I think that RMS probably knows that the Soviet system was not true Communism as defined by Marx; he's a pretty smart guy. But when somebody throws around accusations of Communism, they're generally doing so to paint the target of those accusations with being like the thing that was called Communism under the Soviet system. That means that disclaiming similarity to the Soviet approach is a reasonable response, since it's a response to the comparison intended.
The alternative of launching into a discussion of how the Soviet Union didn't represent true Communism, and you're happy to accept the mantle of being a Communist as described by Marx, etc. isn't the way to convince people that you're OK. Turning the issue around and painting yourself in Red, White, and Blue as the defenders of peoples' rights and your opponents as evil authoritarians out to deprive those rights makes a much better sound bite. And despite people's complaints about RMS, he's actually getting pretty good at coming up with clever sound bites and slogans. The comparison of software code with recipies is an apt way of making his point.
If software is to be free, then who can we expect to write it. Obviously, I have a need for a paycheck. Since I have this need, I have an employer. In order for my employer to pay me, I have to contribute to their revenue.
You have made the classic mistake (and it's an honest and reasonable one given the dual meaning of free) of software that is available at no charge with software that's free of restrictions. Mr. Stallman has never suggested that it's wrong to charge money for software (to the contrary, in fact), only that it should not have obnoxious restrictions placed on it. RedHat, Mandrake, et. al (even non-proft Debian) charge money for Free Software and it doesn't make it non-free.
And, of course, there are ways of funding free software other than trying to sell it. Linus is being paid partly to hack Linux because his employers think that it will help sell their products (microprocessors). Larry Wall is being paid to hack Perl because his employer thinks that it will help them sell their product (reference books). And now a number of big companies like IBM and Sun are paying developers to write Free Software at least in part because they think that it will help them sell their products (mostly expensive hardware).
The status of the disclamers in any EULA is a legal gray area. There just isn't enough case law to be really confident about just how much would fly in court. It's my understanding that some of the things they're trying to disclaim (like the limitation on incidental and consequental damages, or voiding of implied warrant of merchantability) can't be disclaimed even if the buyer wanted to disclaim them for some reason. Of course finding out in court which things were and weren't legal could be pretty interesting and expensive in legal fees. Part of the goal of UCITA was to change the law so that those ridiculous disclaimers would be legally defined to be correct- software merchants would be allowed to disclaim any liability they felt like. That's why it's such a bad idea.
I don't think that NDAs are generally part of a sales agreement. They're something that's required of people who are given privileged information by a manufacturer as part of an agreement to work with that manufacturer. No sale is involved. AFAIK, though, no reverse-engineering clauses as part of a sales contract have been rejected for physical goods, but their legality for software hasn't been tested.
Yes, I think that you did misunderstand. The "first sale" doctrine says that a seller may not arbitrarily remove your rights to something that you've purchased. The specific case in which it was first mentioned was for a book that claimed to prevent purchasers from re-selling it for below a set price. The Supreme Court ruled that this was not allowable; once the buyer has paid for the book it is his to do with as he pleases. (Note that this refers to the physical book itself, and not the words therein, which are protected under copyright.)
Similarly there are some rights that a seller may not require you to waive as a condition of purchasing their product. In most states, for instance, there is an implied warrant of merchantability (i.e. that selling a product is a promise that it is actually fit for sale) that legally can't be signed away. Any contract or license that purports to do so is just trying to convince users not to try suing, and that portion of the license would be ruled invalid under the law if push came to shove.
IIRC, there is an implied licence with any software that the user has the right to do with it any steps required to use it on his computer, like copying it onto his hard-drive, modifying configuration files that come with the program, etc. AFAIK, though, the right of software companies to further constrain users is a legal gray area without enough case law to make it clear what is and isn't permissible. That's what UCITA is/was about. The software companies are unsure that their licenses would actually stand up in court, so they're trying to change the law to explicitly legalize their favorite licensing terms. This suggests that the current legal status of their licenses is dubious and requires explicit validation.
That's easy enough. Just go to your user page, select Customize Homepage, and click the box next to JonKatz in Exclude Stories from Home Page subcategory Authors. It's the only thing I have checked, and my Slashdot experience has been better ever since.
Scrounging and scavenging equipment is a vital skill for all experimental scientists. It's usually more along the lines of finding the unused goodies that somebody has stashed in the back of their lab than finding the expensive stuff described in the article, but everyone without military-class bugets learns to do it. (Actually, I'll bet that even the best funded darlings do a lot of scrounging, too) Figuring out how to use the components is sometimes a bit of a trick, but there are few things as fun as finding a pile of junk and figuring out how how those components are going to help your next project.
Ever since we follow these steps to prevent the introduction of undesirable code:
[Description that includes using add-on software deleted]
Of course an alternative to this approach is for businesses to behave decently and not try to screw their customers by installing a bunch of unwanted junk. Maybe I'm a naive idealist, but I think that users should be allowed to deal with software under the assumption that the default install will only include components that are significant for the nominal function of the program. Extra packages should be clearly marked as extra and require a separate installation. Any company that doesn't follow this simple principle is behaving obnoxiously and customers have a reasonable right to complain, even if they could have prevented the problem by installing defensively. I understand that this isn' the case, but the rule should not be caveat installor, and companies that behave as though it is should be smacked.
One simple, specific act that would should have been done by Microsoft years ago...
If an attachment is executable, drop it on the floor.
For every complex problem there is an answer that is clear, simple, and wrong. -- H L Mencken
Thanks for proving the truth of your sig. Auto-dropping executable attachments is not a good idea. Your contention that they're incredibly rare is less true than you think. I know someone who wound up with a big hassle just the other day because our email system does drop executable attachments and somebody was trying to send him one that he needed for his work. As an alternative to your suggestion, might I suggest the following:
Not let executables disguise themselves as non-executables. Windows currently lets this happen and even sets it as the default behavior.
Require explicit authorization for a file to be executable, rather than implicit as it is currently. This could be something like Unix's executable bit, but could be as simple as prompting the user for permission the first time he tries to open an executable file.
Allow executables to be sent, but only in a way that requires an extra step to unlock them. You could, for instance, automatically zip all executables but not non-executalbes when sending them by email, and then attach a note telling the user that you've done so. This lets you send executables but gives the user fair warning that they are potentially dangerous.
IMO any of these alternatives would be preferable to what you suggest.
Well, telling people not to click on attachments will help at least somewhat. The viruses propagate by social engineering, so it's important to break the cycle of infection by teaching people not to open the attachments uncritically. That won't do anything to keep current infections from sending out messages, but it will cut down on the next generation of infections and may (ha, ha, ha) prevent the next virus from propagating at all. It's certainly better than just ignoring the problem completely, and there's always the risk of deleting a legitimate attachment accidentally if you scan transmitted email for viruses. Just think about the damage that could be done if some cracker inserted a bogus entry that recognized MS Word headers into the virus definitions for the on-server virus scan used by a major ISP. Then you'd have a really nasty lawsuit on your hands.
An important point, though, is that in some cases we want that power to be available to at least some people. That's the reason that some of the databases were created in the first place; when they're used properly and lawfully they can be very valuable. If the police find evidence that John Doe is going on a killing spree, it's nice if they can call up a list of known favorite hideouts, associates, etc. so that they can catch him faster.
Unfortunately the databases are necessarily accessed by people and not all of them are going to be trustworthy, even in jobs like the police where we'd like to think otherwise. The problem is not so much that there's a database with people's information in it as that the police force is full of the kind of people who are willing to abuse the system (and break the law) to "stalk women, threaten motorists and settle scores." This is, IMO, not so much a technological problem as it is a police corruption issue.
I can't disagree more strongly. This is essentially why RMS came up with the GPL in the first place. His immediate motivation wasn't to find a better way of squashing bugs or higher security. It was to avoid the obnoxious behavior of the companies that were selling proprietary software (though most of them were primarily hardware vendors at that time). He felt that their licensing practices were draconian and treated users as enemies. That's why placed so much emphasis on licensing issues; because licensing issues were what got him started down that path in the first place.
I don't think that anyone is trying to change Perl into Python or Ruby, and I don't think that they're trying to water down any of its current abilities. It sounds to me as though the biggest thing that they're trying to do is to simplify the internals by turning stuff that's currently builtin into modules. One potential big advantage of that is, as Larry Wall points out, that it makes it possible for pragmas to warp the syntax more thoroughly. They could then design a perl6 pragma that would warp the syntax back to the perl5 form, for instance, to make backwards compatibility easier. If somebody really wanted to, they could even make a module that made Perl use more Python-like syntax. It's still about expanding what Perl can do, not about forcing people to do things a particular way; IOW making Perl more Perllike, not less.
Someone once said it:
1) You can't win
2) You can't break even
3) You can't get out of the game
These are the laws of thermodynamics, not of economics. For economics, at least, the first two are 100% dead wrong. The whole thing that makes capitalism work in the first place is that you most certainly can win. If I have something that I think is worth $100 and you think is worth $200, I can sell it to you for $150 and we're both ahead of where we started. I have money that I feel is worth more than the object I started out with, and you have an object that you think is worth more than what you payed for it. That's a win/win situation. In fact, there's no reason for anyone to engage in a voluntary economic activity (i.e. not paying taxes) that doesn't leave them ahead of where they started out, so the fact that people are constantly buying and selling things is strong evidence that they think that they're getting ahead by doing so.
There's an easy solution to this. Just charge more for anything that's sent with the urgent flag. If it's email, for instance, you could charge users a penny for each KB that's sent with the urgent flag. An ordinary person who doesn't abuse the system could send the occasional urgent email and not notice the cost, but spammers who wanted to do so wouldn't be able to afford it. Similarly, you could charge double price for IPv6 packets that have the priortiy flag set; just count the total packets, the packets with priority flag set, and multiply the monthly bill by 1 + priority/total. That, IMO, is what they mean when they talk about applying to basic laws of economics to the system. Let people have their priority service but make it cost them in some way so they don't abuse the system.
The weakness of the idea, though, is that it's open for another type of abuse: forgery. Real spammers aren't going to worry about extra charges for priority email, since they're forging headers and don't expect to be presented with the bill in the first place. Similarly, the Code Red and Sircam worms aren't going to have any compunctions about using high priority flags, either, since the designers aren't the ones who are going to be footing the bill. But if that makes people take security a bit more seriously, so much the better.
No, no, no! If they read about something on a public forum, the original publisher could bring forward the date at which it was submitted as proof of prior art. There's no reason that publishing on a web site would be treated any differently than publishing in a book, provided that you could demonstrate a date of publication. In any case, if they really wanted to patent it, they could apply for the patent before submitting it. Hell, most physicists currently make their articles that are under review available on preprint servers (like http://xxx.lanl.gov and nobody's going around and stealing their ideas. The web was invented by physicists specifically to make it easy for them to make their work available before it was formally published on paper.
This is the key point that so many people are missing. We have very strong evidence that what the biologists are requesting would work becuause the physicists have already tried essentially the same thing and made it work. There's no good reason to think that the result would be any different in biology.
I remember reading many years ago that for many publications, the subscription price is pretty much the cost of the postage to send you the journals, all the other costs are covered by the advertisers (Popular Science, etc). If you have a scientific journal with little to no ads, then the only source of revenue is subscriptions.
That's not really relevant to scientific journals, though. Subscription prices to journals are much, much higher than prices for popular magazines like Popular Science. Subscription prices are routinely over $50 per year for monthly journals and $200 for weekly journals, and 5 to 20 times that for institutional subscriptions. Most journals are scarcely free from advertizing, either, and it's generally well targeted advertizing for expensive items- the kind that's most likely to generate really big revenue. Plus many journals add per-page charges to the authors as an additional source of money. In fact, some journals (PNAS is one I particularly remember) are legally required to print "ADVERTIZEMENT" on each page containing an article because their page charges are high enough that the articles are legally classified as paid advertizements!
It's not as though the publishers are crying for money. Big publishers like Elesvier are very profitable (Elesvier's pre-tax proft margin last year was 25%) and are hardly crying for money. Somehow non-profit publishers manage to put out their journals for substantially less- even when they contract with one of the bigger journals to do the actual physical publishing. There's a reason that the big, for-profit publishers are starting new journals very rapidly; they wouldn't be doing so if they didn't think they'd be profitable.
And that $250 price is for the individual subscription. The institutional price is substantially higher, probably in the range of $1000 per year. And Science is by no means the most expensive journal out there. The Journal of the American Chemical Society is $2296 per year for an institutional subscription, for instance, Biochemistry is $2486, and those were just two that I was able to check quickly. Most of the journals' web sites were interested in discussing on-line institutional access, with prices on the order of $500 per IP per year.
I also work in molecular biology (or biochemistry; the line is a bit fuzzy) and I've had essentially the same issues- except for one publication. I helped to write several units for Protocols in Protein Science and was then completely bowled over when it turned out that in exchange for turning over their copyrights, authors received:
A per-article and per-page payment for their units,
A free copy of the book- which is very useful and normally fairly expensive- for each co-author, and
Free updates as long as the unit remains in the book (in exchange for updating the units as needed)
I thought that was a pretty good deal, even if my employer did make me sign over the check because the writing was done on company time; at least the money went into an account that was under my boss's control rather than into the general pot. OTOH, those chapters were the one piece of writing I've done where the publisher solicited the authors for work rather than the other way around, so the apparently had to offer incentives to get people to agree.
The shameful thing isn't that authors have to sign contracts - it's that in the case of scientific journals the authors aren't being compensated and the works that they essentially donate are being restricted.
In a real sense it's worse than that. Science simply doesn't work if the scientists never exhange their ideas; it's much like journalism in the sense that gathering the data is pointless unless it's disseminated in a timely fashion. This is true not only in the sense that publication is important to the authors careers (which it certainly is) but also in that the research doesn't do anyone any good unless the results are made public.
The publishers have figured out that the authors are eager to publish and they can get them to give up their rights in exchange for getting the work published. And then they can turn around and charge outrageous prices for the work because it's critical for other scientists to be able to read it. As long as most of the journals stick together and insist on those rules, the scientists have no choice; they must publish and they must read, so they have to accept the publishers' terms.
The net (and particularly the web) is putting a big kink in that. The physicists have already banded together and forced the publishers to accept that things have changed. Now it's the biologist's turn. If they can create something as powerful as xxx.lanl.gov, they can get somewhere.
But this assumes that the.kids domain would use the same kind of "no questions asked" registration policy that's in place for.com etc. That's not necessarily the case, though. They could as easily require each registrant to submit a brief description of what was on their site, have a person check the description to make sure that it was "kid safe", and revoke registrations for sites that either violated their description (so that you couldn't register nambla.kids as a Barney site) or the general requirements for all.kids sites (i.e. no nudity, no discussion of sex, etc.) Anyone could lodge a complaint with registrar.kids if they felt that a site was inappropriate. It wouldn't necessarily keep nambla.kids from existing temporarily, but as long as the policy was well spelled out in advance it would give them justification for shutting the site down as soon as it published anything "kid unfriendly".
This, IMO, should be one of the goals of all new TLDs. They should have a clearly stated policy regarding who should be allowed to register, and should block and/or revoke the registrations of sites that violate that policy. Thus they could have policies like one domain per person or organization, must not have the same second name under a different TLD, must not be the trademark holder (for.sucks), etc. If.kids had such a clear policy on what was and wasn't allowed, they could do a pretty good job of ensuring that their domains were consistently appropriate for kids.
My elementary school had a couple TRS-80s...and the wires to connect a cassette player (anyone remember those?).
Hell yes. I also remember the importance of writing about ten copies of anything that you wanted to be able to recover again because the damn tapes were so unreliable. What I really remember, though, is when we first got a floppy drive for the thing and thinking about how the 180 K that you could get on a SS/DD 5 1/4" diskette was so huge. Of course it's still impressive how much you can cram into 180 K today, if anyone actually tried to do so.
And, of course, one big advantage of releasing in a BSD or Public Domain license is that they allow relicensing. If the GPL is really as great as its advocates claim, they could simply relicense it under the GPL and watch all of the people who love the GPL flock to their version and stay away from the BSD/public domain version. It would certainly be an interesting test of some of the claims of GPL advocates.
Slashdot Mirror
I think that RMS probably knows that the Soviet system was not true Communism as defined by Marx; he's a pretty smart guy. But when somebody throws around accusations of Communism, they're generally doing so to paint the target of those accusations with being like the thing that was called Communism under the Soviet system. That means that disclaiming similarity to the Soviet approach is a reasonable response, since it's a response to the comparison intended.
The alternative of launching into a discussion of how the Soviet Union didn't represent true Communism, and you're happy to accept the mantle of being a Communist as described by Marx, etc. isn't the way to convince people that you're OK. Turning the issue around and painting yourself in Red, White, and Blue as the defenders of peoples' rights and your opponents as evil authoritarians out to deprive those rights makes a much better sound bite. And despite people's complaints about RMS, he's actually getting pretty good at coming up with clever sound bites and slogans. The comparison of software code with recipies is an apt way of making his point.
You have made the classic mistake (and it's an honest and reasonable one given the dual meaning of free) of software that is available at no charge with software that's free of restrictions. Mr. Stallman has never suggested that it's wrong to charge money for software (to the contrary, in fact), only that it should not have obnoxious restrictions placed on it. RedHat, Mandrake, et. al (even non-proft Debian) charge money for Free Software and it doesn't make it non-free.
And, of course, there are ways of funding free software other than trying to sell it. Linus is being paid partly to hack Linux because his employers think that it will help sell their products (microprocessors). Larry Wall is being paid to hack Perl because his employer thinks that it will help them sell their product (reference books). And now a number of big companies like IBM and Sun are paying developers to write Free Software at least in part because they think that it will help them sell their products (mostly expensive hardware).
The status of the disclamers in any EULA is a legal gray area. There just isn't enough case law to be really confident about just how much would fly in court. It's my understanding that some of the things they're trying to disclaim (like the limitation on incidental and consequental damages, or voiding of implied warrant of merchantability) can't be disclaimed even if the buyer wanted to disclaim them for some reason. Of course finding out in court which things were and weren't legal could be pretty interesting and expensive in legal fees. Part of the goal of UCITA was to change the law so that those ridiculous disclaimers would be legally defined to be correct- software merchants would be allowed to disclaim any liability they felt like. That's why it's such a bad idea.
I don't think that NDAs are generally part of a sales agreement. They're something that's required of people who are given privileged information by a manufacturer as part of an agreement to work with that manufacturer. No sale is involved. AFAIK, though, no reverse-engineering clauses as part of a sales contract have been rejected for physical goods, but their legality for software hasn't been tested.
Yes, I think that you did misunderstand. The "first sale" doctrine says that a seller may not arbitrarily remove your rights to something that you've purchased. The specific case in which it was first mentioned was for a book that claimed to prevent purchasers from re-selling it for below a set price. The Supreme Court ruled that this was not allowable; once the buyer has paid for the book it is his to do with as he pleases. (Note that this refers to the physical book itself, and not the words therein, which are protected under copyright.)
Similarly there are some rights that a seller may not require you to waive as a condition of purchasing their product. In most states, for instance, there is an implied warrant of merchantability (i.e. that selling a product is a promise that it is actually fit for sale) that legally can't be signed away. Any contract or license that purports to do so is just trying to convince users not to try suing, and that portion of the license would be ruled invalid under the law if push came to shove.
IIRC, there is an implied licence with any software that the user has the right to do with it any steps required to use it on his computer, like copying it onto his hard-drive, modifying configuration files that come with the program, etc. AFAIK, though, the right of software companies to further constrain users is a legal gray area without enough case law to make it clear what is and isn't permissible. That's what UCITA is/was about. The software companies are unsure that their licenses would actually stand up in court, so they're trying to change the law to explicitly legalize their favorite licensing terms. This suggests that the current legal status of their licenses is dubious and requires explicit validation.
That's easy enough. Just go to your user page, select Customize Homepage, and click the box next to JonKatz in Exclude Stories from Home Page subcategory Authors. It's the only thing I have checked, and my Slashdot experience has been better ever since.
Scrounging and scavenging equipment is a vital skill for all experimental scientists. It's usually more along the lines of finding the unused goodies that somebody has stashed in the back of their lab than finding the expensive stuff described in the article, but everyone without military-class bugets learns to do it. (Actually, I'll bet that even the best funded darlings do a lot of scrounging, too) Figuring out how to use the components is sometimes a bit of a trick, but there are few things as fun as finding a pile of junk and figuring out how how those components are going to help your next project.
Of course an alternative to this approach is for businesses to behave decently and not try to screw their customers by installing a bunch of unwanted junk. Maybe I'm a naive idealist, but I think that users should be allowed to deal with software under the assumption that the default install will only include components that are significant for the nominal function of the program. Extra packages should be clearly marked as extra and require a separate installation. Any company that doesn't follow this simple principle is behaving obnoxiously and customers have a reasonable right to complain, even if they could have prevented the problem by installing defensively. I understand that this isn' the case, but the rule should not be caveat installor, and companies that behave as though it is should be smacked.
--
Karma down to 50 again. Thanks Karma Kap.
Thanks for proving the truth of your sig. Auto-dropping executable attachments is not a good idea. Your contention that they're incredibly rare is less true than you think. I know someone who wound up with a big hassle just the other day because our email system does drop executable attachments and somebody was trying to send him one that he needed for his work. As an alternative to your suggestion, might I suggest the following:
IMO any of these alternatives would be preferable to what you suggest.
--
Karma down to 50 again. Thanks Karma Kap.
Well, telling people not to click on attachments will help at least somewhat. The viruses propagate by social engineering, so it's important to break the cycle of infection by teaching people not to open the attachments uncritically. That won't do anything to keep current infections from sending out messages, but it will cut down on the next generation of infections and may (ha, ha, ha) prevent the next virus from propagating at all. It's certainly better than just ignoring the problem completely, and there's always the risk of deleting a legitimate attachment accidentally if you scan transmitted email for viruses. Just think about the damage that could be done if some cracker inserted a bogus entry that recognized MS Word headers into the virus definitions for the on-server virus scan used by a major ISP. Then you'd have a really nasty lawsuit on your hands.
--
Karma down to 50 again. Thanks Karma Kap.
Well, the proceeds from the sale of their assets would be divided among the shareholders. It's about the only plausible way of disposing of the money.
--
Karma down to 50 again. Thanks Karma Kap.
An important point, though, is that in some cases we want that power to be available to at least some people. That's the reason that some of the databases were created in the first place; when they're used properly and lawfully they can be very valuable. If the police find evidence that John Doe is going on a killing spree, it's nice if they can call up a list of known favorite hideouts, associates, etc. so that they can catch him faster.
Unfortunately the databases are necessarily accessed by people and not all of them are going to be trustworthy, even in jobs like the police where we'd like to think otherwise. The problem is not so much that there's a database with people's information in it as that the police force is full of the kind of people who are willing to abuse the system (and break the law) to "stalk women, threaten motorists and settle scores." This is, IMO, not so much a technological problem as it is a police corruption issue.
--
Karma down to 50 again. Thanks Karma Kap.
I can't disagree more strongly. This is essentially why RMS came up with the GPL in the first place. His immediate motivation wasn't to find a better way of squashing bugs or higher security. It was to avoid the obnoxious behavior of the companies that were selling proprietary software (though most of them were primarily hardware vendors at that time). He felt that their licensing practices were draconian and treated users as enemies. That's why placed so much emphasis on licensing issues; because licensing issues were what got him started down that path in the first place.
--
Karma down to 50 again. Thanks Karma Kap.
There's a great and classic quote from Star Wars that summarizes the situation well:
--
Karma down to 50 again. Thanks Karma Kap.
I don't think that anyone is trying to change Perl into Python or Ruby, and I don't think that they're trying to water down any of its current abilities. It sounds to me as though the biggest thing that they're trying to do is to simplify the internals by turning stuff that's currently builtin into modules. One potential big advantage of that is, as Larry Wall points out, that it makes it possible for pragmas to warp the syntax more thoroughly. They could then design a perl6 pragma that would warp the syntax back to the perl5 form, for instance, to make backwards compatibility easier. If somebody really wanted to, they could even make a module that made Perl use more Python-like syntax. It's still about expanding what Perl can do, not about forcing people to do things a particular way; IOW making Perl more Perllike, not less.
Karma below 50 again. Thanks Karma Kap.
These are the laws of thermodynamics, not of economics. For economics, at least, the first two are 100% dead wrong. The whole thing that makes capitalism work in the first place is that you most certainly can win. If I have something that I think is worth $100 and you think is worth $200, I can sell it to you for $150 and we're both ahead of where we started. I have money that I feel is worth more than the object I started out with, and you have an object that you think is worth more than what you payed for it. That's a win/win situation. In fact, there's no reason for anyone to engage in a voluntary economic activity (i.e. not paying taxes) that doesn't leave them ahead of where they started out, so the fact that people are constantly buying and selling things is strong evidence that they think that they're getting ahead by doing so.
Karma below 50 again. Thanks Karma Kap.
There's an easy solution to this. Just charge more for anything that's sent with the urgent flag. If it's email, for instance, you could charge users a penny for each KB that's sent with the urgent flag. An ordinary person who doesn't abuse the system could send the occasional urgent email and not notice the cost, but spammers who wanted to do so wouldn't be able to afford it. Similarly, you could charge double price for IPv6 packets that have the priortiy flag set; just count the total packets, the packets with priority flag set, and multiply the monthly bill by 1 + priority/total. That, IMO, is what they mean when they talk about applying to basic laws of economics to the system. Let people have their priority service but make it cost them in some way so they don't abuse the system.
The weakness of the idea, though, is that it's open for another type of abuse: forgery. Real spammers aren't going to worry about extra charges for priority email, since they're forging headers and don't expect to be presented with the bill in the first place. Similarly, the Code Red and Sircam worms aren't going to have any compunctions about using high priority flags, either, since the designers aren't the ones who are going to be footing the bill. But if that makes people take security a bit more seriously, so much the better.
Karma below 50 again. Thanks Karma Kap.
No, no, no! If they read about something on a public forum, the original publisher could bring forward the date at which it was submitted as proof of prior art. There's no reason that publishing on a web site would be treated any differently than publishing in a book, provided that you could demonstrate a date of publication. In any case, if they really wanted to patent it, they could apply for the patent before submitting it. Hell, most physicists currently make their articles that are under review available on preprint servers (like http://xxx.lanl.gov and nobody's going around and stealing their ideas. The web was invented by physicists specifically to make it easy for them to make their work available before it was formally published on paper.
This is the key point that so many people are missing. We have very strong evidence that what the biologists are requesting would work becuause the physicists have already tried essentially the same thing and made it work. There's no good reason to think that the result would be any different in biology.
Karma below 50 again. Thanks Karma Kap.
That's not really relevant to scientific journals, though. Subscription prices to journals are much, much higher than prices for popular magazines like Popular Science. Subscription prices are routinely over $50 per year for monthly journals and $200 for weekly journals, and 5 to 20 times that for institutional subscriptions. Most journals are scarcely free from advertizing, either, and it's generally well targeted advertizing for expensive items- the kind that's most likely to generate really big revenue. Plus many journals add per-page charges to the authors as an additional source of money. In fact, some journals (PNAS is one I particularly remember) are legally required to print "ADVERTIZEMENT" on each page containing an article because their page charges are high enough that the articles are legally classified as paid advertizements!
It's not as though the publishers are crying for money. Big publishers like Elesvier are very profitable (Elesvier's pre-tax proft margin last year was 25%) and are hardly crying for money. Somehow non-profit publishers manage to put out their journals for substantially less- even when they contract with one of the bigger journals to do the actual physical publishing. There's a reason that the big, for-profit publishers are starting new journals very rapidly; they wouldn't be doing so if they didn't think they'd be profitable.
Karma below 50 again. Thanks Karma Kap.
And that $250 price is for the individual subscription. The institutional price is substantially higher, probably in the range of $1000 per year. And Science is by no means the most expensive journal out there. The Journal of the American Chemical Society is $2296 per year for an institutional subscription, for instance, Biochemistry is $2486, and those were just two that I was able to check quickly. Most of the journals' web sites were interested in discussing on-line institutional access, with prices on the order of $500 per IP per year.
Karma below 50 again. Thanks Karma Kap.
I also work in molecular biology (or biochemistry; the line is a bit fuzzy) and I've had essentially the same issues- except for one publication. I helped to write several units for Protocols in Protein Science and was then completely bowled over when it turned out that in exchange for turning over their copyrights, authors received:
I thought that was a pretty good deal, even if my employer did make me sign over the check because the writing was done on company time; at least the money went into an account that was under my boss's control rather than into the general pot. OTOH, those chapters were the one piece of writing I've done where the publisher solicited the authors for work rather than the other way around, so the apparently had to offer incentives to get people to agree.
Karma below 50 again. Thanks Karma Kap.
In a real sense it's worse than that. Science simply doesn't work if the scientists never exhange their ideas; it's much like journalism in the sense that gathering the data is pointless unless it's disseminated in a timely fashion. This is true not only in the sense that publication is important to the authors careers (which it certainly is) but also in that the research doesn't do anyone any good unless the results are made public.
The publishers have figured out that the authors are eager to publish and they can get them to give up their rights in exchange for getting the work published. And then they can turn around and charge outrageous prices for the work because it's critical for other scientists to be able to read it. As long as most of the journals stick together and insist on those rules, the scientists have no choice; they must publish and they must read, so they have to accept the publishers' terms.
The net (and particularly the web) is putting a big kink in that. The physicists have already banded together and forced the publishers to accept that things have changed. Now it's the biologist's turn. If they can create something as powerful as xxx.lanl.gov, they can get somewhere.
Karma below 50 again. Thanks Karma Kap.
But this assumes that the .kids domain would use the same kind of "no questions asked" registration policy that's in place for .com etc. That's not necessarily the case, though. They could as easily require each registrant to submit a brief description of what was on their site, have a person check the description to make sure that it was "kid safe", and revoke registrations for sites that either violated their description (so that you couldn't register nambla.kids as a Barney site) or the general requirements for all .kids sites (i.e. no nudity, no discussion of sex, etc.) Anyone could lodge a complaint with registrar.kids if they felt that a site was inappropriate. It wouldn't necessarily keep nambla.kids from existing temporarily, but as long as the policy was well spelled out in advance it would give them justification for shutting the site down as soon as it published anything "kid unfriendly".
This, IMO, should be one of the goals of all new TLDs. They should have a clearly stated policy regarding who should be allowed to register, and should block and/or revoke the registrations of sites that violate that policy. Thus they could have policies like one domain per person or organization, must not have the same second name under a different TLD, must not be the trademark holder (for .sucks), etc. If .kids had such a clear policy on what was and wasn't allowed, they could do a pretty good job of ensuring that their domains were consistently appropriate for kids.
Hell yes. I also remember the importance of writing about ten copies of anything that you wanted to be able to recover again because the damn tapes were so unreliable. What I really remember, though, is when we first got a floppy drive for the thing and thinking about how the 180 K that you could get on a SS/DD 5 1/4" diskette was so huge. Of course it's still impressive how much you can cram into 180 K today, if anyone actually tried to do so.
And, of course, one big advantage of releasing in a BSD or Public Domain license is that they allow relicensing. If the GPL is really as great as its advocates claim, they could simply relicense it under the GPL and watch all of the people who love the GPL flock to their version and stay away from the BSD/public domain version. It would certainly be an interesting test of some of the claims of GPL advocates.