P4 vs. Athlon XP race is irrelevant. AMD need not take the speed crown again, they just need to keep their existing market share until Hammers hit stores. BTW, Athlon XP is still very fast, moreover it is avaliable now unlike benchmarked P4s. If you can't help lusting for even more speed, a dual athlon xp box is quite affordable unlike a dual P4 Xeon box.
Dos texts are standart? That is news to me. Wish I knew that before all time I spend on cr/lf incompatibilities.
Think about it, how much you can get non-standart with pure text files? The obvious way is you can redefine whole character set. DOS kind of did this by using nonstandart codes for non-latin character sets and redefining some control characters.
Nope, although my original point (or troll, as mods would rather call) was about broken and imcompatible clipboard implementations (try cut&paste between galeon, mozilla, konqueror, gvim & kate) its specification is also braindead. I -like many with macos or windows background- select things I don't intend to copy, cut or paste over. Sometimes I do that to count things using selection as a marker. Sometimes I use it to remedy bad color selection on web pages like light-blue-on-grey. Sometimes I use it to remember where I was while scrolling to see where document ends etc. Under X I can do these only if I don't intend to use active clipboard. Also if I intend to replace text, it is perfectly okay to select and start writing, whatever I write overwrites whole selection, but I can't do that if the input will come from clipboard. This is inconsistent behavior.
I remain unconvinced that they will noticably increase performance. It doesn't really matter how many registers are avaliable to the programmer, the only important thing is how many registers are actually used. Athlon cores already have dozens of extra registers utilized by register renaming; making some of them visible will remove them from the renamer's pool. Any possible performance increase is due to compiler's ability to use registers better than CPU but you should not take that for granted, CPU might make better use of those registers when they are internal. After the x86-64 compilers have matured somewhat, we will most probably see a performance increase due to reduced cache/memory load, but don't expect that to be spectecular.
...but having owned a first gen. athlon classic on a first generation mobo (using a first gen. AMD chipset), and two first gen. Athlon XPs on a first gen. Tiger S2460 mobo (using a first gen. AMD MP chipset), I recommend you to stay away from anything from AMD before its second incarnation. And that will be 2003H1 for hammers, can you wait that long? BTW in both cases problems were related to the chipset, BIOS and lack of workarounds for discontinued motherboards; all 3 cpus were excellent.
I found this website [rennard.org]
on genetic algorithms. I have no idea how accurate it is, since i have never studied it before. Let me know if you think it contains a poor introduction to the topic or not.
As an introduction, it looks ok, but I doubt that you can make use of that page since it is loaded with similarities between natural and artificial evolution. Actually evolution is a matemathical/statistical phenomenon, I gave a reference to GAs only because mathematical analysis of evolution is rarely done elsewhere. Lets call this mathematical phenomenon evolution Ems. Biological entities have also been subject to an evolution, lets call that EB. Whether EB is an instance of Ems, or whether EB is an instance of natural/automatic (as in: without any external influence) Ems are different issues and not my focus now. I just want you to find out about Ems, so that we can discuss why I think EB is an instance of unguided Ems, any you can argue why it isn't.
Under genetic algorithms and real evolution, changes are gradual and minor (the theory, anyway). There are a number of factors that make such changes have an almost non-existent effect on the chances of the creature's survival above everyone else's.
Do genetic algorithms deal with the problem I presented? Namely that a multitude of harmful mutations will lie dormant together with one beneficial mutation - and the beneficial one only expressing itself at the same time as the many harmful mutations also will?
Yes, it is called genetic hitchiking in literature. It is a tradeoff, controlled by mapping and crossover operators; higher the probability of meaningful chunks staying together in an offspring(good thing) and persisting in population in time (another good thing), higher the probability of hitchiking (bad thing) and early convergence (a real bad thing.) But keep in mind that had there been no hitchiking as in all bad mutations are immediatly destroyed, a very early convergence is assured, degenerating the GA to hill climbing. Biological equivalent of early convergence is a single species in a single habitat, and not necessarily best fit one on the long term. The reverse of that is high mutation rates and splicing genes at positions that break coherent interpretation, which degenerates the GA to random searh. This has the biological analogy of subjecting sperms to almost lethal doses of radiation while hoping that the mutations will be somehow benefical for the baby.
Also do genetic algorithms represent changes in information rather than addition of information? When a genetic mutation occurs, it changes some already existing information, rather than being added. This is a serious question and not a criticism of genetic algorithms, because I don't know anything about this topic.
Yes they do. The plain vanilla GAs almost exclusively use both constant length genome representations, and a direct mapping thru a stateless function onto phenotype. Your critisim is valid in that case. GAs are usually used for static and closed problem solutions, like "there is one or many solutions, and we want to find one or some of them that fits these criteria." In a dynamic problem (such as survival in biology or game playing in AI) potential solutions are infinite and may change in time. Celular automata encodings, variable length phenotypes, evolution of genotype->phenotype mapping etc. are solutions to these. You also need a dynamic objective function, which suggests that even the "same" traits will have different effects over time and should be trated as such.
You have a few fairly basic misunderstandings in your text that can be trivially answered by studying changes in a population of evolving agents in a religion neutral context. I suggest you to study artificial evolution in genetic algorithms; that way you will see the pure mathematical reasons why evolution works the way it does, and in what alternative ways it could have worked (since a GA researcher is rarely interested in duplicating natural evolution) or failed. All four facts you mention are common to both artifical evolution systems and natural biological beings, and as such can be explained (in the sense that they are not troublesome) thru a purely population mechanistic viewpoint.
I didn't want to discuss with you just yet because of nonsense in the last paragraph, but once you do your homework about artifical evolution, I will gladly point out how GA concepts map onto natural evolution concepts.
The point is you can't put an NDA on Linux. That's relicensing GPL code.
FWIW someone already posted two quotes (from GPL FAQ I believe) about NDAs with GPL. It seems one can indeed put an NDA for development purposes but not for distribution. But it also seems that by NDAs, you can only put limits redistribution; anyone with binary access still has access to source.
You have a point but my point was not that you can redistribute GPLed software with alternative licenses, open or closed. It was "not everything qualifies for the term 'distribution' and this may be such a case."
If lindows makes this "beta program" the only way to access their software even after it is release quality, then one can definetly argue they are breaking GPL by not providing source. I guess we are not at that point yet.
I, the CEO of LinFlows Inc., hire a bunch of people to do beta testing for my GPL based new product LinFlows. Obviously when I release the product, I must observe GPL's source related articles. Yet do I have to give people I hired for my company source code together with binaries? My understanding is that is not necessary.
Whether lindows case is more like this LinFlows case or the picture you present can only be determined by knowing and analysing license agreement between beta participants and lindows. Without this information the discussion is meaningless.
Well I would like to emphasize "closed" nature of the pay beta program. I am not an expert on this matter and I won't pretend to be one. Yet the way I see this is, the beta program of lindows is actually an internal process. The payers may access this internal process after signing a NDA. They are not selling or giving away their betas, betas are also not free to copy, they are closed and whole proces is internal.
Obviously "distribution of software" mentioned in GPL is not something done internally; I can make modifications to any GPLed software, not release the final software but use it internally at my company with how-many-I-desire copies installed without breaking GPL. I don't see much difference between internal usage and an internal beta program. I may be off the base here, and the very fact that one can participate in beta program might make the process open, no matter how many NDAs people sign. But if the program is indeed closed and internal, he may withold any source until it is released.
Secondly, it is not true that we will remain militarily invincible forever. It would be *more* true if we stopped exporting our best, or next-to best, military technology to whichever fascist regime we wanted to prop up today, but nonetheless, if we stopped improving our military the rest of the world would eventually catch up.
I understand that you have no knowlege about first wave of Turkish attack on Cyprus in 1974. We had all imported equipment mostly from USA and guess what, the minute we tried to use them against will of USA we started to see ghost ships and planes around, communications has been disrupted leading us to sink our own ship, navigation systems failed... Lesson learned and we build our own military industry but I think most third world countries still haven't taken that step. Exporting high tech military equipment makes USA more powerful, not less. Also don't forget military exports finance further research, if you stopped exporting them, it would have been easier for other to catch you.
Apply every possible key with every possible encryption algorithm and you end up with nothing but every possible bit sequence that you know how to derive from that particular bit sequence. Some should make sense, but which one is the one you are looking for? Try decrypt 10000 0s with that, you will arrive a message, but that message will probably have nothing to do with my original message.
Theory 1: Everthing you observe is because I want you to observe that way. You cannot figure anything beyond because you neither have perception beyond that nor you can figure out what I like and in which ways I make you see things different than they actually are. You see, you also observe a "me" which is substantially different from real me. I can't tell you what reality is, it would be no fun. Even this post is partially true but it is as close to reality as you can/will ever manage; I'll see to that.
Theory 2: Whatever I say is internally consistent because I say so. Everthing is either something I would say, or not consistent with reality. You might think othewise time to time but you are obviously mistaken when you do that.
An internal gravitation model would be theorizing far more than is necesarry to account for the data. In cognitive science, there has long been an understanding of encoding specificity. This simply means that data, including skill knowledge, is best retrieved from human memory under the same conditions which it was learned.
All true, but how does this apply in this case? If your theory is that astronout fails to remember how one correctly responds to falling balls in space because he has not learned that skill in that environment, that is also theorizing far more than data suggests. Also that theory will have hard time explaining why non-motor skills related to gravity does not suffer likewise in space.
OTOH, assuming having expectations about how the world and objects in it will behave without resorting to a native bias for that expectations can be easily justified. Under this assumption, astronouts inability to efficiently catch those balls simply results from failure to correctly foresee how objects will behave, gravity-wise. They learn slowly because of a negative interference from long term behavior they enjoyed on Earth. They relearn in Earth's gravity quickly because of the experience's deep roots from childhood.
But this is also a simple to understand layman's theory, requiring no expansive speech to express. So I understand why it is unpopular among us, the cognitive scientists.
I don't take broken promises lightly. They never promised to supply us staroffice, but they did prmise that we will be treated equal. I want them to keep their promise, not fscking staroffice.
I'm a basic level member of mandrake club. I didn't expect anything material in return when I subscribed, it was more like a donation. I tought "well, I'm using their product, and they need money, so let me help them" and subscribed. I wasn't even aware that a commercial product might be free for members. I just thought we would have a few dedicated mirrors and a louder voice when we complain.
And they made SO6 free for higher class members. I have no problem with that either, I could have used SO6 (I'm still using OO 639c) but it is possible to buy it from Sun anyway. All members can buy pro suite at a discounted rate which has SO6. And those silver and gold members did pay more than me, so it is only natural they can profit more from their membership.
My only problem is that mandrake failed to keep their promise. They categorically stated that we would all be same, enjoy membership goodies no matter how much we payed. When I first saw it, I thought it was dumb to say. But they did say that. I didn't request to be treated same with other levels, even if I paid considerably less; it was their promise
Why should I trust mandrake now? Which promise they will break next? These are the real problems with their move. SO6 has nothing to do with discussion.
And another thing is, I just whine. Perhaps somebody will be enough pissed so they won't be satisfied with just whining. Do you think mdk has a chance in the court? How will the company survive if it is forced to make SO6 downloads free for every member?
Once again the key questions are: "Does it work? Is it repeatable". If the answer to those questions are "Yes" nothing else matters. Theory has to be adjusted to fit reality - not the other way around.
I'm an engineer and I do quite a few things that would perhaps require man-years to explain with the current theories. I have never spent those man-years and it is possible that some scientist, after spending those man-years, will argue that what seems (their word) to happen is incompatible with theory. I have no problem with that. I produce, it works, nothing else matters. If those things fly, do your best to make use of that!
I'm willing to accept that some unknown process is causing those kites to float. I just don't see justification for that. I don't reject reality, I just don't buy a particular explanation about it. My problem with the hypothesis is basically:
I would say that a crude electrogravitic kite lifting an 18 gram payload is pretty extraordinary.
this. I don't think it is extraordinary at all. If those things can move out of Van Allen belt and athmosphere, it would be extraordinary. That means -whether utilizing reactionless force or not- they are feasible space engines. I didn't see enough controlled experiments to suggest that they will function in that environment though. As long as those things fly on earth, it doesn't matter how they do it; we have enough flying artifacts here already.
Slashdot Mirror
P4 vs. Athlon XP race is irrelevant. AMD need not take the speed crown again, they just need to keep their existing market share until Hammers hit stores. BTW, Athlon XP is still very fast, moreover it is avaliable now unlike benchmarked P4s. If you can't help lusting for even more speed, a dual athlon xp box is quite affordable unlike a dual P4 Xeon box.
A great (and probably the only) opportunity to run a fanless Athlon and they select a puny Motorola. Duh!
Perhaps they will also notice the real name for the project and product: OpenOffice.org
Think about it, how much you can get non-standart with pure text files? The obvious way is you can redefine whole character set. DOS kind of did this by using nonstandart codes for non-latin character sets and redefining some control characters.
Nope, although my original point (or troll, as mods would rather call) was about broken and imcompatible clipboard implementations (try cut&paste between galeon, mozilla, konqueror, gvim & kate) its specification is also braindead. I -like many with macos or windows background- select things I don't intend to copy, cut or paste over. Sometimes I do that to count things using selection as a marker. Sometimes I use it to remedy bad color selection on web pages like light-blue-on-grey. Sometimes I use it to remember where I was while scrolling to see where document ends etc. Under X I can do these only if I don't intend to use active clipboard. Also if I intend to replace text, it is perfectly okay to select and start writing, whatever I write overwrites whole selection, but I can't do that if the input will come from clipboard. This is inconsistent behavior.
His browser is running under X.
I remain unconvinced that they will noticably increase performance. It doesn't really matter how many registers are avaliable to the programmer, the only important thing is how many registers are actually used. Athlon cores already have dozens of extra registers utilized by register renaming; making some of them visible will remove them from the renamer's pool. Any possible performance increase is due to compiler's ability to use registers better than CPU but you should not take that for granted, CPU might make better use of those registers when they are internal. After the x86-64 compilers have matured somewhat, we will most probably see a performance increase due to reduced cache/memory load, but don't expect that to be spectecular.
...but having owned a first gen. athlon classic on a first generation mobo (using a first gen. AMD chipset), and two first gen. Athlon XPs on a first gen. Tiger S2460 mobo (using a first gen. AMD MP chipset), I recommend you to stay away from anything from AMD before its second incarnation. And that will be 2003H1 for hammers, can you wait that long? BTW in both cases problems were related to the chipset, BIOS and lack of workarounds for discontinued motherboards; all 3 cpus were excellent.
When you are done, feel free to contact me on my mail address. /. is not a good place for serious discussion anyway.
As an introduction, it looks ok, but I doubt that you can make use of that page since it is loaded with similarities between natural and artificial evolution. Actually evolution is a matemathical/statistical phenomenon, I gave a reference to GAs only because mathematical analysis of evolution is rarely done elsewhere. Lets call this mathematical phenomenon evolution Ems. Biological entities have also been subject to an evolution, lets call that EB. Whether EB is an instance of Ems, or whether EB is an instance of natural/automatic (as in: without any external influence) Ems are different issues and not my focus now. I just want you to find out about Ems, so that we can discuss why I think EB is an instance of unguided Ems, any you can argue why it isn't.
Under genetic algorithms and real evolution, changes are gradual and minor (the theory, anyway). There are a number of factors that make such changes have an almost non-existent effect on the chances of the creature's survival above everyone else's.
Do genetic algorithms deal with the problem I presented? Namely that a multitude of harmful mutations will lie dormant together with one beneficial mutation - and the beneficial one only expressing itself at the same time as the many harmful mutations also will?
Yes, it is called genetic hitchiking in literature. It is a tradeoff, controlled by mapping and crossover operators; higher the probability of meaningful chunks staying together in an offspring(good thing) and persisting in population in time (another good thing), higher the probability of hitchiking (bad thing) and early convergence (a real bad thing.) But keep in mind that had there been no hitchiking as in all bad mutations are immediatly destroyed, a very early convergence is assured, degenerating the GA to hill climbing. Biological equivalent of early convergence is a single species in a single habitat, and not necessarily best fit one on the long term. The reverse of that is high mutation rates and splicing genes at positions that break coherent interpretation, which degenerates the GA to random searh. This has the biological analogy of subjecting sperms to almost lethal doses of radiation while hoping that the mutations will be somehow benefical for the baby.
Also do genetic algorithms represent changes in information rather than addition of information? When a genetic mutation occurs, it changes some already existing information, rather than being added. This is a serious question and not a criticism of genetic algorithms, because I don't know anything about this topic.
Yes they do. The plain vanilla GAs almost exclusively use both constant length genome representations, and a direct mapping thru a stateless function onto phenotype. Your critisim is valid in that case. GAs are usually used for static and closed problem solutions, like "there is one or many solutions, and we want to find one or some of them that fits these criteria." In a dynamic problem (such as survival in biology or game playing in AI) potential solutions are infinite and may change in time. Celular automata encodings, variable length phenotypes, evolution of genotype->phenotype mapping etc. are solutions to these. You also need a dynamic objective function, which suggests that even the "same" traits will have different effects over time and should be trated as such.
I didn't want to discuss with you just yet because of nonsense in the last paragraph, but once you do your homework about artifical evolution, I will gladly point out how GA concepts map onto natural evolution concepts.
FWIW someone already posted two quotes (from GPL FAQ I believe) about NDAs with GPL. It seems one can indeed put an NDA for development purposes but not for distribution. But it also seems that by NDAs, you can only put limits redistribution; anyone with binary access still has access to source.
If lindows makes this "beta program" the only way to access their software even after it is release quality, then one can definetly argue they are breaking GPL by not providing source. I guess we are not at that point yet.
Whether lindows case is more like this LinFlows case or the picture you present can only be determined by knowing and analysing license agreement between beta participants and lindows. Without this information the discussion is meaningless.
Obviously "distribution of software" mentioned in GPL is not something done internally; I can make modifications to any GPLed software, not release the final software but use it internally at my company with how-many-I-desire copies installed without breaking GPL. I don't see much difference between internal usage and an internal beta program. I may be off the base here, and the very fact that one can participate in beta program might make the process open, no matter how many NDAs people sign. But if the program is indeed closed and internal, he may withold any source until it is released.
'nuff said.
I understand that you have no knowlege about first wave of Turkish attack on Cyprus in 1974. We had all imported equipment mostly from USA and guess what, the minute we tried to use them against will of USA we started to see ghost ships and planes around, communications has been disrupted leading us to sink our own ship, navigation systems failed... Lesson learned and we build our own military industry but I think most third world countries still haven't taken that step. Exporting high tech military equipment makes USA more powerful, not less. Also don't forget military exports finance further research, if you stopped exporting them, it would have been easier for other to catch you.
One step at a time, Cmdr!
How come editors post offtopic and get away with it? I've been rtlbed (or was it rtbled) for that.
Apply every possible key with every possible encryption algorithm and you end up with nothing but every possible bit sequence that you know how to derive from that particular bit sequence. Some should make sense, but which one is the one you are looking for? Try decrypt 10000 0s with that, you will arrive a message, but that message will probably have nothing to do with my original message.
Theory 1: Everthing you observe is because I want you to observe that way. You cannot figure anything beyond because you neither have perception beyond that nor you can figure out what I like and in which ways I make you see things different than they actually are. You see, you also observe a "me" which is substantially different from real me. I can't tell you what reality is, it would be no fun. Even this post is partially true but it is as close to reality as you can/will ever manage; I'll see to that.
Theory 2: Whatever I say is internally consistent because I say so. Everthing is either something I would say, or not consistent with reality. You might think othewise time to time but you are obviously mistaken when you do that.
All true, but how does this apply in this case? If your theory is that astronout fails to remember how one correctly responds to falling balls in space because he has not learned that skill in that environment, that is also theorizing far more than data suggests. Also that theory will have hard time explaining why non-motor skills related to gravity does not suffer likewise in space.
OTOH, assuming having expectations about how the world and objects in it will behave without resorting to a native bias for that expectations can be easily justified. Under this assumption, astronouts inability to efficiently catch those balls simply results from failure to correctly foresee how objects will behave, gravity-wise. They learn slowly because of a negative interference from long term behavior they enjoyed on Earth. They relearn in Earth's gravity quickly because of the experience's deep roots from childhood.
But this is also a simple to understand layman's theory, requiring no expansive speech to express. So I understand why it is unpopular among us, the cognitive scientists.
If there is such things as RAM copies, how come use of authorized DVD players doesn't violate DMCA?
I don't take broken promises lightly. They never promised to supply us staroffice, but they did prmise that we will be treated equal. I want them to keep their promise, not fscking staroffice.
And they made SO6 free for higher class members. I have no problem with that either, I could have used SO6 (I'm still using OO 639c) but it is possible to buy it from Sun anyway. All members can buy pro suite at a discounted rate which has SO6. And those silver and gold members did pay more than me, so it is only natural they can profit more from their membership.
My only problem is that mandrake failed to keep their promise. They categorically stated that we would all be same, enjoy membership goodies no matter how much we payed. When I first saw it, I thought it was dumb to say. But they did say that. I didn't request to be treated same with other levels, even if I paid considerably less; it was their promise
Why should I trust mandrake now? Which promise they will break next? These are the real problems with their move. SO6 has nothing to do with discussion.
And another thing is, I just whine. Perhaps somebody will be enough pissed so they won't be satisfied with just whining. Do you think mdk has a chance in the court? How will the company survive if it is forced to make SO6 downloads free for every member?
I'm an engineer and I do quite a few things that would perhaps require man-years to explain with the current theories. I have never spent those man-years and it is possible that some scientist, after spending those man-years, will argue that what seems (their word) to happen is incompatible with theory. I have no problem with that. I produce, it works, nothing else matters. If those things fly, do your best to make use of that!
I'm willing to accept that some unknown process is causing those kites to float. I just don't see justification for that. I don't reject reality, I just don't buy a particular explanation about it. My problem with the hypothesis is basically:
I would say that a crude electrogravitic kite lifting an 18 gram payload is pretty extraordinary.
this. I don't think it is extraordinary at all. If those things can move out of Van Allen belt and athmosphere, it would be extraordinary. That means -whether utilizing reactionless force or not- they are feasible space engines. I didn't see enough controlled experiments to suggest that they will function in that environment though. As long as those things fly on earth, it doesn't matter how they do it; we have enough flying artifacts here already.