And what's more, the EFF would need to conjure up someone worth voting for. As it currently stands, the Democrats control both houses of Congress and are poised to pass a telecom-immunity bill. So voting to maintain Democratic control is not going to do anything there. Voting to switch control to the Republicans, meanwhile, will probably also not improve matters (if anything they'd pass the same telecom immunity bill, only with more enthusiasm).
Without this constant, it looks like the algorithm basically depends on the difficulty of solving the elliptic curve discrete log problem. As with most such problems used in crypto systems (like prime-factoring large numbers), it's believed to be difficult but not proven to be.
In much of mathematics and computer science, authors write their papers in LaTeX, and produce final, print-ready PDFs of their own papers, so there are virtually no production costs---the journal just has to collate these PDFs and stamp page numbers on them, or complain to the author if there's something wrong. If you don't want to do the formatting yourself, you, not the journal, are expected to hire someone to do it. Perhaps in biology and medicine authors still use the old system where they provide the journal with an unformatted manuscript and expect the journal to do the print-ready formatting for them?
The main open-access journals in my field don't require author payment at all. They're run through a combination of volunteer labor, frugality, and institutional or professional-society sponsorship. And they're among the top journals in their areas, maybe even the top journals now: JAIR and JMLR. There are a lot of top-tier open-access, no-publication-fee statistics journals as well. Seems to vary by area.
The NIH doesn't only fund research that's useful to researchers and companies, though it does do that (e.g. new drug research), but also research that's useful to practicing physicians, like studies of the real-world efficacy of treatment options, prevalence and severity of side effects, meta-analyses of the literature, etc., etc. There is a strong governmental interest in having the results of these taxpayer-funded studies actually be available to doctors, since that was the purpose of funding them in the first place. That requires: 1) that someone not keeping up with all the literature can still locate studies relevant to their practice; and 2) that they can actually read the study. The NIH created the PubMed online abstract indexing service to address #1; this bill would address #2.
In physics, it's now standard practice to digitally publish preprints on arxiv.org, which is where everyone gets their new results. It serves as an informal peer-review process, since people peruse and comment on new results as they come out. By the time something hits a journal, most people doing relevant work have already seen the preprint, so the journal is basically just a way of archiving the no-longer-new results. Or at least that's especially true in some parts of physics.
In many areas of computer science, conferences are now the standard publication venue, and journals are again mainly an archival depository, though this varies a bit by sub-area. In some sub-areas journals almost don't even exist, or have much lower prestige than the top conferences. And in any case, authors invariably post PDFs of their papers on their personal websites, and many publishers have given in to this de facto practice and officially allow it (it would be unseemly to sue your own authors).
In the world of computer science, statistics, and related areas, many of us have put our volunteer effort where our ideology is and actually do run top journals in the field, completely for free. Some generous assistance is provided by sponsoring institutions in most cases, which isn't hard to get if you just ask, as many institutions are keen to get their name associated with a journal.
The main journal in my field has a similar model: open access, no fees for either publication or reading. Oh, and authors retain full copyright. I guess these examples are all impossible eh?
While the national party flipped in the 1960s, the segregationists controlled state Democratic parties in many southern states at least through the 1970s. Take a look at Alabama: George Wallace ran and won the 1970s Democratic primary against an incumbent governor on an explicitly racist message, attacking Albert Brewer for reaching out to blacks. Wallace remained governor through 1979, and served as governor again 1983-87, though he claimed to no longer be a racist during his last term. I mean it's still mind-boggling to me that the Democratic Party didn't dump George Fucking Wallace until 1987. And things are worse if you look at the Democratic machines in rural areas---the clean-up there is going far slower than at the state level.
One defendant, in UMG v. Lindor, has actually raised this defense, and the judge accepted that it can be added to the case, rejecting the RIAA's arguments that it was implausible or inapplicable. In particular, the judge wrote in support of the defense's plausibility: "...Lindor cites to case law and to law review articles suggesting that, in a proper case, a court may extend its current due process jurisprudence prohibiting grossly excessive punitive jury awards to prohibit the award of statutory damages mandated under the Copyright Act if they are grossly in excess of the actual damages suffered."
There's speculation among some legal authorities that there might be a constitutional limit on how far statutory damages can depart from actual damages. There was a notable Supreme Court case that put such a limit on punitive damages (BMV v. Gore, 1996), holding that punitive damages can only be a relatively small multiplier over actual damages (the court suggested 10x or more would be highly unlikely to qualify). No case about statutory damages has made it up that high in the 11 years since, so whether court would extend a similar ruling to statutory damages is all speculation. In part this is because statutory damages usually are actually reasonably close to estimated actual damages, so there's no occasion for an appeal on those grounds. Copyright laws are an anomaly in that respect, but also don't come up much because convictions like this one are rare.
Disbarment would have to be used much more frequently than it currently is to be an effective deterrent. As it is, you have to get to Nifong levels of craziness before you have to worry about it.
Say, estimate the number of people likely to have downloaded the file, times the value of the file (about $1 for most songs), and you get something on the order of probably no more than $20 estimated damages. So a 5x multiplier would be $100.
The Supreme Court has held several times that monetary awards must have a reasonable relationship to the actual damages to pass constitutional muster. They look skeptically at anything with more than a single-digit multiplier over actual damages--i.e. 5x actual damages might be ok, but not 10x, much less 1000x, actual damages. This has been used several times to reduce large punitive damage awards as unconstitutional.
To my knowledge several law professors are of the opinion that this line of argument has a decent chance of success in copyright cases, but it hasn't yet been tried.
The full statutory damage limit is clearly unconstitutional, and had the jury awarded $millions in damages it would have been very easy to prove that on appeal. It might still be in this case, but it'll be somewhat more difficult than if they'd gone all out.
It's still a developing area of case law, but some recent Supreme Court precedents have frowned on damages that are in excess of a single-digit multiplier of the actual damage incurred. Since a music file has a value of approximately $1 on most sales sites, that would mean a reasonable damage award would be an estimate of the number of users who actually downloaded the file from her times $10 (or less). I.e. probably something like $100 per song at most.
Since the judicial system moves glacially this argument hasn't yet seen a court, though a few law professors have suggested it.
That's one interpretation, the most sensible one and the one the FSF advances.
However, another interpretation, which also has some precedent, is that the GPL creates a contract, under which the distributor has the right to distribute in return for following its terms. In this case, if the distributor fails to comply with the GPL, it's not a copyright violation, because they did have a copyright license (the GPL), but it's a contract violation, because they breached the GPL's agreed licensing terms. Still illegal, but litigated differently.
I don't think it's a slam-dunk which approach a particular court would take.
OS X with 192mb of RAM is painfully slow if you run any actual applications, and seems to swap nearly constantly. Even if you run Safari instead of Firefox. I find a relatively stripped down Linux desktop to be a lot less painful on that kind of hardware, though maybe you're right that some sort of monstrous GNOME setup with all the background crap enabled would be even worse.
For those curious, here are the current positions of various states:
Either oppose or explicitly prohibit state-level implementation, at least without changes: Arkansas, Colorado, Georgia, Hawaii, Idaho, Illinois, Maine, Missouri, Montana, Nebraska, Nevada, New Hampshire, North Dakota, South Carolina, Tennessee, Utah, and Washington.
Support the act and/or have already begun implementing it: Alabama, California, North Carolina, and North Dakota
It looks like this is a regional rather than partisan difference. In the states that have passed anti-Real-ID acts, the vote totals have been hugely lopsided, sometimes even unanimous. And in the pro-real-ID states, the support and implementation is bipartisan as well. Opposition does seem to be somewhat more common in "red states", despite it being a Republican Congress and President who passed the bill: 12 of the 31 states who voted for GWB in '04 have come out against the act (39%), while only 5 of the 19 who voted for Kerry have (26%).
They are completely separate entities. The only relationship is that Wikipedia's founder, Jimmy Wales, started this venture with Angela Beesley, a former member of the Wikimedia Foundation board of directors.
There is some synergy, partly because of the fact that Jimmy Wales runs Wikia and sort-of runs Wikimedia, partly because Wikia needs community goodwill to succeed, and partly because Wikia uses the MediaWiki software on its own servers so has an interest in it working well. However it isn't anywhere near as close a relationship as you described.
If it were mean income your hypothesis might work, but this is the median income, i.e. the income that 50% of people make more than and 50% of people make less than. That means that the graph is basically showing real increases for the middle class (i.e. those in the middle range of incomes).
The reason US people are in debt way over their heads more than any time before is the same reason that there are more SUVs and HDTVs in the United States than at any time before.
Household income in the United States typically rises faster than inflation. This isn't true every single year, but over any 10-year period since World War II, the median household was earning more, even after adjusting for inflation, than they were at the beginning of that 10-year period. Here is a graph showing the upward trend of inflation-adjusted incomes.
Slashdot Mirror
And what's more, the EFF would need to conjure up someone worth voting for. As it currently stands, the Democrats control both houses of Congress and are poised to pass a telecom-immunity bill. So voting to maintain Democratic control is not going to do anything there. Voting to switch control to the Republicans, meanwhile, will probably also not improve matters (if anything they'd pass the same telecom immunity bill, only with more enthusiasm).
Without this constant, it looks like the algorithm basically depends on the difficulty of solving the elliptic curve discrete log problem. As with most such problems used in crypto systems (like prime-factoring large numbers), it's believed to be difficult but not proven to be.
In much of mathematics and computer science, authors write their papers in LaTeX, and produce final, print-ready PDFs of their own papers, so there are virtually no production costs---the journal just has to collate these PDFs and stamp page numbers on them, or complain to the author if there's something wrong. If you don't want to do the formatting yourself, you, not the journal, are expected to hire someone to do it. Perhaps in biology and medicine authors still use the old system where they provide the journal with an unformatted manuscript and expect the journal to do the print-ready formatting for them?
The main open-access journals in my field don't require author payment at all. They're run through a combination of volunteer labor, frugality, and institutional or professional-society sponsorship. And they're among the top journals in their areas, maybe even the top journals now: JAIR and JMLR. There are a lot of top-tier open-access, no-publication-fee statistics journals as well. Seems to vary by area.
The NIH doesn't only fund research that's useful to researchers and companies, though it does do that (e.g. new drug research), but also research that's useful to practicing physicians, like studies of the real-world efficacy of treatment options, prevalence and severity of side effects, meta-analyses of the literature, etc., etc. There is a strong governmental interest in having the results of these taxpayer-funded studies actually be available to doctors, since that was the purpose of funding them in the first place. That requires: 1) that someone not keeping up with all the literature can still locate studies relevant to their practice; and 2) that they can actually read the study. The NIH created the PubMed online abstract indexing service to address #1; this bill would address #2.
In physics, it's now standard practice to digitally publish preprints on arxiv.org, which is where everyone gets their new results. It serves as an informal peer-review process, since people peruse and comment on new results as they come out. By the time something hits a journal, most people doing relevant work have already seen the preprint, so the journal is basically just a way of archiving the no-longer-new results. Or at least that's especially true in some parts of physics.
In many areas of computer science, conferences are now the standard publication venue, and journals are again mainly an archival depository, though this varies a bit by sub-area. In some sub-areas journals almost don't even exist, or have much lower prestige than the top conferences. And in any case, authors invariably post PDFs of their papers on their personal websites, and many publishers have given in to this de facto practice and officially allow it (it would be unseemly to sue your own authors).
In the world of computer science, statistics, and related areas, many of us have put our volunteer effort where our ideology is and actually do run top journals in the field, completely for free. Some generous assistance is provided by sponsoring institutions in most cases, which isn't hard to get if you just ask, as many institutions are keen to get their name associated with a journal.
Exhibit A
Exhibit B
Exhibit C
etc.
In fact, you can just take a look at this directory and scan for the entries that say "Publication fee: no"---hundreds of them.
The main journal in my field has a similar model: open access, no fees for either publication or reading. Oh, and authors retain full copyright. I guess these examples are all impossible eh?
While the national party flipped in the 1960s, the segregationists controlled state Democratic parties in many southern states at least through the 1970s. Take a look at Alabama: George Wallace ran and won the 1970s Democratic primary against an incumbent governor on an explicitly racist message, attacking Albert Brewer for reaching out to blacks. Wallace remained governor through 1979, and served as governor again 1983-87, though he claimed to no longer be a racist during his last term. I mean it's still mind-boggling to me that the Democratic Party didn't dump George Fucking Wallace until 1987. And things are worse if you look at the Democratic machines in rural areas---the clean-up there is going far slower than at the state level.
One defendant, in UMG v. Lindor, has actually raised this defense, and the judge accepted that it can be added to the case, rejecting the RIAA's arguments that it was implausible or inapplicable. In particular, the judge wrote in support of the defense's plausibility: "...Lindor cites to case law and to law review articles suggesting that, in a proper case, a court may extend its current due process jurisprudence prohibiting grossly excessive punitive jury awards to prohibit the award of statutory damages mandated under the Copyright Act if they are grossly in excess of the actual damages suffered."
There's speculation among some legal authorities that there might be a constitutional limit on how far statutory damages can depart from actual damages. There was a notable Supreme Court case that put such a limit on punitive damages (BMV v. Gore, 1996), holding that punitive damages can only be a relatively small multiplier over actual damages (the court suggested 10x or more would be highly unlikely to qualify). No case about statutory damages has made it up that high in the 11 years since, so whether court would extend a similar ruling to statutory damages is all speculation. In part this is because statutory damages usually are actually reasonably close to estimated actual damages, so there's no occasion for an appeal on those grounds. Copyright laws are an anomaly in that respect, but also don't come up much because convictions like this one are rare.
Disbarment would have to be used much more frequently than it currently is to be an effective deterrent. As it is, you have to get to Nifong levels of craziness before you have to worry about it.
So you're telling me Tokyo, one of the densest cities in the world, is easier to wire for broadband than rural Wyoming? Shocking!
Say, estimate the number of people likely to have downloaded the file, times the value of the file (about $1 for most songs), and you get something on the order of probably no more than $20 estimated damages. So a 5x multiplier would be $100.
The Supreme Court has held several times that monetary awards must have a reasonable relationship to the actual damages to pass constitutional muster. They look skeptically at anything with more than a single-digit multiplier over actual damages--i.e. 5x actual damages might be ok, but not 10x, much less 1000x, actual damages. This has been used several times to reduce large punitive damage awards as unconstitutional.
To my knowledge several law professors are of the opinion that this line of argument has a decent chance of success in copyright cases, but it hasn't yet been tried.
The full statutory damage limit is clearly unconstitutional, and had the jury awarded $millions in damages it would have been very easy to prove that on appeal. It might still be in this case, but it'll be somewhat more difficult than if they'd gone all out.
It's still a developing area of case law, but some recent Supreme Court precedents have frowned on damages that are in excess of a single-digit multiplier of the actual damage incurred. Since a music file has a value of approximately $1 on most sales sites, that would mean a reasonable damage award would be an estimate of the number of users who actually downloaded the file from her times $10 (or less). I.e. probably something like $100 per song at most.
Since the judicial system moves glacially this argument hasn't yet seen a court, though a few law professors have suggested it.
That's one interpretation, the most sensible one and the one the FSF advances.
However, another interpretation, which also has some precedent, is that the GPL creates a contract, under which the distributor has the right to distribute in return for following its terms. In this case, if the distributor fails to comply with the GPL, it's not a copyright violation, because they did have a copyright license (the GPL), but it's a contract violation, because they breached the GPL's agreed licensing terms. Still illegal, but litigated differently.
I don't think it's a slam-dunk which approach a particular court would take.
Copyright law is notoriously unclear on what a "derived work" is. But yes, this is a general copyright-law issue, not a GPL-specific issue.
As far as I can tell, most of Nintendo's resources and marketing are going into their "desktop" equivalent, the Wii, not their portable, the DS.
OS X with 192mb of RAM is painfully slow if you run any actual applications, and seems to swap nearly constantly. Even if you run Safari instead of Firefox. I find a relatively stripped down Linux desktop to be a lot less painful on that kind of hardware, though maybe you're right that some sort of monstrous GNOME setup with all the background crap enabled would be even worse.
At 512mb and above it's not an issue, though.
For those curious, here are the current positions of various states:
Either oppose or explicitly prohibit state-level implementation, at least without changes: Arkansas, Colorado, Georgia, Hawaii, Idaho, Illinois, Maine, Missouri, Montana, Nebraska, Nevada, New Hampshire, North Dakota, South Carolina, Tennessee, Utah, and Washington.
Support the act and/or have already begun implementing it: Alabama, California, North Carolina, and North Dakota
It looks like this is a regional rather than partisan difference. In the states that have passed anti-Real-ID acts, the vote totals have been hugely lopsided, sometimes even unanimous. And in the pro-real-ID states, the support and implementation is bipartisan as well. Opposition does seem to be somewhat more common in "red states", despite it being a Republican Congress and President who passed the bill: 12 of the 31 states who voted for GWB in '04 have come out against the act (39%), while only 5 of the 19 who voted for Kerry have (26%).
The pro-BSD and pro-GPL folks have been sniping at each other for years now.
They are completely separate entities. The only relationship is that Wikipedia's founder, Jimmy Wales, started this venture with Angela Beesley, a former member of the Wikimedia Foundation board of directors.
There is some synergy, partly because of the fact that Jimmy Wales runs Wikia and sort-of runs Wikimedia, partly because Wikia needs community goodwill to succeed, and partly because Wikia uses the MediaWiki software on its own servers so has an interest in it working well. However it isn't anywhere near as close a relationship as you described.
If it were mean income your hypothesis might work, but this is the median income, i.e. the income that 50% of people make more than and 50% of people make less than. That means that the graph is basically showing real increases for the middle class (i.e. those in the middle range of incomes).
The reason US people are in debt way over their heads more than any time before is the same reason that there are more SUVs and HDTVs in the United States than at any time before.
Household income in the United States typically rises faster than inflation. This isn't true every single year, but over any 10-year period since World War II, the median household was earning more, even after adjusting for inflation, than they were at the beginning of that 10-year period. Here is a graph showing the upward trend of inflation-adjusted incomes.