Its interesting, I've heard intel talking about this before (wish I remembered a particular link). Reportedly anyone willing to pay enough could buy a license to sign their software. Along with viral protection they mentioned enhanced DRM... meaning the ability to prevent "circumvention" tools from running.
GP was over the top, but the ignorance implied by your blanket allegation that the post had not a shred of useful thought is surprising. You don't seem to realize the legal significance of a 'per-se' rule. DWI statutes establish a 'per-se' relationship between BAC (blood alcohol concentration) and illegal intoxication. This sort of legal policy should give you pause. Not only does it smack of arbitrariness it also offends what was once a basic principle of law: without a harm to redress there is no standing to protest. Meaning: we should be suspicious of laws which pro-actively punish a potential rather than a realized harm.
The arbitrariness declines depending on how the law is policed. If an officer stops erratic drivers, then tests BAC, we can more reasonably assume that the process is not unduly arbitrary; however, once officers adopt check-point tactics the question of arbitrariness arises.
This went to the US Supreme Court. The Hiibel case law is as follows:
* If the police ask your name you must give it, but you cannot be compelled to give any supporting documentation. * The majority also stated that if someone was convicted of a crime as a consequence of giving their name that the issue could be reconsidered under a Fifth Amendment challenge but that such a challenge did not apply in this particular case.
Sorry, but an anti-circumvention argument is a stretch.
No person shall circumvent a technological measure that effectively controls access to a work protected under this title. The prohibition contained in the preceding sentence shall take effect at the end of the 2-year period beginning on the date of the enactment of this chapter. -- USC Title 17, Section 1201(a)(1)(A)
Now, is a technological measure that can be defeated by merely deleting files or removing registry keys "effective"?
I think not. The real gem is why companies would prefer to use the DMCA: fraud is a civil matter requiring them to pursue the case on their own dime. Conversely, the DMCA allows for criminal prosecution. Whether or not the government is likely to play ball, the threat of such action improves the likelihood of "fear-for-your-life" settlement from the defendant.
Yes, considering the insane amount of dark-fiber between major cities and business districts, I'd guess that the problem is not there. Obviously it takes money to light that fiber. I have to say that technology is being driven very fast right--and its being driven by the likes of Google.
Google is pushing vendors for very fast, high density interconnect. 10Gbps from the server to the mesh. An IEEE study group just green lighted work on a 100Gbps ethernet standard. The target market for this is in metropolitan networks.
An OC-192 fiber connection is worth a mere 622.080 Mbps. Layer-3 switches can operate at roughly 240Gbps.
The noise is all about the business model not about the fundamentals. The backbone providers are becoming something of a commodity service. This would be okay if the tax structure let them provide their service + pay dividends. Instead every company has to be a 'growth company'. Ergo, they have a problem. There is no revenue growth future in what they are doing--unless they can dig their teeth into a new revenue stream--e.g., by raising the rents of content providers.
Did you look at the graph? The error wasn't in anybody's favor. It was negligable.
No, the net bias was about 0.15 C. Note this in the context of your own statement of 0.8 C total warming.
Then consider how ironic it is for you to say, "Don't make the mistake of assuming that a small change in temperature won't have a significant effect."
That's why there's criticism, the US's share of the pollution is a lot larger than its share of land area or population.
Uh-huh. Have you considered that the US grows about one-quarter of the worlds grain supply--most of which is exported? Its kind-of funny how grain product strongly correlates with oil consumption.
There are actually two questions. First whether it is true that the UHI effect is compensated. Second whether Urban stations are actually _trending_ differently from rural stations--rather than merely experiencing an Urban offset. I apologize that this was not made more clear.
Paleoclimatologists that I know are very dubious about Steve's methods, and have convincingly demolished his own statistical techniques (sadly, I can't reconstruct their elegant arguments myself, it was around the time of the NAS study that a couple of them did a presentation for my group on the subject, and my memory isn't so great that far outside my own speciality).
Mann basically applied PC1 analysis with a non-standard (and ultimately unsound) normalization procedure. While in context, I can understand why he did so, it nonetheless distorts the aggregate dataset. I've personally checked Steve's paper on this point (intuitively and numerically): when Mann's algorithm is applied to trendless red-noise, a hockey-stick looking graph results. Proper PCA gives a trendless result under a trendless monte-carlo analysis. This does not demonstrate that there is no hockey-stick, but it does show that Mann's paper ought to have no evidential weight either way because the analytical technique used is not appropriate.
What concerned a lot of people was the circle-the-wagons response of other paleoclimate scientists. In this context what you heard about Steve makes sense. So as regards the PC1 business, I urge you to read Steve's publications and talk to someone who studies statistics and is familiar with the techniques used and draw your own conclusions thereafter.
The classic paper on this is: Jones PD, Groisman PYa, Coughlan M, Plummer N, Wangl WC, Karl TR (1990) Assessment of urbanization effects in time series of surface air temperatures over land. Nature 347:169-172. This conclusion was refreshed by Easterling '97.
The IPCC TAR stated:
These results confirm the conclusions of Jones et al. (1990) and Easterling et al. (1997) that urban effects on 20th century globally and hemispherically averaged land air temperature time-series do not exceed about 0.05C over the period 1900 to 1990 (assumed here to represent one standard error in the assessed non-urban trends).
There have been a couple of recent papers that Steve has been looking at, but as his site is down I don't have the citations handy (and I don't know them off-hand).
You should be careful with realclimate.org. While the site is climate science by climate scientists, it is characterized by evangelism rather than objectivity. This isn't to say their evangelism isn't often scientific and correct, but they do distort, obscure, and ignore information that hurts their evangelism.
As it happens, Steve started his blog climateaudit.org after he was subject to smear campaigns on realclimate.org over a couple of papers he published demolishing the statistical techniques used in MBH'98. Judge for yourself: MM'05,
rc1, rc2, Recap
Steve's papers were ultimately vindicated by a NAS panel review. A copy of which Steve posted on his website: Wegman Report
The guy who found this bug in the GISS data is Steven McIntyre. He's been working for the past few months at auditing studies of the Urban Heat Island (UHI) effect. Several studies have dismissed this effect as non-existent. Steve has been pulling those studies apart--making it more likely that a UHI effect actually exists.
If so, this would tend to bring world-wide temperatures more in-line with US numbers. World-wide temperature records are predominated by urban stations--in areas of substantially growing urbanization in the past 100 years. This urbanization itself taints the temperature trends.
If you look at US cities, their temperature profile matches the global trend.
I think that a sheet of paper with instructions would be sufficient, even for the most computer illiterate students, few though they may be.
I'd be more concerned about computer illiteracy among the teaching staff. e.g., the teacher who emails assignments and supplementary material in a format readable only by the school's software.
Text messaging is cheap in the US compared to Europe. Witness this British article cheering that text messaging will now only be 25p. i.e., about 50c per message.
The iPhone screen is not made of plastic as is the Casio's screen. Consequently it will not draw lots of scratches.
The retail price of the iPhone is consistent with manufacturing costs. That this guy could get a subsidized casio product for 41 dollars is not surprising.
This article plays on a disgusting pathology "all the best stuff is made in japan". No, Japan has been struggling through a crippling recession for ten years. What was true for Marty McFly in the 80s is simply dated now. And here's the kicker: some stuff is actually made for the US market first and then released to Japan. Ditto for Samsung in Singapore. You always release to your home market first to test out the waters.
Except that it tells you what the cost has been in removing most of the constitutional prohibitions against Federal action--the perpetual corruption of the political system because the gains from doing so are so large.
Unfortunately the emphasis on voting systems can be misplaced. The idea of aggregating preferences to pick a unitary course of action is flawed in itself.
Alternatively, or in addition, pay Equifax their extortion money of $130/year for their 3-in-1 monitoring
I find it rather interesting that the credit agencies get away with charging people money to correct a problem which their business practices have ultimately created.
The problem right now is that the credit reporting agencies are exempted by law from any liability for repeating false credit history. What we need is a clear reform of the law such that inaccuracies in the credit report issued about you are subject to civil tort law. The CR agency should be required to collect documentary proof of any previously challenged line of the report. It could then be made libelous to convey challenged entries to creditors unless the CR agency can defend their allegation.
I used to curse hardware vendors for their buggy chips, but now having worked for one such vendor, I see the other end of the stick.
It isn't the number of gates per-se that matters. Most of those transistors are in cache anyways. No, what really kills a project is the amount of explicit (configuration) and implicit state. This makes in nearly impossible to test adequately.
Second, Verilog RTL is used extensively. Verilog RTL lacks type-safety. It is also a very primitive language. A common source of error stems from cycle-misalignment. i.e., data from two sources is expected to be coherent on a given clock-cycle but actually arrives skewed. Worse is when this happens only during certain conditions.
This problems can be hard to detect; they may require very specific sequences of events; they may require sustained load, etc. Verilog provides no means to adequately annotate and assert alignment requirements--they have assertions but I find them nearly useless for documenting temporal relationships.
Modern x86 has everything including the kitchen sink inside. Its pretty difficult for designers to have a good sense of what is happening outside their bubble of responsibility. On the other hand, substantial code reuse takes place from generation to generation. I suspect this run of problems stems from invasive changes being made to enhance SMP performance.
Uh, 300%? I don't think so. Weight is critical in airplane. Components tend to be designed with no more than a 2x margin in mind. To do otherwise would be an egregious waste (think operating fuel costs).
Indeed, its very unusual for a company to be so stringently against having visitors in engineering areas. The last time I encountered such a policy was back in the 80s when AT&T Bell Labs worked on classified projects.
Even at Yahoo/Overture there was only one off-limits room: essentially their version of a NOC and then only because they had big protected displays showing graphs that were correlated with real-time revenue.
I'm sorry but that sort of attitude is not only wrong, it's dangerous--and for the record I've not been convinced the NSA program was ever illegal. I think you're infusing Marbury v. Madison with a meaning that it doesn't have.
All officers of the United States, all legislators, and all judges have a responsibility to judge for themselves the meaning and bounds of the constitution and the law. It is true that simply calling something a name doesn't make it so. But this rule applies to the courts and judges as well as to congress and congressmen.
This is quite different than the requirement for deprive some of life, liberty, or property. There all branches of government must concur.
Slashdot Mirror
Its interesting, I've heard intel talking about this before (wish I remembered a particular link). Reportedly anyone willing to pay enough could buy a license to sign their software. Along with viral protection they mentioned enhanced DRM... meaning the ability to prevent "circumvention" tools from running.
GP was over the top, but the ignorance implied by your blanket allegation that the post had not a shred of useful thought is surprising. You don't seem to realize the legal significance of a 'per-se' rule. DWI statutes establish a 'per-se' relationship between BAC (blood alcohol concentration) and illegal intoxication. This sort of legal policy should give you pause. Not only does it smack of arbitrariness it also offends what was once a basic principle of law: without a harm to redress there is no standing to protest. Meaning: we should be suspicious of laws which pro-actively punish a potential rather than a realized harm.
The arbitrariness declines depending on how the law is policed. If an officer stops erratic drivers, then tests BAC, we can more reasonably assume that the process is not unduly arbitrary; however, once officers adopt check-point tactics the question of arbitrariness arises.
This went to the US Supreme Court. The Hiibel case law is as follows:
* If the police ask your name you must give it, but you cannot be compelled to give any supporting documentation.
* The majority also stated that if someone was convicted of a crime as a consequence of giving their name that the issue could be reconsidered under a Fifth Amendment challenge but that such a challenge did not apply in this particular case.
Bingo. I remember very fondly joining companies and discovering that the IT stuff just works. This is the measure of a good sysadmin team.
The common state-of-affairs is not that everything works; its impressive when it actually does.
Sorry, but an anti-circumvention argument is a stretch.
Now, is a technological measure that can be defeated by merely deleting files or removing registry keys "effective"?
I think not. The real gem is why companies would prefer to use the DMCA: fraud is a civil matter requiring them to pursue the case on their own dime. Conversely, the DMCA allows for criminal prosecution. Whether or not the government is likely to play ball, the threat of such action improves the likelihood of "fear-for-your-life" settlement from the defendant.
Yes. Nasty slip on my part.
Yes, considering the insane amount of dark-fiber between major cities and business districts, I'd guess that the problem is not there. Obviously it takes money to light that fiber. I have to say that technology is being driven very fast right--and its being driven by the likes of Google.
Google is pushing vendors for very fast, high density interconnect. 10Gbps from the server to the mesh. An IEEE study group just green lighted work on a 100Gbps ethernet standard. The target market for this is in metropolitan networks.
An OC-192 fiber connection is worth a mere 622.080 Mbps. Layer-3 switches can operate at roughly 240Gbps.
The noise is all about the business model not about the fundamentals. The backbone providers are becoming something of a commodity service. This would be okay if the tax structure let them provide their service + pay dividends. Instead every company has to be a 'growth company'. Ergo, they have a problem. There is no revenue growth future in what they are doing--unless they can dig their teeth into a new revenue stream--e.g., by raising the rents of content providers.
Mann basically applied PC1 analysis with a non-standard (and ultimately unsound) normalization procedure. While in context, I can understand why he did so, it nonetheless distorts the aggregate dataset. I've personally checked Steve's paper on this point (intuitively and numerically): when Mann's algorithm is applied to trendless red-noise, a hockey-stick looking graph results. Proper PCA gives a trendless result under a trendless monte-carlo analysis. This does not demonstrate that there is no hockey-stick, but it does show that Mann's paper ought to have no evidential weight either way because the analytical technique used is not appropriate.
What concerned a lot of people was the circle-the-wagons response of other paleoclimate scientists. In this context what you heard about Steve makes sense. So as regards the PC1 business, I urge you to read Steve's publications and talk to someone who studies statistics and is familiar with the techniques used and draw your own conclusions thereafter.
The classic paper on this is: Jones PD, Groisman PYa, Coughlan M, Plummer N, Wangl WC, Karl TR (1990) Assessment of urbanization effects in time series of surface air temperatures over land. Nature 347:169-172. This conclusion was refreshed by Easterling '97.
The IPCC TAR stated:
There have been a couple of recent papers that Steve has been looking at, but as his site is down I don't have the citations handy (and I don't know them off-hand).
You should be careful with realclimate.org. While the site is climate science by climate scientists, it is characterized by evangelism rather than objectivity. This isn't to say their evangelism isn't often scientific and correct, but they do distort, obscure, and ignore information that hurts their evangelism.
As it happens, Steve started his blog climateaudit.org after he was subject to smear campaigns on realclimate.org over a couple of papers he published demolishing the statistical techniques used in MBH'98. Judge for yourself: MM'05, rc1, rc2, Recap
Steve's papers were ultimately vindicated by a NAS panel review. A copy of which Steve posted on his website: Wegman Report
.The guy who found this bug in the GISS data is Steven McIntyre. He's been working for the past few months at auditing studies of the Urban Heat Island (UHI) effect. Several studies have dismissed this effect as non-existent. Steve has been pulling those studies apart--making it more likely that a UHI effect actually exists.
If so, this would tend to bring world-wide temperatures more in-line with US numbers. World-wide temperature records are predominated by urban stations--in areas of substantially growing urbanization in the past 100 years. This urbanization itself taints the temperature trends.
If you look at US cities, their temperature profile matches the global trend.
The parent is obviously not a troll.
Yes. The article was about roaming voice charges with one sentence about SMS. I took that to be flat SMS roaming or not, but I could have been wrong.
US SMS costs 10-15 cents, so about 5-7p. Usually $5/mo gets you unlimited SMS.
Except that it tells you what the cost has been in removing most of the constitutional prohibitions against Federal action--the perpetual corruption of the political system because the gains from doing so are so large.
Unfortunately the emphasis on voting systems can be misplaced. The idea of aggregating preferences to pick a unitary course of action is flawed in itself.
I find it rather interesting that the credit agencies get away with charging people money to correct a problem which their business practices have ultimately created.
The problem right now is that the credit reporting agencies are exempted by law from any liability for repeating false credit history. What we need is a clear reform of the law such that inaccuracies in the credit report issued about you are subject to civil tort law. The CR agency should be required to collect documentary proof of any previously challenged line of the report. It could then be made libelous to convey challenged entries to creditors unless the CR agency can defend their allegation.
I used to curse hardware vendors for their buggy chips, but now having worked for one such vendor, I see the other end of the stick.
It isn't the number of gates per-se that matters. Most of those transistors are in cache anyways. No, what really kills a project is the amount of explicit (configuration) and implicit state. This makes in nearly impossible to test adequately.
Second, Verilog RTL is used extensively. Verilog RTL lacks type-safety. It is also a very primitive language. A common source of error stems from cycle-misalignment. i.e., data from two sources is expected to be coherent on a given clock-cycle but actually arrives skewed. Worse is when this happens only during certain conditions.
This problems can be hard to detect; they may require very specific sequences of events; they may require sustained load, etc. Verilog provides no means to adequately annotate and assert alignment requirements--they have assertions but I find them nearly useless for documenting temporal relationships.
Modern x86 has everything including the kitchen sink inside. Its pretty difficult for designers to have a good sense of what is happening outside their bubble of responsibility. On the other hand, substantial code reuse takes place from generation to generation. I suspect this run of problems stems from invasive changes being made to enhance SMP performance.
Uh, 300%? I don't think so. Weight is critical in airplane. Components tend to be designed with no more than a 2x margin in mind. To do otherwise would be an egregious waste (think operating fuel costs).
Indeed, its very unusual for a company to be so stringently against having visitors in engineering areas. The last time I encountered such a policy was back in the 80s when AT&T Bell Labs worked on classified projects.
Even at Yahoo/Overture there was only one off-limits room: essentially their version of a NOC and then only because they had big protected displays showing graphs that were correlated with real-time revenue.
No. It was an attempt at illegal price-fixing by way of copyright.
See Bobbs-Merrill Co. v. Straus, 210 U.S. 339 (1908)
I'm sorry but that sort of attitude is not only wrong, it's dangerous--and for the record I've not been convinced the NSA program was ever illegal. I think you're infusing Marbury v. Madison with a meaning that it doesn't have.
All officers of the United States, all legislators, and all judges have a responsibility to judge for themselves the meaning and bounds of the constitution and the law. It is true that simply calling something a name doesn't make it so. But this rule applies to the courts and judges as well as to congress and congressmen.
This is quite different than the requirement for deprive some of life, liberty, or property. There all branches of government must concur.
Please see: http://www.legislature.mi.gov/(S(blyro0f2tkntfnak0 ueptb55))/mileg.aspx?page=getObject&objectName=mcl -752-797