nanotech/biotech bandwagon of the time. Songs of Earth and Power was amazing, but its really fantasy, not sci-fi. Still, its a great story, with some very interesting ideas.
Bear's amazing when he allows himself to be himself. (Songs, Moving Mars) When he tries to jump on a trend (Anvil, Slant), he sucks.
Give Wil McCarthy's The Collapsium a try. It reads very much like Niven and Asimov at their best. You might also like some of David Weber's stand-alone books (Path of the Fury, The Apocalypse Troll). Greg Bear might interest you, though one must be careful - some of his books are great, while others are absolutely terrible. I found Moving Mars to be one of the best, having both interesting ideas and good writing. You might also like Timothy Zahn's Manta's Gift or Angelmass - again, they have a similar feel to Asimov and Niven.
The thing is, "competent businessmen" these days are defined as those that only think about the short term. After all, that's what gets them the big bonuses, and when things start going all wahoonie-shaped, all the decision-makers can bail out with their golden parachutes and leave everyone else to die. Sacrificing short-term profits, even a little, for long-term stability and growth and an even bigger payoff ten years down the road is seen as stupid, unprofitable, irresponsible, and professional suicide.
And what do we have to show for it? A domestic economy that's in shambles, that's what.
Been watching that, but I didn't mention it here. Its more steampunk or fantasy than science fiction so far. There are others that are borderline sci-fi, but they tend to be more "X in space". (Stellvia, for example, feels a lot like Azumanga in space.)
Nope, they're all running 2K. And in the four months I've been here, I've seen about a half-dozen Trillian crashes that have somehow managed to bring the machine to a grinding halt. (Not necessarily kill the operating system, but consume/smash enough other resources to require a reboot)
And IMHO, its found it in Wil McCarthy. If you haven't read The Collapsium, do so now. Reasonably hard science fiction by someone who can actually write? Sign me up!
Seriously, there's a lot of good new science fiction and space opera authors popping up recently. Some good fantasy authors too. Most are still finding their way and struggling to stand out amidst the tide of mindless Tolkien clones and talentless hacks, but they are there. And then there's older space opera/sci-fi authors, like Bujold, that're still doing good work. I think the '90s was the slow decade, and what we're seeing now is a resurgance.
As for TV, give up on the American networks already. They haven't produced anything worthwhile since Next Generation and Babylon-5, and they're not going to anytime soon. Not while they treat the viewer as an inconvenient obstacle between them and their money and use legislation instead of innovation to protect their revenue. Turn to Japan for your sci-fi TV fix, and watch shows like Crest of the Stars (and Banner of the Stars, and the other soon-to-follow sequel series), Legend of the Galactic Heroes, Gundam, Macross, and the new Ghost in the Shell TV series.
(Anime sci-fi shows named off the top of my head. There's a couple dozen other great ones you can find if you look.)
Right, but an open standard makes it the next best thing. Anyone can write a Jabber client or server and take full advantage of the basic protocol. (I believe the protocol allows for extensions which can, of course, be proprietary)
Never had a problem with any of the Jabber clients or servers I've tried, unlike the often flaky and unreliable AIM/ICQ clients and servers my co-workers are always complaining about. (EG, Trillian, which often seems responsible for more work-destroying crashes than Windows ever was.)
When I actually CAN 'follow the jobs' the way people from other countries can, we can talk.
A-men! I say that a primary rule of any globalization policy should be that corporations seeking to move jobs around (by whatever means, be it opening a branch office or hiring from another country) must meet the same inane, frustrating, and often contradictory or impossible requirements that workers seeking to immigrate to that country must meet. So if you think its just for Joe's Multinational to be able to hire programmers from India instead of its native France, then you must also agree that its just for French programmers to move to India and take advantage of the job boom and lower cost of living there, right?
And hey, this also means that India can't toss out their Muslim population, and the middle eastern Arabs have no leg to stand on when calling for the Jews to leave. To say nothing of all the poor Mexicans and Cubans that now have free entry to the US.
Except that the Bush administration is well on its way to banning ALL stem cell research. Including the research on stem cells drawn (or manufactured) from adults. Consenting adults. Consenting, rational adults. Consenting, rational adults unharmed by the procedure. Why? Because their knee-jerk puritanical reaction is biotech == bad.
Slashdot Mirror
That was sort of my point - small programs chained together in simple ways used to produce complex results IS the Unix Way. Object-oriented techniques allow this at the programmer level, making systems like KDE possible, but don't do anything for the user-level. So what we need is a way to support the Unix Way for a graphical environment.
I'd do it myself, but my "UI sense" is horrible.
Thinking about this, he seems to be accurate on one point - there hasn't been much UI innovation in the open-source community. (And after all, everyone knows that's all that matters!) There has been a lot of innovation in other areas, mostly places the user doesn't see but which improve the overall experience. Things like operating system internals, file-distribution protocols (BitTorrent), server architecture (look at Apache, and all the stuff they do!), build tools, programming languages, software packaging/installation, software frameworks, compression algorithms, file formats, system administration tools... And that's just off the top of my head.
There's definitely room for improvement. Look at the noises coming out from Microsoft about their next-generation database filesystem. Coders who are interested in filesystems should be looking at that and thinking "how can this be done better?" Or .Net - instead of marching to Microsoft's drum, "we" should be asking "how can we do this better?" And there's always the UI and graphics infrastructure issue...
One problem is that a lot of OSS projects (UI ones, mostly) have moved away from the Unix philosophy: small, simple, dedicated programs that do a job well and can be connected with simple tools to perform complex tasks. Sure, you can feed data from one program into another with modern GUIs, but it typically requires a lot of user intervention and the programs are usually monolithic blobs of functionality. Find a way to escape from that limitation, and develop a graphical equivalent to pipes and I/O redirection, and you'll have some real innovation.
Oh yeah, and there's one little open-source innovation he seems to be forgetting about. Its this minor, inconsequential technology that no-one cares about or uses, called "the Internet".
You know, this "unlimited local calls" thing sounds suspiciously like the "unlimited always-on internet access" bait-and-switch that the broadband companies are pulling now.
This IS assuming that all of the parties aren't colluding based on (say) corporate interests pulling the strings behind the scenes. Even with two parties colluding, they could concievably "force" two wasted rounds with exactly the same sort of tactics the Republicans and Democrats have been using to marginalize third parties for decades. Specifically, the "wasted vote" rhetoric: if you don't like our candidate but vote for one of the other right/left candidates instead of NOTA, you've helped the other guy win! By this, they can "force" a draw until the final round, where they can employ traditional wasted vote rhetoric.
Ranked ballots or elimination rounds seem to minimize this by giving me more than one vote, though combining one with a NOTA system also sounds effective. I can "risk" voting for a third party and still be confident that my vote won't be "helping" the left/right candidate I don't want in office win. I think that a modern electoral system not only has to fairly report the will of the people and prevent fraud, but also blunt propaganda attacks aimed at spreading disinformation about the system itself. Oregon-style voter's informational books sound like a promising way to do this...
I'm not directly comparing against the last election, where massive quantities of fraud seem to have occured on both sides. (Making it a biased sample in the first place) Especially not since the US Federal electoral system is currently inherently biased against third parties - they don't get funding, they're barred from televised debates, etc.
Ditto. Having a conversation with someone who is capable of presenting a sane argument on Slashdot has been novel.
I think the problem occurs because we're using DRM to mean different things. You're right, a good encryption algorithm (for example) is effective even if its not closed, because all the secrets are outside the algorithm itself. (The keys, usually) The algorithm's just a way of producing suitable keys and applying them to data. And the algorithms are usually really simple, relying on properties of (usually prime) numbers to do all the heavy lifting. But when I refer to DRM, I'm referring to something like CSS or watermarking, used to prevent someone without a license from accessing digital content in a fashion that the publisher disapproves of.
Hmm... To be accurate, you're right. The algorithms don't have to be secret. However, I still hold that it defeats the purpose of the device for them not to be. Because then the only secret is the key, which the device must somehow be able to access to decode the data. And if the device can access it, I can plop a couple of high-grade probes onto the board and yank the bits out of its nice, secure hardware. And I think that only one person has to do this - once its been done and posted on the 'net, the parrot's out of the bag. Having a secret algorithm adds another step to the reverse-engineering. Not an insurmountable one, but another step all the same. (I no longer know where to look for the key, or how to apply it to the data.)
(I'm assuming, again, DRM like what what the RIAA wants. DRM whose purpose is to technologically prevent you from copying/watching things you haven't paid for. This requires some sort of encryption algorithm, which requires a key for decryption... And which still fails to protect against a bit-for-bit copy, requiring hardware protection of dubious value, such as Palladium.)
Now, if we want to broaden the meaning of DRM, there are other uses for it. Signing programs and content, for example, could be used to verify the source of a movie or program. Assuming a secure method of public key distribution (which is by no means easy), this could work. However, it only garuntees that no changes were made in transit. It doesn't stop me from giving a copy to all one million of my buddies-down-the-street on Kazaa. ;) It also doesn't help the 48-hour rental thing. You could have an expiring key, yes, but there's still nothing to prevent me from capturing the key in-transit and applying it for my own nefarious purposes. Or just saving the decrypted data, or any number of other simple attacks.
DRM in general seems to run afoul of the man-in-the-middle attack. (Not that most crypto doesn't - its a very hard puzzle to solve!) But the added problem seems to be that the intended recipient (the viewer) is also the man-in-the-middle.
The 48-hour problem has piqued my curiousity, and I can't really see a solution to it with my (admittedly not universal) knowledge of crypto techniques and other technologies. I'm curious - short of sealed "black box" hardware, what sort of technological solutions do you see to it?
(I specify technological because it is my longstanding belief that illegal copying is a social problem, produced by overpriced content and default treatment of content customers as criminals. And as we all know, there is no way to craft a technological solution to a social problem.)
Actually, capping them at three sounds like a really bad idea. That sounds exploitable - for example, fielding horribly objectionable candidates the first two rounds and then putting out a slightly-less-objectionable but still horrible one (eg, Bush or Gore) for the last, when the people can't reject them. This is the problem with this sort of cyclic voting.
The anti-voting proposed in another post sounds good (instead of voting for candidates, you vote against them). Ranking systems also seem to work well, or at least, give third parties a chance.
Right, and you know what? Its looking more and more like music isn't something that anyone has any business selling. The market seems to have been manufactured by technological limitations. Just as most markets are manufactured by resource limitations - what happens to them when technology provides near-limitless resources? They - naturally - die off in favour of technology markets. And its looking like technology will do this to music. Not only make it easy to distribute, but make it easy to make, requiring very little investment.
As for CDs, the transition was slow. I still remember tapes being common in the mid-90s, with most record stores split between tapes and CDs. It was only after burners and CD-ROMs took off that tapes finally died completely.
As for technology and government, those are both invalid examples. Why? The very nature of DRM is a closed, secret system. It has to be. Opening it up and letting people see how it works makes it trivially circumventable, and destroys the monopoly on playback devices it tries to manufacture. (Even keeping just the keys secret isn't enough, because the devices have to have the keys in them somewhere, or access them...) Technology and government, or at least good technology and government, is open and transparent. Democracy, open protocols, open source, etc. You can see what's going on inside, and the process that created it, and if you don't like it, you can change it or replace it. (Or for a government, ideally, move somewhere with one you like.)
Linux is a great example of this. Linus is really only in control because he makes decisions the majority of the community seems to like. (Those that don't are a minority and tend to move to other projects, like the BSDs or HURD) Anyone who doesn't like what's going on can fork the project. If he creates enough dissatisfaction, the community could just move to one of these forks. Of course, this would take a pretty big, unpopular, and hard-to-work-around decision, but it could happen.
Here's why DRM won't work: Joe Sixpack is used to being able to record a show off TV and do whatever he wants with the tape. Any device that gives him less freedom than that will be ignored. (Ref: DivX, which many business interests tried to ram down the market's throat) "Business" can't force changes on the market if the market doesn't want the changes - the market will dry up. (And if "business" can force the market to accept a new product, and pay money for it, even if the market does not want that product, then there's something wrong. Very probably criminal conduct.)
And your rental point defeats itself. That's an example of where DRM is useless because its redundant. There's no mass copying of rentals, so why do you need to protect against it? If, OTOH, you want digital transmission and storage of "rentals", then you're going to have to accept that people are going to copy it. There is no technological way around this once your data's entered the user's hands in an easy-to-copy format.
I'm not saying we don't seem to fight DRM, but that we SHOULD fight it, not put on the tight leather outfit, bend over, and resign ourselves to the inevitable. Like prohibition, it can be defeated because there's more money in free copying than in restricted copying. (How? Well, people need devices to play things on. They need bandwidth. They need indexing services and reviews. They make small donations to artists/works they enjoy, if its convenient to. Take a look at TV - viewers don't pay for the content, and they can copy it freely... And networks STILL make money.)
I'm not saying bandwidth caps are a bad thing. Rate limiting is good, as long as we're talking a sane rate - 150-250 kbps, max. Most net connections aren't going to be anywhere near that, even with smart swarming technology like BitTorrent. Reasonable volume limits are also good - 30 gigs/month, for example, is well outside of standard usage, even for people trading video files. (Especially with new codecs, like Xvid)
What's unreasonable is:
I agree that Gnutella/Kazaa/etc protocols suck, which was one of my points - that suckage is why P2P is percieved as being such a bandwidth hog. Hopefully, we'll see more net-friendly, well-designed software like BitTorrent in the near future. Hopefully we'll also see broadband companies with infrastructure that actually reflects the well-documented usage patterns of broadband users!
Yes, that's what I mean. "Bandwidth costs" are often artificial, imposed by poor end-user technology, business decisions ("User usage patterns will be the same as with dialup, so we can assume that they won't ever use more than the capacity of a 56K modem!"), and telco monopolies that see cheap bandwidth as a threat to their "core business". Sure, there's problems when traffic exceeds your infrastructure, which just means that you have to plan for it - we've known for 5 years plus that people like filesharing, so why not account for that when designing your network?
(Answer: Carefully rationed shortages provide an excellent, continuing excuse for a power grab.)
What he actually seems to be supporting is the stated (note stated, not actual) purpose of Palladium/TCPA. Signing code and verification of signed code to ensure that programs are who/what they say they are. The nature of GPL'd software makes "DRM" impossible - if your GPL'd program does X to verify that its allowed to access a file, I can write a program that does X, accesses the file, and then writes the file in a form that I can access without doing X.
DRM relies on secrets buried in the binary and removing elements of the system from the user's control. (As if the user had control over the verification bits, the system would be useless) The GPL is designed to place all elements of the system in the user's control.
Sorry, you haven't convinced me DRM is inevitable. Saying Linux must support DRM because DRM is inevitable because in order to support their current business plans, companies will require it doesn't work. Business plans change, attitudes change. Especially in response to customer feedback. You give Joe Average a set-top box that he can record his favorite shows on, but only play them back for 48 hours after recording, and Joe will flip you off and go back to using his VCR. Which doesn't try to tell him how he's allowed to use it.
I'd argue the opposite. That the eventual elimination of DRM is inevitable, as customers refuse technology that employs it, and companies see the technology they purchased for billions from some "security company" defeated in 15 seconds by a grad student. Read Bruce Schnieder sometime - "encryption" and "signing" are not the answer.
Though the problems from Kazaa etc. are often because most modern P2P applications are very wasteful of bandwidth. IIRC, Kazaa does a bunch of "tricks" to try to detect and circumvent attempts to throttle its bandwidth, restrict outgoing connections, etc. The end result is a lot of garbage traffic that is not, strictly speaking, necessary for the protocol.
I believe (though haven't checked personally) that newer open-source P2P software like BitTorrent is more responsible about this.
Also, note that you're drastically overstating the cost of bandwidth to the ISP. Bandwidth is naturally cheap, the only time its really expensive is when a telco monopoly controls it. (As cheap bandwidth is against their percieved best interest) The real problem is often upstream bandwidth - many broadband ISPs seem to have assumed that usage patterns would be unchanged from dialup. The idea that an always-on, high-speed connection might lead to people uploading/hosting more never seems to have occurred to them. (Nor that this is desirable, as it creates value for their other customers at no real cost to them)
Yup. This has been pointed out on Slashdot before, and is just as true now as it was then. Either Microsoft is guilty of terrorism, treason, and espionage, or Jim Allchin is guilty of perjury. Of course, no charges will ever be filed. Hitting a student who made the mistake of putting a few hundred MP3s on an SMB share with a $96 billion lawsuit is a much better use of the Justice Department's time and effort.
In fact, Britain's RIP act is an open invitation to throw anyone who uses a computer in jail for an indefinite period of time. Why? Well, all the act says is that you must turn over encryption keys to your data to the cops upon request, or be tossed in jail without trial. No-where does it require the cops to produce any evidence that the data is encrypted. And since you can get thrown in jail for telling your lawyer that they asked...
Bushcroft still has a lot to learn from Blair about how to exploit a terrorist attack into a civil rights feeding frenzy.
Show me where, on the outside of the box in prominent writing, it tells me that this is a license, not a sale. The text must be clearly visible to me before I hand over my money - any terms they try to add after that are not only unethical, but blatantly illegal.
Then again, we've never had a White House softer on real crime than the one we have now, so it'll be at least another year before anyone will get slapped for this.
I'd disagree about Eon, though that's most likely me getting fed up with cold-war-era sci-fi at about the time I read it. Some interesting ideas, but I think the writing quality was poor. I agree about avoiding Anvil of Stars - avoid, avoid, avoid! Slant is also pretty bad, it reads like an attempt to jump onto the nanotech/biotech bandwagon of the time. Songs of Earth and Power was amazing, but its really fantasy, not sci-fi. Still, its a great story, with some very interesting ideas.
Bear's amazing when he allows himself to be himself. (Songs, Moving Mars) When he tries to jump on a trend (Anvil, Slant), he sucks.
Give Wil McCarthy's The Collapsium a try. It reads very much like Niven and Asimov at their best. You might also like some of David Weber's stand-alone books (Path of the Fury, The Apocalypse Troll). Greg Bear might interest you, though one must be careful - some of his books are great, while others are absolutely terrible. I found Moving Mars to be one of the best, having both interesting ideas and good writing. You might also like Timothy Zahn's Manta's Gift or Angelmass - again, they have a similar feel to Asimov and Niven.
The thing is, "competent businessmen" these days are defined as those that only think about the short term. After all, that's what gets them the big bonuses, and when things start going all wahoonie-shaped, all the decision-makers can bail out with their golden parachutes and leave everyone else to die. Sacrificing short-term profits, even a little, for long-term stability and growth and an even bigger payoff ten years down the road is seen as stupid, unprofitable, irresponsible, and professional suicide.
And what do we have to show for it? A domestic economy that's in shambles, that's what.
Been watching that, but I didn't mention it here. Its more steampunk or fantasy than science fiction so far. There are others that are borderline sci-fi, but they tend to be more "X in space". (Stellvia, for example, feels a lot like Azumanga in space.)
Nope, they're all running 2K. And in the four months I've been here, I've seen about a half-dozen Trillian crashes that have somehow managed to bring the machine to a grinding halt. (Not necessarily kill the operating system, but consume/smash enough other resources to require a reboot)
And IMHO, its found it in Wil McCarthy. If you haven't read The Collapsium, do so now. Reasonably hard science fiction by someone who can actually write? Sign me up!
Seriously, there's a lot of good new science fiction and space opera authors popping up recently. Some good fantasy authors too. Most are still finding their way and struggling to stand out amidst the tide of mindless Tolkien clones and talentless hacks, but they are there. And then there's older space opera/sci-fi authors, like Bujold, that're still doing good work. I think the '90s was the slow decade, and what we're seeing now is a resurgance.
As for TV, give up on the American networks already. They haven't produced anything worthwhile since Next Generation and Babylon-5, and they're not going to anytime soon. Not while they treat the viewer as an inconvenient obstacle between them and their money and use legislation instead of innovation to protect their revenue. Turn to Japan for your sci-fi TV fix, and watch shows like Crest of the Stars (and Banner of the Stars, and the other soon-to-follow sequel series), Legend of the Galactic Heroes, Gundam, Macross, and the new Ghost in the Shell TV series.
(Anime sci-fi shows named off the top of my head. There's a couple dozen other great ones you can find if you look.)
Right, but an open standard makes it the next best thing. Anyone can write a Jabber client or server and take full advantage of the basic protocol. (I believe the protocol allows for extensions which can, of course, be proprietary)
Never had a problem with any of the Jabber clients or servers I've tried, unlike the often flaky and unreliable AIM/ICQ clients and servers my co-workers are always complaining about. (EG, Trillian, which often seems responsible for more work-destroying crashes than Windows ever was.)
When I actually CAN 'follow the jobs' the way people from other countries can, we can talk.
A-men! I say that a primary rule of any globalization policy should be that corporations seeking to move jobs around (by whatever means, be it opening a branch office or hiring from another country) must meet the same inane, frustrating, and often contradictory or impossible requirements that workers seeking to immigrate to that country must meet. So if you think its just for Joe's Multinational to be able to hire programmers from India instead of its native France, then you must also agree that its just for French programmers to move to India and take advantage of the job boom and lower cost of living there, right?
And hey, this also means that India can't toss out their Muslim population, and the middle eastern Arabs have no leg to stand on when calling for the Jews to leave. To say nothing of all the poor Mexicans and Cubans that now have free entry to the US.
Isn't globalization wonderful?
Except that the Bush administration is well on its way to banning ALL stem cell research. Including the research on stem cells drawn (or manufactured) from adults. Consenting adults. Consenting, rational adults. Consenting, rational adults unharmed by the procedure. Why? Because their knee-jerk puritanical reaction is biotech == bad.