It wasn't a press release. It was an interview with Bill Buxton, a well-known pioneering computer scientist (SIGCHI Lifetime Achievement Award, Chief Scientist at Alias Wavefront and SGI, pioneered multi-touch interfaces in the '70s, now a principle researcher at Microsoft Research). When the press interviews well-known scientists, it is customary to ask about what new things are coming in the next few years.
Almost all of the vocal critics of Ubunutu I've seen have been trolls, FUDsters, and other worthless people. Has anyone raised serious legitimate criticism of Ubuntu?
He does a great job playing a parody of himself in Free Enterprise. If you haven't seen it, do so. It is a perfect movie for the geek crowd, and Shatner is hilarious.
He can't address a largely made up issue. Grassroots provides hosting and other services for pay to a wide range of organizations. Their parent organization has done some work for Microsoft, and so now to Schestowitz everyone associated with them in any way is secretly working for Microsoft.
What is a little weird about this model is that it ultimately creates a quasi-governmental funding basis for the arts: everyone pays a flat fee that gives them unfettered access to all the world's music (film, etc.) - then, who decides how that money is allocated?
Yes, that's the problem with that solution. It can be shown by the mathematical economists that for goods with certain attributes, a free market is an optimal way to determine how those goods should be produced and distributed. Unfortunately, goods like music lack the necessary attributes, leading to what economists call a "market failure" when you try to use a free market to handle music.
There are two known viable solutions to this kind of market failure. One is to artificially, by force of law, imbue things like music copies with the properties that physical goods have that make them work with a free market. The free market then can determine which music to produce, based on consumer preferences, and funnel the money for that to artists, and all are sort of happy. However, with this solution consumers are paying more than the marginal cost of production for their copies, which is more than they "should" be paying, and that leads to underconsumption. So the tradeoff here is that we get underconsumption, but the good side is that we let the free market deal with resource allocation--no need for government interference other than to provide the legal framework that gives copies the necessary attributes. This is the approach the current copyright system takes in most places.
The other known viable solution is to make copying legal, so that consumers do get their copies for essentially zero, and can consume all the music they wish to consume. Artists are paid by the government, possibly via a special tax on something hopefully at least somewhat correlated with music consumption (an internet tax is a decent approximation for this), or perhaps just out of some general "cultural improvement" budget. The problem with this solution is that someone has to decide which artists get funded. If it is not done right, you could end up with a Bureau of Music that decides what stuffy old safe music gets funded. Better would be to find some way to measure what people are actually choosing to copy and allocate the money accordingly.
Personally, I think it is time to give the second approach a serious test. The first approach worked well when copying was hard enough that most people obeyed the law. Copying is easy now, and far too many people are willing to cheat and get their music for zero relying on the honest people to keep the artists going for the leeches. Time to recognize that use of music is nearly universal and treat it as a public good to be supported by the public.
Here's the total evidence presented at Boycott Novell (oops...Techrights now) for their claims:
1. Consumer Watchdog buys web space from a hosting site that is associated with some lobbying sites.
2. They once had a different name.
Everything else is just random speculation that they provide no evidence whatsoever for.
Considering Consumer Watchdog's long record of involvement in consumer rights issues that have nothing to do with Microsoft, something better than hosting on the same hosting provider as some lobbyists and changing their name once needs to be provided. Just because Boycott Novell changed their name to try to escape their reputation doesn't mean everyone who changes their name is doing it to hide something.
As others have noted, incorrect patent marking stifles innovation.
Letting the public enforce this is efficient. It reminds me of how certain forms of illegal stock trading were discouraged. Certain stockholders are not allowed to engage in something called "short swing trading". If they do, and are caught, they have to give all their profits from the trade to the company. The brilliant way Congress and the SEC came up with to enforce this was to make it so any shareholder can sue on behalf of the corporation. If the shareholder wins (and he always does, because the people who aren't allowed to do these trades are the same set of people that have to report all their trades to the SEC, and so their illegal short swing trades will quickly come to light), the illegal trader has to pay the shareholder's attorney fees. Finally, in the most brilliant part of all, the shareholder only has to be one at the time of filing the suit--not at the time of the illegal trade.
Net result: law firms get the SEC data, run programs to identify short swing traders, go out and buy one share of stock in the company, and sue.
To make it worse, profits are calculated in a way that is very unfavorable to the defendant. Suppose you bought stock at 100/share, later sold that all at 90/share, then later bought the same amount at 80/share, and then sold that at 70/share. You've had a net loss of 20/share, right? That's what you bank account reflects--but that's not how the court calculates it. The court finds the lowest you paid and the highest you sold for and matches them. Repeat until as much is matched as possible. So, the court would just look at that 90/share sale and the later 80/share purchase, and order you to pay 10/share to the company. The remaining 100/share purchase and 70/share sale are ignored. So in addition to losing in reality 20/share on your transactions, and having to pay plaintiff's attorney fees, you also have to pay 10/share to the company!
This has made short swing trading so scary that among those who have to report their trades it virtually stopped shortly after these rules went into effect.
NoSQL never was necessary. Traditional SQL database - not just terascale, but even simple ones like MySQL - regularly deal with data volumes at Google and Walmart that make the sites that built these databases in desperation look positively tiny.
It isn't data volume that is the problem. It is often data organization. Traditional SQL databases are row stores. For some applications that is not a good way to store data. Column stores make more sense in data warehousing, for example. Michael Stonebraker has blogged about this a few times at the same blog site cited by the submitter.
If you did a line by line rewrite, you would likely include copyrightable elements from the original. However, if you just use the original to find out what services the code provides and what computations it is required to perform, and then you implement from that spec, you will likely not have produced a derivative work.
Yes, I've heard of derivative works. That's why I said to just use the original as a spec. If you don't copy any copyrightable elements from the original, you have not produced a derivative work.
I still don't understand why someone didn't just rewrite the code from scratch, using the original as a spec. The original code was released 25 years ago, and is not that huge an amount of code.
A friend from college got into considerable trouble as a kid when he asked that question in Catholic school, although he rephrased it slightly. His form was "Can Jesus make a dildo so big He can't shove it up His ass?" The Nuns showed no interest in discussing the philosophical aspects of that question.
By requiring that every connection/tube be a different color/size, you've now made medical care even more expensive than it already is - hospitals will be forced to stock all manner of spare parts, in sufficient qty for all possible applications - they will no longer be able to stock a huge spool of bulk tubing and cut it down for the application required
All they have to do is have different connectors on the ends of the tubes for different applications. A stock of connectors plus spool of bulk tubing plus a tool to attach the connectors to the tube is all they would need to then cover all their tubing needs. The completely addresses your ridiculous points.
The right way to use DNA is this. First, you find some suspects using traditional methods. When you have a set of suspects, you then do a DNA test against DNA from the crime scene that is known not to be from the victims or witnesses. It likely only matches one of your suspects. That's very likely the criminal.
The wrong way to use DNA is this. First, you find some DNA at the crime scene that doesn't belong to anyone there who is known not to be the criminal (e.g., it doesn't belong to the victims or witnesses). Then you check that DNA against some database of DNA samples. One match is spit out. You arrest that guy and try to convict him solely on the DNA match, telling jurors the odds of a match are 1 in some-large-number.
The problem with the second approach is that they don't test the whole sequence. They just test part of it--not enough to uniquely identify a single person. There may be a dozen people who match the criminal's DNA, depending on the test they use. If only one of those people happens to have done something to get himself into a DNA database that is available to law enforcement, he's screwed if any of the others who match decide to commit a crime.
The first approach (pick the suspects first, then use DNA to single out one) does not suffer from that problem.
I have no idea what the linked article is talking about, because the site is not responding for me. However, in theory, quality of a digital signal cable can in fact affect the quality of subsequent analog output.
Signals are actually analog on the cable, and circuitry reads those signals and classifies them into 0s and 1s. That power consumption of that circuitry is going to change depending on the characteristics of the analog signal that is encoding the digital data. If the digital information is being transmitted using a protocol that includes ECC information, and the poor cable is causing bit errors requiring correction, that too is going to cause variations in power consumption. So you end up with small variations in the power draw of the digital parts of the system. That can cause the power output of the power supply to have small variations. Those could show up as noise in the analog components of the system, such as the audio outputs.
Slashdot Mirror
Why didn't you RTFA?
A stupid press release all round
It wasn't a press release. It was an interview with Bill Buxton, a well-known pioneering computer scientist (SIGCHI Lifetime Achievement Award, Chief Scientist at Alias Wavefront and SGI, pioneered multi-touch interfaces in the '70s, now a principle researcher at Microsoft Research). When the press interviews well-known scientists, it is customary to ask about what new things are coming in the next few years.
Almost all of the vocal critics of Ubunutu I've seen have been trolls, FUDsters, and other worthless people. Has anyone raised serious legitimate criticism of Ubuntu?
A rotating universe around a stationary earth would give the same experimental results under General Relativity.
He's assuming General Relativity, not just Special Relativity.
If ever there was a perfect reason to switch to open source, this is a prime example. Sheesh!
Wouldn't help. The police would simply claim they are checking for GPL violations.
Assuming this is true, so what? Google has tried to get regulator's onto Microsoft's ass. What's wrong with Microsoft returning the favor?
He does a great job playing a parody of himself in Free Enterprise. If you haven't seen it, do so. It is a perfect movie for the geek crowd, and Shatner is hilarious.
He can't address a largely made up issue. Grassroots provides hosting and other services for pay to a wide range of organizations. Their parent organization has done some work for Microsoft, and so now to Schestowitz everyone associated with them in any way is secretly working for Microsoft.
What is a little weird about this model is that it ultimately creates a quasi-governmental funding basis for the arts: everyone pays a flat fee that gives them unfettered access to all the world's music (film, etc.) - then, who decides how that money is allocated?
Yes, that's the problem with that solution. It can be shown by the mathematical economists that for goods with certain attributes, a free market is an optimal way to determine how those goods should be produced and distributed. Unfortunately, goods like music lack the necessary attributes, leading to what economists call a "market failure" when you try to use a free market to handle music.
There are two known viable solutions to this kind of market failure. One is to artificially, by force of law, imbue things like music copies with the properties that physical goods have that make them work with a free market. The free market then can determine which music to produce, based on consumer preferences, and funnel the money for that to artists, and all are sort of happy. However, with this solution consumers are paying more than the marginal cost of production for their copies, which is more than they "should" be paying, and that leads to underconsumption. So the tradeoff here is that we get underconsumption, but the good side is that we let the free market deal with resource allocation--no need for government interference other than to provide the legal framework that gives copies the necessary attributes. This is the approach the current copyright system takes in most places.
The other known viable solution is to make copying legal, so that consumers do get their copies for essentially zero, and can consume all the music they wish to consume. Artists are paid by the government, possibly via a special tax on something hopefully at least somewhat correlated with music consumption (an internet tax is a decent approximation for this), or perhaps just out of some general "cultural improvement" budget. The problem with this solution is that someone has to decide which artists get funded. If it is not done right, you could end up with a Bureau of Music that decides what stuffy old safe music gets funded. Better would be to find some way to measure what people are actually choosing to copy and allocate the money accordingly.
Personally, I think it is time to give the second approach a serious test. The first approach worked well when copying was hard enough that most people obeyed the law. Copying is easy now, and far too many people are willing to cheat and get their music for zero relying on the honest people to keep the artists going for the leeches. Time to recognize that use of music is nearly universal and treat it as a public good to be supported by the public.
How is a tax for music on your broadband connection "free"?
Here's the total evidence presented at Boycott Novell (oops...Techrights now) for their claims:
1. Consumer Watchdog buys web space from a hosting site that is associated with some lobbying sites.
2. They once had a different name.
Everything else is just random speculation that they provide no evidence whatsoever for.
Considering Consumer Watchdog's long record of involvement in consumer rights issues that have nothing to do with Microsoft, something better than hosting on the same hosting provider as some lobbyists and changing their name once needs to be provided. Just because Boycott Novell changed their name to try to escape their reputation doesn't mean everyone who changes their name is doing it to hide something.
That shouldn't matter, because EULA's have no legal weight anyway
Someone better tell the judges, then, because the case law so far rather decisively disagrees with you.
And how, exactly, are you going to know that it has no current patents without looking anything up?
Patent holders are required to mark their wares with their current patents, if they wish to be able to recover damages for infringement.
As others have noted, incorrect patent marking stifles innovation.
Letting the public enforce this is efficient. It reminds me of how certain forms of illegal stock trading were discouraged. Certain stockholders are not allowed to engage in something called "short swing trading". If they do, and are caught, they have to give all their profits from the trade to the company. The brilliant way Congress and the SEC came up with to enforce this was to make it so any shareholder can sue on behalf of the corporation. If the shareholder wins (and he always does, because the people who aren't allowed to do these trades are the same set of people that have to report all their trades to the SEC, and so their illegal short swing trades will quickly come to light), the illegal trader has to pay the shareholder's attorney fees. Finally, in the most brilliant part of all, the shareholder only has to be one at the time of filing the suit--not at the time of the illegal trade.
Net result: law firms get the SEC data, run programs to identify short swing traders, go out and buy one share of stock in the company, and sue.
To make it worse, profits are calculated in a way that is very unfavorable to the defendant. Suppose you bought stock at 100/share, later sold that all at 90/share, then later bought the same amount at 80/share, and then sold that at 70/share. You've had a net loss of 20/share, right? That's what you bank account reflects--but that's not how the court calculates it. The court finds the lowest you paid and the highest you sold for and matches them. Repeat until as much is matched as possible. So, the court would just look at that 90/share sale and the later 80/share purchase, and order you to pay 10/share to the company. The remaining 100/share purchase and 70/share sale are ignored. So in addition to losing in reality 20/share on your transactions, and having to pay plaintiff's attorney fees, you also have to pay 10/share to the company!
This has made short swing trading so scary that among those who have to report their trades it virtually stopped shortly after these rules went into effect.
NoSQL never was necessary. Traditional SQL database - not just terascale, but even simple ones like MySQL - regularly deal with data volumes at Google and Walmart that make the sites that built these databases in desperation look positively tiny.
It isn't data volume that is the problem. It is often data organization. Traditional SQL databases are row stores. For some applications that is not a good way to store data. Column stores make more sense in data warehousing, for example. Michael Stonebraker has blogged about this a few times at the same blog site cited by the submitter.
If you did a line by line rewrite, you would likely include copyrightable elements from the original. However, if you just use the original to find out what services the code provides and what computations it is required to perform, and then you implement from that spec, you will likely not have produced a derivative work.
Probably 90+% of those Slashdot readers who have full-time jobs are invested in those companies and in "big Pharma". Do you have a point?
Yes, I've heard of derivative works. That's why I said to just use the original as a spec. If you don't copy any copyrightable elements from the original, you have not produced a derivative work.
I still don't understand why someone didn't just rewrite the code from scratch, using the original as a spec. The original code was released 25 years ago, and is not that huge an amount of code.
A friend from college got into considerable trouble as a kid when he asked that question in Catholic school, although he rephrased it slightly. His form was "Can Jesus make a dildo so big He can't shove it up His ass?" The Nuns showed no interest in discussing the philosophical aspects of that question.
By requiring that every connection/tube be a different color/size, you've now made medical care even more expensive than it already is - hospitals will be forced to stock all manner of spare parts, in sufficient qty for all possible applications - they will no longer be able to stock a huge spool of bulk tubing and cut it down for the application required
All they have to do is have different connectors on the ends of the tubes for different applications. A stock of connectors plus spool of bulk tubing plus a tool to attach the connectors to the tube is all they would need to then cover all their tubing needs. The completely addresses your ridiculous points.
The right way to use DNA is this. First, you find some suspects using traditional methods. When you have a set of suspects, you then do a DNA test against DNA from the crime scene that is known not to be from the victims or witnesses. It likely only matches one of your suspects. That's very likely the criminal.
The wrong way to use DNA is this. First, you find some DNA at the crime scene that doesn't belong to anyone there who is known not to be the criminal (e.g., it doesn't belong to the victims or witnesses). Then you check that DNA against some database of DNA samples. One match is spit out. You arrest that guy and try to convict him solely on the DNA match, telling jurors the odds of a match are 1 in some-large-number.
The problem with the second approach is that they don't test the whole sequence. They just test part of it--not enough to uniquely identify a single person. There may be a dozen people who match the criminal's DNA, depending on the test they use. If only one of those people happens to have done something to get himself into a DNA database that is available to law enforcement, he's screwed if any of the others who match decide to commit a crime.
The first approach (pick the suspects first, then use DNA to single out one) does not suffer from that problem.
I have no idea what the linked article is talking about, because the site is not responding for me. However, in theory, quality of a digital signal cable can in fact affect the quality of subsequent analog output.
Signals are actually analog on the cable, and circuitry reads those signals and classifies them into 0s and 1s. That power consumption of that circuitry is going to change depending on the characteristics of the analog signal that is encoding the digital data. If the digital information is being transmitted using a protocol that includes ECC information, and the poor cable is causing bit errors requiring correction, that too is going to cause variations in power consumption. So you end up with small variations in the power draw of the digital parts of the system. That can cause the power output of the power supply to have small variations. Those could show up as noise in the analog components of the system, such as the audio outputs.
Is this likely? Probably not. But it is possible.
How about "14 million people effected by floods in Pakistan"...
I want to hear more about this. How did the floods create 14 million people?