Yes, I was being simplistic in the differences between human and computer chessplayers, but the differences are very large. Anyhow, the point I was trying to make is that getting a computer program to play chess well isn't the same as understanding the game from a human perspective, i.e. what goes on in the mind of a human chessplayer. This is AFAIK the consensus in cognitive science nowadays.
For example, one consequence of this is that the knowledge used to construct a good computer chessplayer is not very useful in constructing a good human chessplayer.
Put another way: in the end, chess is completely deterministic. If you also put a time cap on it, then it is also finite (games that go on too long are ties). So there is a finite number of boards. If you know all of them, then you can do as well as is possible. Given enough processor power to calculate them, and enough space to save them (and speed to retrieve them later), you can have a great computer player - which is basically a retrieval engine. But this isn't the same as understanding what goes on in the mind of a human chessplayer.
This is why I don't like comparing things to language in the way parent poster did; I know of nothing else that cannot be learned later in life, and learned well.
There are some other things. For example, children born blind but that see later on in life (because they had something occluding their vision, that was removed by surgery) have problems with 3D vision. It is said that they are surprised that things 'get larger when they move closer', and IIRC never attain normal functionality. Studies have been done in animals to test this, and they found that indeed there is a 'critical period' for vision (this is much easier to test in animals than to see in humans, since we don't conduct such experiments on humans - the children born blind that see later are 'natural experiments', but not perfect ones).
The 'critical period' aspect of language is probably because of its innateness, and not its complexity. As Chomsky said, "it isn't a coincidence that all children in a household learn a language, while none of the pets do so."
Children actually learn language fairly slowly, when you think about it. Most children raised in environments that are saturated with well-spoken english still take 6-7 years before they speak without making many grammatical errors in the course of a conversation. An adult with some talent at languages, lots of free time, and access to native speakers ought to be able to speak like a reasonably well-educated adult within 2-3 years.
Well, the issue is that learning a primary language and learning a secondary language are different. My original post was a bit vague. I'll try to elaborate.
What you are saying is that it takes an adult less time to learn a secondary language than for a child to learn a primary one. This is true AFAIK. However,
1. Only children can learn a primary language, during the 'critical period'. Adults who never learned a language never will.
2. Children generally learn secondary languages faster than adults. For example, children taken to a foreign country often learn the local language faster than their parents.
Re:Partial credit
on
The Expert Mind
·
· Score: 4, Insightful
Chess isn't a good measure either. A COMPUTER can play chess. The rules and strategies are almost all worked out, so it takes only practice to learn them.
Software chessplayers can beat human ones, but they play completely differently. For example, human chessplayers see only a few moves ahead, while software chessplayers rely more on brute-force search to find good moves.
Computers beat humans at chess not because we understand chess, but because we found a way to make computers do it well, which is different.
Re:Partial credit
on
The Expert Mind
·
· Score: 5, Interesting
You can't be a chess master or genius mathematician or amazing athlete without some genetic preponderance toward intelligence or coordination or speed. This becomes extremely evident in bodybuilding; genetic makeup matters big time. Yes, I realize the article is focused on intellectual pursuits, but the same thing is still true.
So, to argue that intellectual experts are partially born, you compare them to a field where we know that being an expert is mostly born (bodybuilding)?
There are no studies showing a trainer taking a few average joes and getting them into the world championships of bodybuilding. But there are such examples in chess, as TFA states.
I remember learning about the "10-year theory" of genius in a graduate course in psychology (that it takes around 10 years of practice to make an expert, not innate talent). It was portrayed as a 'radical' theory in that it flew in the face of the common belief of innateness. But the evidence does support it.
The one area where the theory wasn't completely fleshed out in TFA, however, was the issue of age. While it is possible that nearly any child can be turned into a chess master with appropriate training and time, it isn't at all clear that the same is true for adults. Whether this is because adults have less time (or motivation), or because they are missing some biological advantage that children have, we don't know. But compare this to language: we know that children learn languages very fast during a 'critical period' of childhood. Children who don't learn a language at that age cannot learn one later in life. So perhaps there is a 'critical period' for being trained to be an expert at chess. We just don't know that yet (or didn't when I was taking the class 4 years ago).
One thing I couldn't tell from the article was [...]
It's odd that you have questions remaining after such a clear-cut article, which lacks no details. Consider the following quote from it:
According to MacWorld, "early reports from people claiming to use the software claim that certain parts of it 'feel' incomplete."
What could be more definite? First, this is 'according to MacWorld', so it's not direct knowledge. Second, these are 'early reports'. Third, these are claims from people claiming to use it (double use of 'claim' originating in TFA). Finally, they use the highly precise terminology of "'feel' incomplete".
After reading such an exact and meticulously-prepared document, I was left with no questions at all. Except perhaps why I read the entire thing.
This will be the first year that Python will be allowed along side the traditional TopCoder languages (Java, C++, C#, and VB). I think it will be very interesting to see how Python, which more often than not get can get more done with less code and less time, will compare to the heavier languages like C#, C++, and Java.
Yes, but read the fine print. Programs must run in under a certain amount of time. And they write that because of it's slower speed, Python may not be able to solve some of the problems. Python's strength is certainly in speed of coding and not execution, admittably, but it's still sad to me that a better solution wasn't found for this problem than "some problems may not be solvable".
Sorry for my English. Instead of "Olga (and similar sites) do not publish recordings of songs, which is a clear copyright violation", I should have had
Olga (and similar sites) do not publish recordings of songs, which would have been a clear copyright violation.
First, I gather from the tone of your reply that I may have offended you. After rereading my reply to you before, I apologize. I feel strongly about this issue and I went a bit overboard.
Well, OLGA supplies more than just lists of chords
I agree. This brings us back to my original post, where I said that the case of chord files is different than tabulature. A highly-accurate tabulature is high in information, even if not entirely complete. And perhaps this is indeed copyright infringement (but I am not sure). However, a large portion of the content is just chord files, which I believe is not infringement, as I claimed before (the information present in 4 chords isn't much).
You could also say, however, that you don't really need a script to learn a play, just a recording, but try telling that to the script publishers if you post one online. They're covered by copyright.
I would compare a list of chords to a very brief summary of a script. I believe that such summaries are legal to post online (e.g. Wikipedia does it). But, a highly accurate tabulature is something else, as I said before. I am less sure about that case.
The problem is that the lines can be blurry between the two cases (such as a very partial tabulature, which is the majority of content on Olga), so we may see them all lumped together as legal or not. It would be a sad thing (to me) if chord files were deemed copyright infringement because of that.
Olga (and similar sites) do not publish recordings of songs
They don't? Then how is it you can learn to play songs from them?
Actually, you can't! From a list of chords, you can't learn to play a song. They aren't meant to be used that way. The correct way to use such a file is to first know the song very well, and preferably have a recording of it.
As any guitar player will tell you, knowing the song is most of it; the list of chords is just an aid. From your knowledge of the song, you will know the rythym, speed, phrasing, points of emphasis, and so forth. Those don't appear in a chord file.
No, this is not apparent at all. The site owner of Olga says so, and I agree with him completely.
Olga (and similar sites) do not publish recordings of songs, which is a clear copyright violation. Nor do they publish lyrics (although they do sometimes). They primarily publish tab files (tabulature), which are specifications of where to place your fingers on a guitar (not even notes), and chord files, which are lists of chords. Now, to see how ridiculous the 'copyright' claim is, consider the case of chord files.
A typical chord file contains something like "G D Em C" - which are all the chords you need to play for quite a lot of rock/pop songs (up to modulation to a different key). A lot of others are covered by "G D C" (even simpler). There are only 6 basic chords on the guitar (in a specific key). Most songs use only a few of those (except for people like e.g. David Bowie, who uses dozens of chords in some songs). Basically, to claim copyright violation here, is to claim that "G D C" is copyrighted. But by which of the 1,000,000 songs that use it? It isn't unique in any way (unlike, say, lyrics or mp3s). Chord files (usually) only contain names of chords, not rythym or anything else. They are brief and nonunique in the extreme. To claim copyright violation would be amusing if it weren't sad.
The case of tab files is different, as these can be fairly specific to a song. However, even here it is far from clear that a copyright violation is being committed. A perfect, note-for-note transcription may seem to be an obvious copyright violation, but 99% of tabs are far from that. They are more like a guess or an interpretation of the song (for example, in nearly all cases they contain only notes, not durations of notes - and again, not even notes, but positions on the guitar).
As a guitar player who has enjoyed Olga for many years, this (repeat) development is sad, and I believe unjustified.
What matters is that small changes in the hardware make it possible to stop having to depend on costly, proprietary, and complex software--like that sold by VMware.
I am 100% in favor of cheap and open solutions. But I don't agree that this will soon be the case for virtualization. VMWare and the few other major vendors do a lot more than software virtualization of a CPU (which is all TFA was talking about). To have a complete virtualization solution, you need to also virtualize the rest of the hardware: storage, graphics, input/output, etc. In particular graphics is a serious issue (attaining hardware acceleration in a virtual environment safely), which from last I heard VMWare were working hard on.
Furthermore, Virtualization complements well with software that can migrate VMs (based on load or failure), and so forth. So, even if hardware CPU virtualization is to be desired - I agree with you on that - that won't suddenly make virtualization as a whole a simple task.
In the example, (1) the software company isn't to blame - they just put out a software product (in theory, anyone could make hardware to run it). (2) The hardware company isn't to blame - sure, they only run software by a specific company, but they do NOT sell software - and therefore cannot be held to a software license (which the GPL is). (3) A retailer that sells the two products together at a discount also isn't to blame - they just give a discount for a certain type of purchase. They would also sell the two products separately.
To be specific, the method by which the hardware manufacturer makes sure only certain software runs on his hardware can be done WITHOUT help from the software company. For example, the hardware manufacturer could buy the software product, create a signature from the binary, and then create hardware that only runs with that binary.
So, while absolutely true that a proprietary vendor can run the code checker on their code as well as an open source project, there is a huge difference when it comes to the customer/user of said software: with Open Source the user has the freedom to run such a tool over the source code themselves.
Actually, I would argue that it isn't just a freedom, it's a necessity. Having the source open means that wrongdoers can use bug-seeking programs to find exploits (presumably they have already been doing so for a while). So just to even things out, the Open Source community should scan them as well. This issue shouldn't be ignored.
Of course, closed-source programs are also being scanned by exploit-seeking software (it's not too hard to e.g. search for all calls to strcpy in a binary). So this isn't a 'new' problem with open-source projects. But, with the advantages of Open Source come a few risks, which we should deal with, as mentioned in the previous paragraph.
I agree that, morally, the hardware-seller is the one at fault. However, I don't see how to make them legally at fault. They sell a hardware product, containing no software (they just check a key when you do try to install software on it), so they aren't bound by a software license like the GPL. The GPL is based on copyright law for its power, but here they aren't copying software, so they can't be in violation.
The end-user who installs the software IS copying it. However, there is no redistribution, so no problem there (this is the same as making a personal backup copy, i.e. fair use, I would assume).
You stop users from changing the code, which is the whole purpose of the GPL.
You do that intentionally, not out of technical need.
So you remove the freedom to run any code.
So you have no license to run GPL-code.
But the point of my example was that it is NOT me that does the removal of freedom. I put out software. Some (not even all) of the hardware manufacturers prevent unsigned code from running. How am I to blame for that?
IANAL also, of course:). But something seems wrong to me here. Maybe I'm not putting my finger on it yet, though, or not being clear enough in my arguments (apologies if so).
There are grey areas, but legal questions are full of grey areas that need to be decided on a case-by-case basis. The Tivo situation, and the situation you describe, are both clear-cut cases by the text of the GPL.
I see your point, and I hope that this will not turn out to be a problem for the GPL3. I suspect, however, that Tivo-type companies will see the grey area as grey enough to be able to be abused. Again, I hope I am wrong.
I do have a more fundamental question, however. Say that there is a single software manufacturer, Soft, and three hardware manufacturers, HardA, HardB and HardC. HardA runs only signed code, HardB and HardC run any code. Say that each hardware company has 1/3 of the market. So Soft is in compliance with the GPL3.
Now, suppose that HardB and HardC vanish from the market - suddenly, only HardA supplies hardware, and it requires signed code. Is Soft suddenly in violation of the GPL3? If so, isn't this odd - that events unrelated to and uncontrollable by Soft cause them to be in violation of the GPL3? What if tomorrow HardB goes back in business, are Soft in compliance once more?
Something about the concept of "recommended or principle context of use" seems 'wrong', if that context of use is not in control of the software manufacturer themselves.
Ok, I read your other post; however, I still don't agree. Quoting your main argument from there:
- If Ovit's software runs only on the Tivo hardware, then the signing key is "necessary to... execute modified versions in the... recommended or principle context of use," and Ovit is guilty of copyright infringement (since the GPLv3 does not apply to their redistribution).
- If Ovit's software runs on other hardware than the Tivo (with "all the same functionality"), then their software is legal by the terms of the GPLv3, which is correct, because they really are making general-purpose media center software, and the lack of freedom on Tivo hardware is merely an irritation rather than a menace.
What if Ovit's software DOES run on other hardware, but for some reason, in practice, that is not a good solution. For example, some third party hardware manufacturer can make hardware that works like Tivo, but doesn't check signed code (so Ovit's software will run), BUT the hardware is expensive/unreliable/etc (yet, it still "provides all the same functionality", in a legal sense). So Ovit are complying with the GPL3 (you can run modified code somewhere), but in practice no one will.
The GPL3 says the requirement is to be able to run modified code "in the recommended or principal context of use". Say Ovit even recommend the 3rd party hardware manufacturer and not Tivo. Or, they could even BE that manufacturer, making hardware that runs modified code just fine. But since Tivo is far superior, 99.9% of people buy Tivo + Ovit software, and cannot run modified code on it, but legally all is fine.
Since the GPL3 can't prevent this, I presume that this is exactly what Tivo et al will be doing, making this aspect of the GPL3 powerless (as far as I can tell).
Besides Xen, a few other interesting tidbits appear in the article, but are missing from the summary (and, were also missing in the post on Digg... suspiciously).
1. All desktops in Novell have been using OpenOffice for a year now.
2. 80% of desktops in Novell now use Linux (I presume the remainder use Windows).
3. The article mentions some explanations for the recent personell changes in Novell. Not much content, though, just "we are in a different place now and need different people" (where have I heard that before).
If he made that computer, and required that his end users download a kernel.org kernel signed by Linus in order for his computer to operate, he would be in the clear, as would his end users (since they aren't copying any GPLed work, the provisions don't have to apply). This situation would make RMS slightly unhappy, since the end user isn't free to modify his computer's software, but it's perfectly legal according to the terms of the GPL v3.
Of course, the DRM provisions aren't designed to attack that farfetched example; they're designed to counter the much more plausible example of Tivo-style DRMization of GPLed works
But all Tivo has to do is exactly what you stated in your post, in order to comply with the GPL3 (separate the software from the hardware). This won't be 'farfetched' at all, it will become commonplace. Circumventing the GPL3's clause will be a minor inconvenience to people like Tivo. So, why bother with all of this effort? It seems (sadly) futile to me.
But hardware isn't the same as software in this area, I don't think. Perhaps I'm wrong, but here is how I see the problem:
Say there are two people making hardware that can run my software. One of them runs only digitally signed software, and I have my code signed by them. The other hardware vendor will run any code.
I, the software vendor, am complying with the GPL3. Correct me if you disagree at this point. But as I see it, I give out my product with accompanying source code, and people can run it on hardware from either hardware vendor. If they want to modify it, they can use hardware by the second vendor. You can't blame me for paying a modest fee to the first hardware vendor to get my code digitally signed, can you? (I might get more sales that way)
Now, what happens if the second hardware vendor goes out of business? Am I all of a sudden in violation of the GPL3, because now there is no hardware to run modified versions of my code?
If you say yes to this, but thought I was complying before, then the GPL violation can be circumvented by having a THIRD party, which makes hardware that can run modified versions of my software - but that hardware is very expensive/poorly made/produced in small numbers, so in practice no-one buys it.
There is no linking here. Company A sells hardware, Company B sells software. Separately.
To reiterate, the software works on the hardware only if signed with a key, known only to Company A, the hardware manufacturer. Anyone is free to reach an arrangement with Company A in order to get their code signed by them. As it happens - by 'chance' - the only one that actually does so is the software manufacturer 'Company B'. This could be because Company A charges a ridiculous sum of money from everyone but Company B, for example.
The end result is that there is NO freedom to modify and run Company B's code on Company A's hardware. But there is no violation of the GPL3. The hardware and software are two 'separate' products, by two different manufacturers.
The argument is that since it is impossible to prevent this from happening, there is no reason to even try. As I said in the previous post, I hope I am wrong here, please correct me.
It seems to me this loophole is already closed by (all of) the drafts of GPLv3. It does not matter which key owned by whom, the KEY TEST (sorry, can't resist) is whether a modified version will run. If TivoV3 uses Linus' signature as DRM, then TivoV3 must give the user a way to sign using Linus' key; which means TivoV3 would be stuck.
This doesn't answer the issue raised by the grandparent post (which has also been raised before in Slashdot and elsewhere, and to which I have yet to see a good response).
Put simply:
Company A releases a 'Tivo-type' hardware device. Only the hardware. Now, the hardware runs only digitally signed software with some secret key. But since this is only hardware, no problem there.
Company B releases a software package. This software package happens to work perfectly on Company A's hardware, as it happens to have the right digital key signature (and recognize the hardware correctly). But, this is only software (and distributed with source, etc.), so, no problem there.
Now, some retailer offers a 'special discount' if you buy Company A's hardware and Company B's software together. When you get home, all you need to do is stick a CD into the hardware device, and everything installs and is ready to run. But - you cannot modify the software and run it on the hardware. Yet, no GPL3 violation has been made.
why would Microsoft license its WMV technologies to a Linux distribution? I can understand someone like Sun licensing Java, or ATI/nVidia licensing their drivers, but Microsoft? Why would they want to make it easier for a Linux distribution to compete with Windows[?]
Because (1) they have had enough 'fun' with monopoly-related investigations and the ensuing fines, and (2) they probably (sadly, realistically) calculate the risk of Linspire stealing serious market share from Windows at about 0.1% or less, and (3) they make money off of licensing WMV, so why the hell not.
Slashdot Mirror
Yes, I was being simplistic in the differences between human and computer chessplayers, but the differences are very large. Anyhow, the point I was trying to make is that getting a computer program to play chess well isn't the same as understanding the game from a human perspective, i.e. what goes on in the mind of a human chessplayer. This is AFAIK the consensus in cognitive science nowadays.
For example, one consequence of this is that the knowledge used to construct a good computer chessplayer is not very useful in constructing a good human chessplayer.
Put another way: in the end, chess is completely deterministic. If you also put a time cap on it, then it is also finite (games that go on too long are ties). So there is a finite number of boards. If you know all of them, then you can do as well as is possible. Given enough processor power to calculate them, and enough space to save them (and speed to retrieve them later), you can have a great computer player - which is basically a retrieval engine. But this isn't the same as understanding what goes on in the mind of a human chessplayer.
This is why I don't like comparing things to language in the way parent poster did; I know of nothing else that cannot be learned later in life, and learned well.
There are some other things. For example, children born blind but that see later on in life (because they had something occluding their vision, that was removed by surgery) have problems with 3D vision. It is said that they are surprised that things 'get larger when they move closer', and IIRC never attain normal functionality. Studies have been done in animals to test this, and they found that indeed there is a 'critical period' for vision (this is much easier to test in animals than to see in humans, since we don't conduct such experiments on humans - the children born blind that see later are 'natural experiments', but not perfect ones).
The 'critical period' aspect of language is probably because of its innateness, and not its complexity. As Chomsky said, "it isn't a coincidence that all children in a household learn a language, while none of the pets do so."
Children actually learn language fairly slowly, when you think about it. Most children raised in environments that are saturated with well-spoken english still take 6-7 years before they speak without making many grammatical errors in the course of a conversation. An adult with some talent at languages, lots of free time, and access to native speakers ought to be able to speak like a reasonably well-educated adult within 2-3 years.
Well, the issue is that learning a primary language and learning a secondary language are different. My original post was a bit vague. I'll try to elaborate.
What you are saying is that it takes an adult less time to learn a secondary language than for a child to learn a primary one. This is true AFAIK. However,
1. Only children can learn a primary language, during the 'critical period'. Adults who never learned a language never will.
2. Children generally learn secondary languages faster than adults. For example, children taken to a foreign country often learn the local language faster than their parents.
Chess isn't a good measure either. A COMPUTER can play chess. The rules and strategies are almost all worked out, so it takes only practice to learn them.
Software chessplayers can beat human ones, but they play completely differently. For example, human chessplayers see only a few moves ahead, while software chessplayers rely more on brute-force search to find good moves.
Computers beat humans at chess not because we understand chess, but because we found a way to make computers do it well, which is different.
You can't be a chess master or genius mathematician or amazing athlete without some genetic preponderance toward intelligence or coordination or speed. This becomes extremely evident in bodybuilding; genetic makeup matters big time. Yes, I realize the article is focused on intellectual pursuits, but the same thing is still true.
So, to argue that intellectual experts are partially born, you compare them to a field where we know that being an expert is mostly born (bodybuilding)?
There are no studies showing a trainer taking a few average joes and getting them into the world championships of bodybuilding. But there are such examples in chess, as TFA states.
I remember learning about the "10-year theory" of genius in a graduate course in psychology (that it takes around 10 years of practice to make an expert, not innate talent). It was portrayed as a 'radical' theory in that it flew in the face of the common belief of innateness. But the evidence does support it.
The one area where the theory wasn't completely fleshed out in TFA, however, was the issue of age. While it is possible that nearly any child can be turned into a chess master with appropriate training and time, it isn't at all clear that the same is true for adults. Whether this is because adults have less time (or motivation), or because they are missing some biological advantage that children have, we don't know. But compare this to language: we know that children learn languages very fast during a 'critical period' of childhood. Children who don't learn a language at that age cannot learn one later in life. So perhaps there is a 'critical period' for being trained to be an expert at chess. We just don't know that yet (or didn't when I was taking the class 4 years ago).
One thing I couldn't tell from the article was [...]
It's odd that you have questions remaining after such a clear-cut article, which lacks no details. Consider the following quote from it:
According to MacWorld, "early reports from people claiming to use the software claim that certain parts of it 'feel' incomplete."
What could be more definite? First, this is 'according to MacWorld', so it's not direct knowledge. Second, these are 'early reports'. Third, these are claims from people claiming to use it (double use of 'claim' originating in TFA). Finally, they use the highly precise terminology of "'feel' incomplete".
After reading such an exact and meticulously-prepared document, I was left with no questions at all. Except perhaps why I read the entire thing.
This will be the first year that Python will be allowed along side the traditional TopCoder languages (Java, C++, C#, and VB). I think it will be very interesting to see how Python, which more often than not get can get more done with less code and less time, will compare to the heavier languages like C#, C++, and Java.
Yes, but read the fine print. Programs must run in under a certain amount of time. And they write that because of it's slower speed, Python may not be able to solve some of the problems. Python's strength is certainly in speed of coding and not execution, admittably, but it's still sad to me that a better solution wasn't found for this problem than "some problems may not be solvable".
Sorry for my English. Instead of "Olga (and similar sites) do not publish recordings of songs, which is a clear copyright violation", I should have had
Olga (and similar sites) do not publish recordings of songs, which would have been a clear copyright violation.
Some chord files contain lyrics, just so guitarists won't need to search for a lyrics site containing them.
My point was that most chord files don't contain lyrics.
First, I gather from the tone of your reply that I may have offended you. After rereading my reply to you before, I apologize. I feel strongly about this issue and I went a bit overboard.
Well, OLGA supplies more than just lists of chords
I agree. This brings us back to my original post, where I said that the case of chord files is different than tabulature. A highly-accurate tabulature is high in information, even if not entirely complete. And perhaps this is indeed copyright infringement (but I am not sure). However, a large portion of the content is just chord files, which I believe is not infringement, as I claimed before (the information present in 4 chords isn't much).
You could also say, however, that you don't really need a script to learn a play, just a recording, but try telling that to the script publishers if you post one online. They're covered by copyright.
I would compare a list of chords to a very brief summary of a script. I believe that such summaries are legal to post online (e.g. Wikipedia does it). But, a highly accurate tabulature is something else, as I said before. I am less sure about that case.
The problem is that the lines can be blurry between the two cases (such as a very partial tabulature, which is the majority of content on Olga), so we may see them all lumped together as legal or not. It would be a sad thing (to me) if chord files were deemed copyright infringement because of that.
Olga (and similar sites) do not publish recordings of songs
They don't? Then how is it you can learn to play songs from them?
Actually, you can't! From a list of chords, you can't learn to play a song. They aren't meant to be used that way. The correct way to use such a file is to first know the song very well, and preferably have a recording of it.
As any guitar player will tell you, knowing the song is most of it; the list of chords is just an aid. From your knowledge of the song, you will know the rythym, speed, phrasing, points of emphasis, and so forth. Those don't appear in a chord file.
Yes, OLGA has (apparently) broken the law.
No, this is not apparent at all. The site owner of Olga says so, and I agree with him completely.
Olga (and similar sites) do not publish recordings of songs, which is a clear copyright violation. Nor do they publish lyrics (although they do sometimes). They primarily publish tab files (tabulature), which are specifications of where to place your fingers on a guitar (not even notes), and chord files, which are lists of chords. Now, to see how ridiculous the 'copyright' claim is, consider the case of chord files.
A typical chord file contains something like "G D Em C" - which are all the chords you need to play for quite a lot of rock/pop songs (up to modulation to a different key). A lot of others are covered by "G D C" (even simpler). There are only 6 basic chords on the guitar (in a specific key). Most songs use only a few of those (except for people like e.g. David Bowie, who uses dozens of chords in some songs). Basically, to claim copyright violation here, is to claim that "G D C" is copyrighted. But by which of the 1,000,000 songs that use it? It isn't unique in any way (unlike, say, lyrics or mp3s). Chord files (usually) only contain names of chords, not rythym or anything else. They are brief and nonunique in the extreme. To claim copyright violation would be amusing if it weren't sad.
The case of tab files is different, as these can be fairly specific to a song. However, even here it is far from clear that a copyright violation is being committed. A perfect, note-for-note transcription may seem to be an obvious copyright violation, but 99% of tabs are far from that. They are more like a guess or an interpretation of the song (for example, in nearly all cases they contain only notes, not durations of notes - and again, not even notes, but positions on the guitar).
As a guitar player who has enjoyed Olga for many years, this (repeat) development is sad, and I believe unjustified.
What matters is that small changes in the hardware make it possible to stop having to depend on costly, proprietary, and complex software--like that sold by VMware.
I am 100% in favor of cheap and open solutions. But I don't agree that this will soon be the case for virtualization. VMWare and the few other major vendors do a lot more than software virtualization of a CPU (which is all TFA was talking about). To have a complete virtualization solution, you need to also virtualize the rest of the hardware: storage, graphics, input/output, etc. In particular graphics is a serious issue (attaining hardware acceleration in a virtual environment safely), which from last I heard VMWare were working hard on.
Furthermore, Virtualization complements well with software that can migrate VMs (based on load or failure), and so forth. So, even if hardware CPU virtualization is to be desired - I agree with you on that - that won't suddenly make virtualization as a whole a simple task.
In the example, (1) the software company isn't to blame - they just put out a software product (in theory, anyone could make hardware to run it). (2) The hardware company isn't to blame - sure, they only run software by a specific company, but they do NOT sell software - and therefore cannot be held to a software license (which the GPL is). (3) A retailer that sells the two products together at a discount also isn't to blame - they just give a discount for a certain type of purchase. They would also sell the two products separately.
To be specific, the method by which the hardware manufacturer makes sure only certain software runs on his hardware can be done WITHOUT help from the software company. For example, the hardware manufacturer could buy the software product, create a signature from the binary, and then create hardware that only runs with that binary.
Tell me where you disagree.
So, while absolutely true that a proprietary vendor can run the code checker on their code as well as an open source project, there is a huge difference when it comes to the customer/user of said software: with Open Source the user has the freedom to run such a tool over the source code themselves.
Actually, I would argue that it isn't just a freedom, it's a necessity. Having the source open means that wrongdoers can use bug-seeking programs to find exploits (presumably they have already been doing so for a while). So just to even things out, the Open Source community should scan them as well. This issue shouldn't be ignored.
Of course, closed-source programs are also being scanned by exploit-seeking software (it's not too hard to e.g. search for all calls to strcpy in a binary). So this isn't a 'new' problem with open-source projects. But, with the advantages of Open Source come a few risks, which we should deal with, as mentioned in the previous paragraph.
I agree that, morally, the hardware-seller is the one at fault. However, I don't see how to make them legally at fault. They sell a hardware product, containing no software (they just check a key when you do try to install software on it), so they aren't bound by a software license like the GPL. The GPL is based on copyright law for its power, but here they aren't copying software, so they can't be in violation.
The end-user who installs the software IS copying it. However, there is no redistribution, so no problem there (this is the same as making a personal backup copy, i.e. fair use, I would assume).
You stop users from changing the code, which is the whole purpose of the GPL.
:). But something seems wrong to me here. Maybe I'm not putting my finger on it yet, though, or not being clear enough in my arguments (apologies if so).
You do that intentionally, not out of technical need.
So you remove the freedom to run any code.
So you have no license to run GPL-code.
But the point of my example was that it is NOT me that does the removal of freedom. I put out software. Some (not even all) of the hardware manufacturers prevent unsigned code from running. How am I to blame for that?
IANAL also, of course
There are grey areas, but legal questions are full of grey areas that need to be decided on a case-by-case basis. The Tivo situation, and the situation you describe, are both clear-cut cases by the text of the GPL.
I see your point, and I hope that this will not turn out to be a problem for the GPL3. I suspect, however, that Tivo-type companies will see the grey area as grey enough to be able to be abused. Again, I hope I am wrong.
I do have a more fundamental question, however. Say that there is a single software manufacturer, Soft, and three hardware manufacturers, HardA, HardB and HardC. HardA runs only signed code, HardB and HardC run any code. Say that each hardware company has 1/3 of the market. So Soft is in compliance with the GPL3.
Now, suppose that HardB and HardC vanish from the market - suddenly, only HardA supplies hardware, and it requires signed code. Is Soft suddenly in violation of the GPL3? If so, isn't this odd - that events unrelated to and uncontrollable by Soft cause them to be in violation of the GPL3? What if tomorrow HardB goes back in business, are Soft in compliance once more?
Something about the concept of "recommended or principle context of use" seems 'wrong', if that context of use is not in control of the software manufacturer themselves.
Ok, I read your other post; however, I still don't agree. Quoting your main argument from there:
... execute modified versions in the ... recommended or principle context of use," and Ovit is guilty of copyright infringement (since the GPLv3 does not apply to their redistribution).
- If Ovit's software runs only on the Tivo hardware, then the signing key is "necessary to
- If Ovit's software runs on other hardware than the Tivo (with "all the same functionality"), then their software is legal by the terms of the GPLv3, which is correct, because they really are making general-purpose media center software, and the lack of freedom on Tivo hardware is merely an irritation rather than a menace.
What if Ovit's software DOES run on other hardware, but for some reason, in practice, that is not a good solution. For example, some third party hardware manufacturer can make hardware that works like Tivo, but doesn't check signed code (so Ovit's software will run), BUT the hardware is expensive/unreliable/etc (yet, it still "provides all the same functionality", in a legal sense). So Ovit are complying with the GPL3 (you can run modified code somewhere), but in practice no one will.
The GPL3 says the requirement is to be able to run modified code "in the recommended or principal context of use". Say Ovit even recommend the 3rd party hardware manufacturer and not Tivo. Or, they could even BE that manufacturer, making hardware that runs modified code just fine. But since Tivo is far superior, 99.9% of people buy Tivo + Ovit software, and cannot run modified code on it, but legally all is fine.
Since the GPL3 can't prevent this, I presume that this is exactly what Tivo et al will be doing, making this aspect of the GPL3 powerless (as far as I can tell).
Besides Xen, a few other interesting tidbits appear in the article, but are missing from the summary (and, were also missing in the post on Digg... suspiciously).
1. All desktops in Novell have been using OpenOffice for a year now.
2. 80% of desktops in Novell now use Linux (I presume the remainder use Windows).
3. The article mentions some explanations for the recent personell changes in Novell. Not much content, though, just "we are in a different place now and need different people" (where have I heard that before).
If he made that computer, and required that his end users download a kernel.org kernel signed by Linus in order for his computer to operate, he would be in the clear, as would his end users (since they aren't copying any GPLed work, the provisions don't have to apply). This situation would make RMS slightly unhappy, since the end user isn't free to modify his computer's software, but it's perfectly legal according to the terms of the GPL v3.
Of course, the DRM provisions aren't designed to attack that farfetched example; they're designed to counter the much more plausible example of Tivo-style DRMization of GPLed works
But all Tivo has to do is exactly what you stated in your post, in order to comply with the GPL3 (separate the software from the hardware). This won't be 'farfetched' at all, it will become commonplace. Circumventing the GPL3's clause will be a minor inconvenience to people like Tivo. So, why bother with all of this effort? It seems (sadly) futile to me.
But hardware isn't the same as software in this area, I don't think. Perhaps I'm wrong, but here is how I see the problem:
Say there are two people making hardware that can run my software. One of them runs only digitally signed software, and I have my code signed by them. The other hardware vendor will run any code.
I, the software vendor, am complying with the GPL3. Correct me if you disagree at this point. But as I see it, I give out my product with accompanying source code, and people can run it on hardware from either hardware vendor. If they want to modify it, they can use hardware by the second vendor. You can't blame me for paying a modest fee to the first hardware vendor to get my code digitally signed, can you? (I might get more sales that way)
Now, what happens if the second hardware vendor goes out of business? Am I all of a sudden in violation of the GPL3, because now there is no hardware to run modified versions of my code?
If you say yes to this, but thought I was complying before, then the GPL violation can be circumvented by having a THIRD party, which makes hardware that can run modified versions of my software - but that hardware is very expensive/poorly made/produced in small numbers, so in practice no-one buys it.
There is no linking here. Company A sells hardware, Company B sells software. Separately.
To reiterate, the software works on the hardware only if signed with a key, known only to Company A, the hardware manufacturer. Anyone is free to reach an arrangement with Company A in order to get their code signed by them. As it happens - by 'chance' - the only one that actually does so is the software manufacturer 'Company B'. This could be because Company A charges a ridiculous sum of money from everyone but Company B, for example.
The end result is that there is NO freedom to modify and run Company B's code on Company A's hardware. But there is no violation of the GPL3. The hardware and software are two 'separate' products, by two different manufacturers.
The argument is that since it is impossible to prevent this from happening, there is no reason to even try. As I said in the previous post, I hope I am wrong here, please correct me.
It seems to me this loophole is already closed by (all of) the drafts of GPLv3. It does not matter which key owned by whom, the KEY TEST (sorry, can't resist) is whether a modified version will run. If TivoV3 uses Linus' signature as DRM, then TivoV3 must give the user a way to sign using Linus' key; which means TivoV3 would be stuck.
This doesn't answer the issue raised by the grandparent post (which has also been raised before in Slashdot and elsewhere, and to which I have yet to see a good response).
Put simply:
Company A releases a 'Tivo-type' hardware device. Only the hardware. Now, the hardware runs only digitally signed software with some secret key. But since this is only hardware, no problem there.
Company B releases a software package. This software package happens to work perfectly on Company A's hardware, as it happens to have the right digital key signature (and recognize the hardware correctly). But, this is only software (and distributed with source, etc.), so, no problem there.
Now, some retailer offers a 'special discount' if you buy Company A's hardware and Company B's software together. When you get home, all you need to do is stick a CD into the hardware device, and everything installs and is ready to run. But - you cannot modify the software and run it on the hardware. Yet, no GPL3 violation has been made.
I hope I am wrong about this. Please correct me.
why would Microsoft license its WMV technologies to a Linux distribution? I can understand someone like Sun licensing Java, or ATI/nVidia licensing their drivers, but Microsoft? Why would they want to make it easier for a Linux distribution to compete with Windows[?]
Because (1) they have had enough 'fun' with monopoly-related investigations and the ensuing fines, and (2) they probably (sadly, realistically) calculate the risk of Linspire stealing serious market share from Windows at about 0.1% or less, and (3) they make money off of licensing WMV, so why the hell not.