First, where else would they go? Nike has to make shoes, it is their job. If they don't do that, they go out of business and default on all their sponsorship contracts. If people won't buy shoes made by workers earning a living wage, they are not a viable product and don't deserve to be produced. But I don't think that will happen. Like I said, Nike might have to raise prices, or pay sports figures less for endorsments, but they will still stay in business.
What would be bad for them is not all shoe companies had to operate on an even footing--if they were held to higher standards than other companies--that is why we need laws and governments for this sort of thing, rather than relying on the goodwill of individual companies.
Too many people don't pay attention in economics class, and all they get out is "humans are rational, self interested parties." This is a questionable hypothesis to begin with, largely misunderstood, and ignores the more important thesis "market forces drive a free market economy". By manipulating those market forces, it is possible (within limits) to control a "free market" economy. Depending on the means and the ends, such manipulations can be monopolistic behavior or collusion (bad), other times it is optimization, and good.
Neither are particularly related to Soviet socailims where the market was driven by government quotas, not supply and demand.
Very often, that is true, though I would say you underestimate the output of the "million monkeys". Go to freshmeat and just browse. Sure, much of it isn't earth shattering, or "pure research", but there are lots and lots of programs that are valuable in their own right. Other programs for Windows or UNIX may have similar functionality, but each one was designed to perform in a particular way as well as possible. In the words of ESR, they all started to scratch an itch.
A few notable examples:
licq -- implements similar functionality to windows ICQ, but IMO has the best UI of any instant messanger I have seen. Also, recently added support for SSL encrypted communications, and has a wide array of plugins (nmap, finger, etc) that windows ICQ and other ICQ programs lack.
lame: conforms to the MPEG audio encoder standard. Similar to several other encoders around, but has (arguably) the best sound quality of any available encoder.
Mame: free arcade emulator that runs on an incredibly number of platforms, including laser light show controllers and digital camers. These two are good examples of features that have no commercial value, and thus would never be implemented by a commercial developer, yet are way cool, and some people might use.
Apache: originally based on the NCSA httpd, Apache is now the most flexible web server in existence.
apt-get: IMNSHO, debian has nearly perfected the process of software distribution, installation, and version management. "apt-get install mozilla" makes inserting a cd and clicking OK from the autorun dialog look horribly complicated. And no commercial software company will ever match it, because they can't keep track of licensing then. Maybe MS will come up with something 10% as good with.NET and "subscription software", where you can install off the net and get billed a subscription, but it isn't worth the cost of giving up control over your computer to get.
Linux Kernel: The linux kernel isn't just another POSIX implementation. It has an unheard of combination of features for desktop, workstation, server, and embedded system all in one package. It runs more or less the same on many, many platforms, and has features like loopback block devices that are useful, but rare on commercial systems.
Finally, the availability of free software systems encourages and enables many, many "traditional" research projects at universities and government labs that may have profound impact. Clustering (beowulf's channel bonding, and the KLAT2 flat network neighborhood archetecture), load distribution/fault tolerance, distributed filesystem research, like coda. The NSA's secure linux project. Linuxbios -- 32 bit boot code that can boot up a PC in seconds flat. All these things would have been harder or impossible without free software that was easily tweakable to perform some new task.
Innovation isn't always about "revolutionizing the world" with a previously unheard of technology, which MS has never done. It can be about taking an existing idea and running with it, to produce the most stable/useful/powerful/efficient/ whatever implementation possible.
To give credit where credit is due, MS has done some of this. They took the GUI from Apple and others, and reimplemented it. It sucked, but they kept working at it long after most people would have given up. In the end, like it or not, the brought the GUI to the masses. And while windows 98 is not the most stable, efficient, or reliable, it is a GUI that most people can be taught to use for simple tasks, with a relatively small amount of pain.
Linux is getting there. Installation is still a little rougher than Windows installation (at least when windows installation actually works), but I think I could teach a computer illterate person to use linux for web, email, and word processing.
It is easy to say "free software is a threat to innovation" because to a traditional software industry point of view it sounds like it has to be true. But it is 100% contrary to fact -- free software encourages more innovation than proprietary software every could. And of course, there is the #1 innovation of free software that all others are lesser than, it puts the user back in control of his computer.
Everybody knows that corporations, Nike or otherwise, are not around to promote social justice.
However, we forget that what governments are *supposed* to do, what we pay taxes for and give up numerous personal freedoms for, is to promote social justice. The goverment's one and only true job is to create and enforce a legal system that promotes social justice and reduces suffering. That is rarely what they do, and it seems most people (including our elected representatives in the US) forget that this is what they are there for, but it is.
I don't object to Nike paying people a wage that is penuts on the US wealth scale. What I object to is Nike being allowed (by either the US or foriegn contries) to pay people less than a living wage for the area they live in, or to allow a manifestly unsafe work environment. That is Just Plain Wrong.
It is pure fallacy to think that if Nike couldn't pay people $.50/day to make shoes (or whatever) that all the people working for those wages would be jobless. Nike would just have to pay more -- and the cost of tennis shoes might go up.
A free market isn't about "let people do whatever they want". It is about letting the market forces drive supply and demand, rather than government price and production fixing. But we can and should manipulate those market forces (via taxes, tarrifs, tax credits, etc) to encourage socially responsible behavior. In a perfectly efficient market, this would result in all goods and services having its "true cost" -- defined as "how much does it cost the rest of the world for me to have this." which must take into account things like pollution, deforestation, poverty, crime, etc.
You analogy is seriously flawed. You can look at the river all you like. Within certain boundaries, you can do whatever you want with a river flowing through your property. What you cannot do is dump polution in the river that travles downstream, or divert the river, or dam it up. Likewise, my braodcast of radio waves is regulated by the FCC. But as far as I know, there is no government regulation on radio receivers.
The DMCA makes it illegal to make devices to decrypt these transmissions. So yes, dss cards are illegal under the DMCA.
Now, I personally believe that the DMCA is *wrong* (never confuse the law with what is right), and possibly unconstitutional (which would mean it was not merely unjust, but illegal as well). But until it is demonstrated unconstitutional by the Supreme Court, or otherwise repealed, descrambling
dss signals is illegal.
Someone mentioned wireless LANs. This falls into the same category. It is NOT illegal to intercept wireless LAN traffic on your own property. However, what you do with the information gained may or may not be illegal.
This is just Not True. Though no encryption agorithm other than a one time pad has been proven unbreakable, the foundations of computer science are based on the ability to calculate (for some problems) with 100% certainty that "you have to do operation o at least f(n) times to solve a this problem", and that certain problems (ie, the halting problem) cannot be solved by computers. Even quantum computing doesn't get around this, it just allows many parallel computations to take place at once.
I don't think that any wide spread encryption algorithm falls into either the "unsolvable" or "known scale super-polynomically", and I don't expect to see any of the former (that would make it kind of hard to decrypt), but super-polynomic encryption algorithms are certainly possible. That kind of algorithm, while crackable, can be made arbitrairly hard to crack, at much lower cost to the encryptor (assuming the actual alg. runs as a polynomial of the key size).
Quantum Encryption (which really isn't an encryption algorithm, but a protocol for securely exchanging one time pads) looks like it is provably secure. It is based on the principle that it is impossible to duplicate a 2-state system exactly.
In fact, in order to break RSA, you'd have to find a way to discover prime numbers without having to check them individually - that'd be quite an achievement and a mathematical triumph.
First, you mean prime factors, not prime numbers. It is very instructive to consider that there are indeed ways of finding prime numbers without checking them individually -- otherwise RSA encryption would be intractable, since it would take exponential runtime to generate keys, not just to break them. The interesting thing is, it is a probabalistic approach... it is indeed very difficult to generate known prime numbers, but relatively easy to generate numbers that are "probably" prime with arbitrarily small error margins. See _Applied Cryptograph_ for more info.
Given that, it seems a whole lot less "obvious" that it is not possible to find large prime factors in polynomial time--though based on the amount of research that has gone into it, I am inclined to believe it.
Likewise, Quantum Computing algorithms for factoring prime numbers are not perfect. Since QM is inherently statistical, it only has a statistical chance of getting the right answer -- of course, you can check it on a regular computer and try again if it doesn't work, but the point is, Quantum Computing algorithms may have very good runtime for various algorithms, but in general have completely unbounded worst case runtimes.
Quite a bit, actually--even as far out as they are, the satelites are still bound rather tightly to earth. Shoving them out a ways would just give them eccentric orbits, causing them to eventually make an uncontrolled reentry. It would take quite a kick to free them, or put them into a solar orbit.
However, I have a hard time imagning the odds they quote. Frankly, humans don't cover that much of the surface area of the earth. Maybe if you randomly dropped ~70 titanium fuel tanks on the earth from space you would get those odds, but even a lame attempt at deorbiting into the ocean should drop the odds substantially.
No question about the BX issue--especially if you like SMP (once you try it, you never go back). If I were to buy a motherboard today for my main workstation, it would be a SuperMicro P6DBE. Sadly enough, this is the same motherboard that I purchased 2 years ago. It is an excellent board, but I would hope for some progress in two years.
Hm. While their computer knowledge has never impressed me, I have found them to be quite reasonable when I say "just looking".
I have had to wait (seemingly) 6 hours while the dude explains the extended warranty on and HP Pavillion to the granny in front of me when all I want is for him to hand me a hard drive off the shelf behind the desk.
Fun way to torment Best Buy/Staples/CompUSA employees of the day: Go conspiciously hang out in the networking aisle, say you are looking at, say a linksys 10/100 hub versus switch, and then when he tries to explain the difference, ask "Can you tell me if this switch uses store-and-forward or cut-through switching"?
RE: prior art, please read Transmeta's patents before flying off the handle here. To the best of my knowledge, TM didn't patent dynamic translation, but they have several patents on optimizations for dynamic translation. Most of them are particularly suited to the situation where the hardware can be specifically designed to help out the translator, like the shadow registers used to insure precice exception behavior.
The biggest drags on performance of (say) x86 systems is not lack of runtime profile data to the compiler, but from hard-to-parallelize restrictions placed on operations by both the language and the host ISA. Dynamic translation allows you to optimze for the 99% case, and patch up the other 1% with an exception handler. Register allocation may be a little hard to deal with, but scheduling, prefeching and instruction selection are exactly what dynamic translation should fix up.
It might be faster to use an IR or VLIW->VLIW translator rather than using a "real" ISA as your source, but I think that is the wrong comparison. The advantage of dynamic translation is that it can potentially be used to execute x86 (or whatever) code on a VLIW machine much faster than on a native machine. BTW, one of their source ISAs is JVM, which really is an IR.
I think this set of people would be isomorphic to the set of people who would have child porn, credit card fraud, copyright violation, or be operating against the acceptable use policy of their ISP.
Thus, it becomes a simple matter of procecuting them for fraud or whatever, rather than "illegal porn on.com". I doubt that in the presence of.xxx, any legally operated porn site would have find an economic advantage to use non.xxx domains, in any capacity other than as "redirectors" to the.xxx domain.
Content based filtering might still be deemed necessary by some people, but far fewer, and it could be made more effective, with fewer false positives (which is all I personally care about -- I think it is totally inappropriate for schools/libraries/etc to use filters with the level of false positives that they have today).
I think the.xxx domain should be created, and be completely voluntary. I believe the business case for exclusive, or nearly exclusive use of.xxx domains is good for porn sites, and most of them would voluntarily use.xxx domains.
Combine that with ISPs that require use of a.xxx domain for porn sites, and most, if not all of the "problem" here would be solved without any government regulation or use of error prone keyword or site list filtering software.
The idea of the.xxx domain name is (or should be) to give a universal mechanism for regulation w/o mandating any specific regulations.
Why on earth would a porn site not *want* to be filtered? Why would a porn site want to be in the.kids domain? The answer is, no reason. Nobody behind any sort of filtering software is going to be in a position to pay for pr0n, and certainly kids are not.
For the most part pr0n webmasters are not out to corrupt the youth, but to make a buck off of paying subscribers.
"Non-freakshow"? Not sure what you mean by that. My empeg player was expensive, but is an incredibly well designed and built machine. It isn't mainstream but it isn't frankenstiens monster, either.
The idea behind the "quantum leap" phrase is that it is discreet. A quantum transition from eg. An N=1 to N=2 state never passes through "N=1.5" (which doesn't even exist, or make sense at all). So people adopted the term in a colloquial fashion to refer to an instantaneous jump, as opposed to a gradual "classical" advancment.
Of course, like every other colloquialism, it quickly became grossly over- and mis- used.
Tux2 will eventually support snapshots as well. Basically, the way it works is that you have a "snapshot" command that clones the metadata tree. Then, like the name implies, you can access the filesystem as it was the moment the snapshot was made.
First, however, we are going to do the data integrity.
VMWare works at the block device level for its rollback system--all dirty blocks on a device are stored in memory while VMWare is running, then at shutdown you can either discard them or flush them. While it allows you to bail out of a horked session, it makes no guarantees about data integrity while the block device is actually being written to.
Basically, in Intel nomenclature, CPUs are grouped into 4-CPU prossor groups, which use a shared bus, and the majority of the bus arbitration is done by the CPUs themselves, which is a big cost saver on motherboard design.
To go above 4 CPUs, you have a couple of options. You can "fake it" which how the ALR 6 CPU ppro motherboard you can buy on eBay work -- Put 2 groups of three CPUs on a shared bus, and with a relatively simple, but very clever hack, you can make the fourth CPU in each group a "phantom" that represents the other CPU group.
If you want to go beyond that you need to handle arbitration between the processor groups on the motherboard. This could be done either shared data bus or with a switched interconnect, with a price/performance tradeoff. I don't know how current systems work, but I imagine it is difficult to get a CPU designed for shared-bus MP to handle locked bus cycles and cache coherency when they aren't on a shared bus.
They (GCC steering committee) think RH did something irresponsible and foolish that will cause lots of trouble for them. This is different than illegal/against the license. They are also upset that RH slapped a GCC release number on an unofficial snapshot, which misrepresents the software as a GCC committee endorsed release.
They also wish RH had talked to them before doing this, as they might have been able to mitigate the damage on both sides.
These are all reasons to be upset that RH did something entirely within their rights, but which was very annoying anyway.
Binary compatability with other distributions is much less important. You can't install a.deb on RH, you can't really install a SuSE.rpm, and these days you push your luck with a mdk rpm anyway.
It isn't really that hard to recompile OSS code for a different distribution, anyway. Comercial software companies, I guess are SOL until most distributions have a gcc 3.0 blessed set of libraries, so they will probably have to statically link against libstdc++ if they want portability.
RH does ship compatabilitiy libraries whenever a new release breaks binary compatability, though I don't know if they tricks they play will help out a binary compiled on a different distribution.
C++ name munging will change between gcc 2.96 and gcc 3.0. So, any program which links against a C++ library will get unresolved symbols if it can't find a version of said library that was compiled with the old compiler.
RH always includes compatability libraries for these situations, though they aren't always installed by default (if you are lazy: rpm -Uvh/redhat/i386/RedHat/RPMS/*compat* usually works). Thus an rpm built on RH 6.2 should work on RH 7, but not the other way around (since RH6.2 won't have the 7.0 libraries).
If you always build on 6.2, or you don't use C++, or you don't linke against C++.so's, you will almost certainly be fine.
If your code doesn't conform to the C++ standard, people may not be able to compile it on RH 7, since that compiler is more strict than older versions. Theoretically no valid C++ programs that compiled under egcs or gcc 2.95 should break, but there may be more bugs in this snapshot.
Slashdot Mirror
First, where else would they go? Nike has to make shoes, it is their job. If they don't do that, they go out of business and default on all their sponsorship contracts. If people won't buy shoes made by workers earning a living wage, they are not a viable product and don't deserve to be produced. But I don't think that will happen. Like I said, Nike might have to raise prices, or pay sports figures less for endorsments, but they will still stay in business.
What would be bad for them is not all shoe companies had to operate on an even footing--if they were held to higher standards than other companies--that is why we need laws and governments for this sort of thing, rather than relying on the goodwill of individual companies.
Too many people don't pay attention in economics class, and all they get out is "humans are rational, self interested parties." This is a questionable hypothesis to begin with, largely misunderstood, and ignores the more important thesis "market forces drive a free market economy". By manipulating those market forces, it is possible (within limits) to control a "free market" economy. Depending on the means and the ends, such manipulations can be monopolistic behavior or collusion (bad), other times it is optimization, and good.
Neither are particularly related to Soviet socailims where the market was driven by government quotas, not supply and demand.
What part of the "THESE DATES ARE NOT REALISTIC" enclosed in blink tags did you not understand?
He didn't stake a beer on that date, much less his life.
Very often, that is true, though I would say you underestimate the output of the "million monkeys". Go to freshmeat and just browse. Sure, much of it isn't earth shattering, or "pure research", but there are lots and lots of programs that are valuable in their own right. Other programs for Windows or UNIX may have similar functionality, but each one was designed to perform in a particular way as well as possible. In the words of ESR, they all started to scratch an itch.
.NET and "subscription software", where you can install off the net and get billed a subscription, but it isn't worth the cost of giving up control over your computer to get.
A few notable examples:
licq -- implements similar functionality to windows ICQ, but IMO has the best UI of any instant messanger I have seen. Also, recently added support for SSL encrypted communications, and has a wide array of plugins (nmap, finger, etc) that windows ICQ and other ICQ programs lack.
lame: conforms to the MPEG audio encoder standard. Similar to several other encoders around, but has (arguably) the best sound quality of any available encoder.
Mame: free arcade emulator that runs on an incredibly number of platforms, including laser light show controllers and digital camers. These two are good examples of features that have no commercial value, and thus would never be implemented by a commercial developer, yet are way cool, and some people might use.
Apache: originally based on the NCSA httpd, Apache is now the most flexible web server in existence.
apt-get: IMNSHO, debian has nearly perfected the process of software distribution, installation, and version management. "apt-get install mozilla" makes inserting a cd and clicking OK from the autorun dialog look horribly complicated. And no commercial software company will ever match it, because they can't keep track of licensing then. Maybe MS will come up with something 10% as good with
Linux Kernel: The linux kernel isn't just another POSIX implementation. It has an unheard of combination of features for desktop, workstation, server, and embedded system all in one package. It runs more or less the same on many, many platforms, and has features like loopback block devices that are useful, but rare on commercial systems.
Finally, the availability of free software systems encourages and enables many, many "traditional" research projects at universities and government labs that may have profound impact. Clustering (beowulf's channel bonding, and the KLAT2 flat network neighborhood archetecture), load distribution/fault tolerance, distributed filesystem research, like coda. The NSA's secure linux project. Linuxbios -- 32 bit boot code that can boot up a PC in seconds flat. All these things would have been harder or impossible without free software that was easily tweakable to perform some new task.
Innovation isn't always about "revolutionizing the world" with a previously unheard of technology, which MS has never done. It can be about taking an existing idea and running with it, to produce the most stable/useful/powerful/efficient/ whatever implementation possible.
To give credit where credit is due, MS has done some of this. They took the GUI from Apple and others, and reimplemented it. It sucked, but they kept working at it long after most people would have given up. In the end, like it or not, the brought the GUI to the masses. And while windows 98 is not the most stable, efficient, or reliable, it is a GUI that most people can be taught to use for simple tasks, with a relatively small amount of pain.
Linux is getting there. Installation is still a little rougher than Windows installation (at least when windows installation actually works), but I think I could teach a computer illterate person to use linux for web, email, and word processing.
It is easy to say "free software is a threat to innovation" because to a traditional software industry point of view it sounds like it has to be true. But it is 100% contrary to fact -- free software encourages more innovation than proprietary software every could. And of course, there is the #1 innovation of free software that all others are lesser than, it puts the user back in control of his computer.
Everybody knows that corporations, Nike or otherwise, are not around to promote social justice.
However, we forget that what governments are *supposed* to do, what we pay taxes for and give up numerous personal freedoms for, is to promote social justice. The goverment's one and only true job is to create and enforce a legal system that promotes social justice and reduces suffering. That is rarely what they do, and it seems most people (including our elected representatives in the US) forget that this is what they are there for, but it is.
I don't object to Nike paying people a wage that is penuts on the US wealth scale. What I object to is Nike being allowed (by either the US or foriegn contries) to pay people less than a living wage for the area they live in, or to allow a manifestly unsafe work environment. That is Just Plain Wrong.
It is pure fallacy to think that if Nike couldn't pay people $.50/day to make shoes (or whatever) that all the people working for those wages would be jobless. Nike would just have to pay more -- and the cost of tennis shoes might go up.
A free market isn't about "let people do whatever they want". It is about letting the market forces drive supply and demand, rather than government price and production fixing. But we can and should manipulate those market forces (via taxes, tarrifs, tax credits, etc) to encourage socially responsible behavior. In a perfectly efficient market, this would result in all goods and services having its "true cost" -- defined as "how much does it cost the rest of the world for me to have this." which must take into account things like pollution, deforestation, poverty, crime, etc.
You analogy is seriously flawed. You can look at the river all you like. Within certain boundaries, you can do whatever you want with a river flowing through your property. What you cannot do is dump polution in the river that travles downstream, or divert the river, or dam it up. Likewise, my braodcast of radio waves is regulated by the FCC. But as far as I know, there is no government regulation on radio receivers.
The DMCA makes it illegal to make devices to decrypt these transmissions. So yes, dss cards are illegal under the DMCA.
Now, I personally believe that the DMCA is *wrong* (never confuse the law with what is right), and possibly unconstitutional (which would mean it was not merely unjust, but illegal as well). But until it is demonstrated unconstitutional by the Supreme Court, or otherwise repealed, descrambling
dss signals is illegal.
Someone mentioned wireless LANs. This falls into the same category. It is NOT illegal to intercept wireless LAN traffic on your own property. However, what you do with the information gained may or may not be illegal.
This is just Not True. Though no encryption agorithm other than a one time pad has been proven unbreakable, the foundations of computer science are based on the ability to calculate (for some problems) with 100% certainty that "you have to do operation o at least f(n) times to solve a this problem", and that certain problems (ie, the halting problem) cannot be solved by computers. Even quantum computing doesn't get around this, it just allows many parallel computations to take place at once.
I don't think that any wide spread encryption algorithm falls into either the "unsolvable" or "known scale super-polynomically", and I don't expect to see any of the former (that would make it kind of hard to decrypt), but super-polynomic encryption algorithms are certainly possible. That kind of algorithm, while crackable, can be made arbitrairly hard to crack, at much lower cost to the encryptor (assuming the actual alg. runs as a polynomial of the key size).
Quantum Encryption (which really isn't an encryption algorithm, but a protocol for securely exchanging one time pads) looks like it is provably secure. It is based on the principle that it is impossible to duplicate a 2-state system exactly.
Given that, it seems a whole lot less "obvious" that it is not possible to find large prime factors in polynomial time--though based on the amount of research that has gone into it, I am inclined to believe it.
Likewise, Quantum Computing algorithms for factoring prime numbers are not perfect. Since QM is inherently statistical, it only has a statistical chance of getting the right answer -- of course, you can check it on a regular computer and try again if it doesn't work, but the point is, Quantum Computing algorithms may have very good runtime for various algorithms, but in general have completely unbounded worst case runtimes.
Don't worry, but the time it reaches earth, it will be no larger than a chihuahua's head :)
Quite a bit, actually--even as far out as they are, the satelites are still bound rather tightly to earth. Shoving them out a ways would just give them eccentric orbits, causing them to eventually make an uncontrolled reentry. It would take quite a kick to free them, or put them into a solar orbit.
However, I have a hard time imagning the odds they quote. Frankly, humans don't cover that much of the surface area of the earth. Maybe if you randomly dropped ~70 titanium fuel tanks on the earth from space you would get those odds, but even a lame attempt at deorbiting into the ocean should drop the odds substantially.
No question about the BX issue--especially if you like SMP (once you try it, you never go back). If I were to buy a motherboard today for my main workstation, it would be a SuperMicro P6DBE. Sadly enough, this is the same motherboard that I purchased 2 years ago. It is an excellent board, but I would hope for some progress in two years.
/me waits for the 760MP with dual tbirds.
Hm. I like the blue dudes. Better than the dust bunnies at least.
Hm. While their computer knowledge has never impressed me, I have found them to be quite reasonable when I say "just looking".
I have had to wait (seemingly) 6 hours while the dude explains the extended warranty on and HP Pavillion to the granny in front of me when all I want is for him to hand me a hard drive off the shelf behind the desk.
Fun way to torment Best Buy/Staples/CompUSA employees of the day: Go conspiciously hang out in the networking aisle, say you are looking at, say a linksys 10/100 hub versus switch, and then when he tries to explain the difference, ask "Can you tell me if this switch uses store-and-forward or cut-through switching"?
RE: prior art, please read Transmeta's patents before flying off the handle here. To the best of my knowledge, TM didn't patent dynamic translation, but they have several patents on optimizations for dynamic translation. Most of them are particularly suited to the situation where the hardware can be specifically designed to help out the translator, like the shadow registers used to insure precice exception behavior.
The biggest drags on performance of (say) x86 systems is not lack of runtime profile data to the compiler, but from hard-to-parallelize restrictions placed on operations by both the language and the host ISA. Dynamic translation allows you to optimze for the 99% case, and patch up the other 1% with an exception handler. Register allocation may be a little hard to deal with, but scheduling, prefeching and instruction selection are exactly what dynamic translation should fix up. It might be faster to use an IR or VLIW->VLIW translator rather than using a "real" ISA as your source, but I think that is the wrong comparison. The advantage of dynamic translation is that it can potentially be used to execute x86 (or whatever) code on a VLIW machine much faster than on a native machine. BTW, one of their source ISAs is JVM, which really is an IR.
I think this set of people would be isomorphic to the set of people who would have child porn, credit card fraud, copyright violation, or be operating against the acceptable use policy of their ISP.
.com". I doubt that in the presence of .xxx, any legally operated porn site would have find an economic advantage to use non .xxx domains, in any capacity other than as "redirectors" to the .xxx domain.
Thus, it becomes a simple matter of procecuting them for fraud or whatever, rather than "illegal porn on
Content based filtering might still be deemed necessary by some people, but far fewer, and it could be made more effective, with fewer false positives (which is all I personally care about -- I think it is totally inappropriate for schools/libraries/etc to use filters with the level of false positives that they have today).
I think the .xxx domain should be created, and be completely voluntary. I believe the business case for exclusive, or nearly exclusive use of .xxx domains is good for porn sites, and most of them would voluntarily use .xxx domains.
.xxx domain for porn sites, and most, if not all of the "problem" here would be solved without any government regulation or use of error prone keyword or site list filtering software.
.xxx domain name is (or should be) to give a universal mechanism for regulation w/o mandating any specific regulations.
Combine that with ISPs that require use of a
The idea of the
Why on earth would a porn site not *want* to be filtered? Why would a porn site want to be in the .kids domain? The answer is, no reason. Nobody behind any sort of filtering software is going to be in a position to pay for pr0n, and certainly kids are not.
For the most part pr0n webmasters are not out to corrupt the youth, but to make a buck off of paying subscribers.
"Non-freakshow"? Not sure what you mean by that. My empeg player was expensive, but is an incredibly well designed and built machine. It isn't mainstream but it isn't frankenstiens monster, either.
The idea behind the "quantum leap" phrase is that it is discreet. A quantum transition from eg. An N=1 to N=2 state never passes through "N=1.5" (which doesn't even exist, or make sense at all). So people adopted the term in a colloquial fashion to refer to an instantaneous jump, as opposed to a gradual "classical" advancment.
Of course, like every other colloquialism, it quickly became grossly over- and mis- used.
Tux2 will eventually support snapshots as well. Basically, the way it works is that you have a "snapshot" command that clones the metadata tree. Then, like the name implies, you can access the filesystem as it was the moment the snapshot was made.
First, however, we are going to do the data integrity.
VMWare works at the block device level for its rollback system--all dirty blocks on a device are stored in memory while VMWare is running, then at shutdown you can either discard them or flush them. While it allows you to bail out of a horked session, it makes no guarantees about data integrity while the block device is actually being written to.
Who else is waiting for the PIII 1337?
Basically, in Intel nomenclature, CPUs are grouped into 4-CPU prossor groups, which use a shared bus, and the majority of the bus arbitration is done by the CPUs themselves, which is a big cost saver on motherboard design.
To go above 4 CPUs, you have a couple of options. You can "fake it" which how the ALR 6 CPU ppro motherboard you can buy on eBay work -- Put 2 groups of three CPUs on a shared bus, and with a relatively simple, but very clever hack, you can make the fourth CPU in each group a "phantom" that represents the other CPU group.
If you want to go beyond that you need to handle arbitration between the processor groups on the motherboard. This could be done either shared data bus or with a switched interconnect, with a price/performance tradeoff. I don't know how current systems work, but I imagine it is difficult to get a CPU designed for shared-bus MP to handle locked bus cycles and cache coherency when they aren't on a shared bus.
They (GCC steering committee) think RH did something irresponsible and foolish that will cause lots of trouble for them. This is different than illegal/against the license. They are also upset that RH slapped a GCC release number on an unofficial snapshot, which misrepresents the software as a GCC committee endorsed release.
They also wish RH had talked to them before doing this, as they might have been able to mitigate the damage on both sides.
These are all reasons to be upset that RH did something entirely within their rights, but which was very annoying anyway.
Binary compatability with other distributions is much less important. You can't install a .deb on RH, you can't really install a SuSE .rpm, and these days you push your luck with a mdk rpm anyway.
It isn't really that hard to recompile OSS code for a different distribution, anyway. Comercial software companies, I guess are SOL until most distributions have a gcc 3.0 blessed set of libraries, so they will probably have to statically link against libstdc++ if they want portability.
RH does ship compatabilitiy libraries whenever a new release breaks binary compatability, though I don't know if they tricks they play will help out a binary compiled on a different distribution.
C++ name munging will change between gcc 2.96 and gcc 3.0. So, any program which links against a C++ library will get unresolved symbols if it can't find a version of said library that was compiled with the old compiler.
/redhat/i386/RedHat/RPMS/*compat* usually works). Thus an rpm built on RH 6.2 should work on RH 7, but not the other way around (since RH6.2 won't have the 7.0 libraries).
.so's, you will almost certainly be fine.
RH always includes compatability libraries for these situations, though they aren't always installed by default (if you are lazy: rpm -Uvh
If you always build on 6.2, or you don't use C++, or you don't linke against C++
If your code doesn't conform to the C++ standard, people may not be able to compile it on RH 7, since that compiler is more strict than older versions. Theoretically no valid C++ programs that compiled under egcs or gcc 2.95 should break, but there may be more bugs in this snapshot.