"Red Hat Linux 9 was the last in the line. Instead of being "Red Hat Linux 10" it's going to be "Fedora Linux 1[.0]" when it's released within the next few weeks/months."
From reading their web pages, it certainly sounds like that is more or less the plan. It seems that RH wants to drop the consumer version of their distro. This amply clear from the packages that have disappeared in RH8 and RH9. Considering the hacker/hobbyist base of Linux, I was shocked to see them dropping mature popular window managers (fvwm et al), and classics like xtetris and xevil, as well as UNIX staples like fortune.
In all cases, it is because these programs conflict with the goal of selling the Redhat distro as a business desktop system, with minimum variations between installations and nothing "non-professional". RH employees have said this on mailing lists and in bugzilla comments.
They don't care about those of us that go to Fry's and buy their boxed sets. They need the businesses that will buy installation and support for 10,000 seats.
The obvious possibility is that some of the investors buying stock do not expect any return; they are in effect tossing money to the SCO execs and increasing SCO's credibility by keeping the stock high.
I have also read that there are discussions on stock boards where people are guessing that SCO will sell licenses to X percent of a bazillion Linux users at $1400 a pop. This gives a huge income for a company worth $250M, enticing day-trader types, unless you know enough to realize that the most likely value for X is 0%.
"This is the current state of the Linux community."
No it isn't. I am on a local LUG mailing list, and people are politely helping newbies all the time, going out of their way to explain things that weren't even asked, just in case it might help.
"Not only the case on Slashdot, but go to any IRC help channel and you'll find the same the majority of the time."
IRC and/. were not exactly designed for thoughtful interaction.
"Uh, maybe I'm missing something here. Why would you not want a customer to see all the data associated with his server."
You have two customers, X and Y, and you do extensive monitoring. If you tell X everything you know, then X knows what monitoring you are doing and can make the reasonable guess that you are monitoring Y in exactly the same way. This is information that would make an attack on Y easier.
"A lot of people don't need rock solid though, and don't need to pay for it."
Even five years ago, there was a large gap in performance between a decent PC and a cheap workstation. If you needed more than a what a PC could offer, you had to jump up to a workstation/server, which means people were routinely buying much beefier hardware than they needed. This gap has closed. You can buy the hardware that is right for your needs, and this translates into lower sales for sellers of high end hardware.
"May I suggest an even more radical solution? Market research! Get a large group of windows users, give them a Linux desktop and tell them to complain and make requests!"
Market research like this has been done. The results are that the windows users want the Linux system to act exactly like their Windows machines, because that is what they are used to. People don't know what will be better or easier to use or most especially make them more productive.
Thus you get the abhorrent tangling of the issues of useability and marketability. If you want to sell software, you want people to see a shiny, non-threatening interface. This has nothing to do with whether the software is actually good at getting things done. It is very difficult for market research to distinguish these. And if you want a shiny, non-threatening interface, use Windows. That's what it's there for, and there's nothing wrong with that.
"Finding large mammals previously unknown to science"
Reading comprehension, please. I specifically said that I didn't care about when the species became known to western science.
As I said, a quick web search shows that the Vu Quang was well known to the locals, but it was unknown to western science. This has nothing to do with any search for bigfoot in the US. There are not piles of bigfoot bones and pelts sitting around in Oregon, unknown to scientists because nobody thought it worth mentioning to anyone. That is the comparison you have to make. You can't expect villagers in southeast Asia to go searching through European zoological literature to see if some of the critters haven't been described there. On the other hand, there is nobody in the Pacific northwest who wouldn't go screaming to the media if he found some concrete evidence of a bigfoot.
Ah, I just found the page you cut-and-pasted. Plagiarist.
"These are, at heart, text matching algorithms, which are easily defeated and of little relevance."
It is of extreme relevance, from a legal standpoint. If the code is not literally copied unchanged, then someone claiming copyright violation has to prove that the aspect of the material copied is protected by copyright. For example, SCO would have to prove that some snippet they claim was copied from SysV and obfuscated is sufficiently different from public domain ancient Unix code that it deserves copyright protection.
This sort of test is not invoked when direct, verbatim copying occurs. So there is a big legal difference between cut-and-paste jobs and obfuscated code.
There is also the aspect that Linux kernel coders are not, for the most part, college freshmen. Kernel coders would not have wasted time obfuscating things they could more easily have written from scratch. If they were copying to save time, they would have copied verbatim. As far as trivial changes such as reindenting, shred has the option to ignore whitespace differences.
You are right that this doesn't prove anything. All it does is point out parts of the code that deserve more scrutiny by the human coders.
"ome cryptozoological successes: -Acionyx rex, a giant cheetah, in 1873 -Tratratratra, a giant lemur, in the 1800s -Chacoan peccary -Vu Quang ox, an ox with antelope antlers, 1993...the list can go on.
Oh, do go on. Discoveries in the 1800's aren't very interesting, though, as communications and travel were nothing compared to today. After all, sequoias weren't discoved until the mid-1800's, and they are 200 feet tall and don't run around hiding. So stick to the late twentieth century.
Now lets look at the Vu Quang. I did a quick google. The species was not discovered then, it was merely "discovered" by western science. Not the same thing at all, as the locals knew about them.
So how about some examples of large mammalian species found in the second half of the twentieth century that were previously unknown to any humans. That would be an interesting list. If there is anything on it.
Worth noting that stuff insured with the USPS is insured to arrive, but not at any particular time. I saw a guy going ballistic in the post office once because someone had mailed him airplane tickets. They arrived too late for him to use them, but they did arrive, so the insurance paid nothing.
"When he used the wireless key to unlock his car, it also unlocked another car two to three spots over."
Are mechanical car locks any better? One time I unlocked and opened the door to my car. It didn't feel quite right as it unlocked, so I looked again. I had unlocked the door using the key to a different car. The cars are from different manufacturers, but it looks like they use the same blanks for the keys.
Re:Doing their work for them
on
Back To SCO
·
· Score: 1
"But my biggest concern is that at 7.5 cents per line of alleged copyright violation, even if the community ultimately ID's 100 lines of code inadvertantly included in Linux, it could mean SCO seeking $7.50 per license, and with the current copyright law they could then go after significant penalties and legal costs far in excess of that"
Huh? I have been reading pretty thoroughly about SCO stuff, and this is the first I have seen mention of a fine at 7.5 cents per line per copy. Googling turns up zilch. Could you provide a reference?
"There are hundreds of ways to do a single thing, and if the GNU/Linux took ideas from the SCO kernel, SCO may be as eligible for compensation as if it were directly copied from SCO."
This is true. But proving direct verbatim copying is straightforward. It is much harder to prove copyright infringement when the copying is not verbatim, because the plaintiff has to justify that the copied material is copyrightable.
If I look at your copyrighted code and retype it exactly into my code, then all that has to be shown is that I copied it (assuming the code is not GPLed or whatever). If I look at your copyrighted code and write a very similar routine for myself, you have to prove that I was looking at your code and you have to prove that whatever I took is copyrightable.
There is no hard line between what is copying and what is legal use of the ideas embodied in the code. It is for the courts to decide on the merits of each case.
"So, yes, SGI could have copied the 32V code, but they didn't."
This may be important when damages are considered. If code was copied from SysV but is essentially identical to code in the public domain, a judge may rule, "Defendent is guilty of violating copyrights, and we award to the plaintiff damages of 37 cents plus whatever loose change they can find under the cushions in the jury box." Of course, SCO would have to sue SGI for this, which they haven't done yet. The fact that this code was in Itanium specific code and was probably never run by more than a handful of people doesn't help SCO's quest for damages either.
"McBride is no fool. He is actually exploiting weaknesses in the US media, business structure and investment industries, in order to massage the SCO share price.
Indeed, SCOX is up 10% today as of 3PM EDT under heavy trading. Not quite as much as I had expected, though. I figured it would go up $2 after the release of this "letter", but it's only up a $1.50.
This was possibly the single genuinely new item in the diatribe. I consider it a real possibility that code ended up in the kernel that was in books or on web sites authorized with a copyright notice permitting copying for non-commercial use only. Such code could not be legally placed in the kernel because the "non-commercial use" clause conflicts with the GPL.
But if this were true, they would not have to spin it quite so hard. "Books and Internet sites intended and authorized for the purpose of teaching and other non-commercial use cannot be copied for commercial use." Why does he use the word "intendend"? Intention has nothing to do with it. What matters is the copyright notices and any authorization. You can't legally copy an entire book that is copyrighted, but what about code snippets in a book? Numerical Recipes for example claims that that each routine is individually covered by copyrights, but are the rights under fair use different in the absence of such specific claims?
It sounds as if they are going to say that a simple statement like "This material was made available in hopes that someone could learn something from it" is equivalent to "Any commercial use of any part of the code shown here is forbidden." They could also be planning to claim copyright to bits of code that are in SysV but also present in ancient UNIX no longer under copyright, for which they need to argue that the code was never put in the public domain and the copyrights to the pieces has never lapsed.
Time will tell.
"GNU/Linux. We still have old distributions available."
Which are of little value if you care about security. Security is what drives the upgrade train. Old programs have known security holes. New programs require new compilers and libraries. If you use your machine on a network or over a modem, you can't go with old distributions.
"It's simply not feasible to hold back development just for the sake of old machines running at 1/20 or even 1/50 the clock speed of today. Such machines can't handle demanding modern applications."
Like emacs, LaTeX, X, squid, ppp, ftpd, sshd, xfig, xv, and mozilla? Sorry, they run great on a P200 (well, the lizard is marginal).
"What they are trying to say is that you cannot unwittingly or unknowingly license your code under the GPL; you must do so intentionally."
I am sure that most people would agree with this statement. But my point is that they are not honestly making their case about this, they are throwing out true but irrelevent legalistic statements to make their argument appear more convincing to the uncareful reader. They know that transfer of copyright has nothing to do with putting code under the GPL. They want people to think, "Hmm, SCO never signed anything transfering their copyrights, so their code can't be under the GPL," despite the lack of legal or logical connection between these two statements.
This is why we hate them. Well, this is one reason why we hate them.
They now admit that SGI was responsible for one example of "stolen" code they have been touting, but forget to mention that was removed from the kernel and was only compiled into intanium kernels.
"To date, we claim that more than one million lines of Unix System V protected code have been contributed to Linux through this model."
This is much stronger than the "derivative code" claim, because they are saying one million actual specific lines of code were taken, not just ideas and algorithms, and that they are from SysV code. IBM/Sequent AIX stuff is not System V code. They are going to get in trouble for telling such a blatant lie.
They are specifically claiming that SysV code was legally put into books and posted on public web sites but for non-commercial use only, and that Linux programmers illegally copied this code. This is new. And weakens their claims by making the chain from SysV code to my Redhat CD one link longer. It also sounds like they are claiming that any ancient Unix code that is still in their SysV codebase is SysV code.
"Some have claimed that, because SCO software code was present in software distributed under the GPL, SCO has forfeited its rights to this code. Not so - SCO never gave permission, or granted rights, for this to happen."
Once again a big lie: No one says the code is GPL because it was distributed under the GPL, they say that is GPL because SCO distributed it under the GPL. They can't claim they never gave permission to distribute the kernel under the GPL with their code in it, when they themselves knowingly did it. It's the distribution that's key, as they admit here. Doesn't matter who fired up emacs and typed the code in.
"Transfer of copyright ownership without express written authority of all proper parties is null and void."
A true, but irrelevent statement, as GPLed code does not generally involve the "transfer" of copyrights. Sounds good, though, doesn't it? Also ignoring that they don't own copyright to code written by IBM. Think about it. Does SCO own the copyrights to AIX? If they do, then they can sell it themselves without giving a penny to IBM. They can't sell AIX themselves? Then they don't own the copyright to the code.
Blah blah. More of the same legalistic, but not legally specific, mumbo jumbo about liability of Linux users and SCO's "IP". But the comment about programmers popping in code that is publically available but not public domain is new, and I expect to hear a lot more about it from them.
"bought out by IBM...You can be sure as hell that this is one thing that will not happen."
It occurs to me that IBM would not want to own control of the UNIX(TM)* sysV codebase. It would put IBM in too powerful a position with respect to other vendors, and cause them to gang up on IBM. There could be lawsuits about IBM abusing its monopolistic control of the source code and licenses that its competitors need to operate. Such accusations would be a problem even if they are false.
* UNIX(TM) is a trademark of The Open Group, and it is not the name of a single operating system no matter what SCO puts in its court documents.
With Redhat 9, the sound card and video card in my desktop don't work as well as they did with my RH4 distribution. Some days I am this close -->||<-- to reinstalling it.
I have seen complaints that Linux users are too resistant to change. Well, it worked, and now it doesn't, so why did somebody change it? And yes, I have looked at some of the source code to try to find the problem, but there is no documentation in the code or explication of the code structure, so there is no way someone who isn't already up to his eyeballs in X could make sense of it.
Not many sensible investors looking for long-term profit would invest in SCO. But they aren't the only ones who buy stock. As an example, say some company or person wanted to reward the SCO execs for their litigious nonsense. This company could do so legally and largely unnoticeably by buying SCO stock. With the execs dumping stock which is really worthless, money is just going from the purchasers to the officers' pockets.
After all, SCO is being sued by IBM for patent infringement, and they will almost certainly lose. They are going down. It is merely a question of when.
Slashdot Mirror
From reading their web pages, it certainly sounds like that is more or less the plan. It seems that RH wants to drop the consumer version of their distro. This amply clear from the packages that have disappeared in RH8 and RH9. Considering the hacker/hobbyist base of Linux, I was shocked to see them dropping mature popular window managers (fvwm et al), and classics like xtetris and xevil, as well as UNIX staples like fortune.
In all cases, it is because these programs conflict with the goal of selling the Redhat distro as a business desktop system, with minimum variations between installations and nothing "non-professional". RH employees have said this on mailing lists and in bugzilla comments.
They don't care about those of us that go to Fry's and buy their boxed sets. They need the businesses that will buy installation and support for 10,000 seats.
I have also read that there are discussions on stock boards where people are guessing that SCO will sell licenses to X percent of a bazillion Linux users at $1400 a pop. This gives a huge income for a company worth $250M, enticing day-trader types, unless you know enough to realize that the most likely value for X is 0%.
Right! Does "gigantic" refer to one billion ntics? Of course not!
I have to admit that grep '^giga' /usr/share/dict/words did not prove nearly as amusing as I had hoped.
No it isn't. I am on a local LUG mailing list, and people are politely helping newbies all the time, going out of their way to explain things that weren't even asked, just in case it might help.
"Not only the case on Slashdot, but go to any IRC help channel and you'll find the same the majority of the time."
IRC and /. were not exactly designed for thoughtful interaction.
You have two customers, X and Y, and you do extensive monitoring. If you tell X everything you know, then X knows what monitoring you are doing and can make the reasonable guess that you are monitoring Y in exactly the same way. This is information that would make an attack on Y easier.
Even five years ago, there was a large gap in performance between a decent PC and a cheap workstation. If you needed more than a what a PC could offer, you had to jump up to a workstation/server, which means people were routinely buying much beefier hardware than they needed. This gap has closed. You can buy the hardware that is right for your needs, and this translates into lower sales for sellers of high end hardware.
So, the primary purpose of graduate school is to make people feel so exploited and undervalued that being a post doc seems good in comparison?
Market research like this has been done. The results are that the windows users want the Linux system to act exactly like their Windows machines, because that is what they are used to. People don't know what will be better or easier to use or most especially make them more productive.
Thus you get the abhorrent tangling of the issues of useability and marketability. If you want to sell software, you want people to see a shiny, non-threatening interface. This has nothing to do with whether the software is actually good at getting things done. It is very difficult for market research to distinguish these. And if you want a shiny, non-threatening interface, use Windows. That's what it's there for, and there's nothing wrong with that.
Reading comprehension, please. I specifically said that I didn't care about when the species became known to western science.
As I said, a quick web search shows that the Vu Quang was well known to the locals, but it was unknown to western science. This has nothing to do with any search for bigfoot in the US. There are not piles of bigfoot bones and pelts sitting around in Oregon, unknown to scientists because nobody thought it worth mentioning to anyone. That is the comparison you have to make. You can't expect villagers in southeast Asia to go searching through European zoological literature to see if some of the critters haven't been described there. On the other hand, there is nobody in the Pacific northwest who wouldn't go screaming to the media if he found some concrete evidence of a bigfoot.
Ah, I just found the page you cut-and-pasted. Plagiarist.
It is of extreme relevance, from a legal standpoint. If the code is not literally copied unchanged, then someone claiming copyright violation has to prove that the aspect of the material copied is protected by copyright. For example, SCO would have to prove that some snippet they claim was copied from SysV and obfuscated is sufficiently different from public domain ancient Unix code that it deserves copyright protection.
This sort of test is not invoked when direct, verbatim copying occurs. So there is a big legal difference between cut-and-paste jobs and obfuscated code.
There is also the aspect that Linux kernel coders are not, for the most part, college freshmen. Kernel coders would not have wasted time obfuscating things they could more easily have written from scratch. If they were copying to save time, they would have copied verbatim. As far as trivial changes such as reindenting, shred has the option to ignore whitespace differences.
You are right that this doesn't prove anything. All it does is point out parts of the code that deserve more scrutiny by the human coders.
Oh, do go on. Discoveries in the 1800's aren't very interesting, though, as communications and travel were nothing compared to today. After all, sequoias weren't discoved until the mid-1800's, and they are 200 feet tall and don't run around hiding. So stick to the late twentieth century.
Now lets look at the Vu Quang. I did a quick google. The species was not discovered then, it was merely "discovered" by western science. Not the same thing at all, as the locals knew about them.
So how about some examples of large mammalian species found in the second half of the twentieth century that were previously unknown to any humans. That would be an interesting list. If there is anything on it.
Maybe they have been kept out of sight working in the labs that have perfected cold fusion.
Worth noting that stuff insured with the USPS is insured to arrive, but not at any particular time. I saw a guy going ballistic in the post office once because someone had mailed him airplane tickets. They arrived too late for him to use them, but they did arrive, so the insurance paid nothing.
Are mechanical car locks any better? One time I unlocked and opened the door to my car. It didn't feel quite right as it unlocked, so I looked again. I had unlocked the door using the key to a different car. The cars are from different manufacturers, but it looks like they use the same blanks for the keys.
Huh? I have been reading pretty thoroughly about SCO stuff, and this is the first I have seen mention of a fine at 7.5 cents per line per copy. Googling turns up zilch. Could you provide a reference?
This is true. But proving direct verbatim copying is straightforward. It is much harder to prove copyright infringement when the copying is not verbatim, because the plaintiff has to justify that the copied material is copyrightable.
If I look at your copyrighted code and retype it exactly into my code, then all that has to be shown is that I copied it (assuming the code is not GPLed or whatever). If I look at your copyrighted code and write a very similar routine for myself, you have to prove that I was looking at your code and you have to prove that whatever I took is copyrightable.
There is no hard line between what is copying and what is legal use of the ideas embodied in the code. It is for the courts to decide on the merits of each case.
This may be important when damages are considered. If code was copied from SysV but is essentially identical to code in the public domain, a judge may rule, "Defendent is guilty of violating copyrights, and we award to the plaintiff damages of 37 cents plus whatever loose change they can find under the cushions in the jury box." Of course, SCO would have to sue SGI for this, which they haven't done yet. The fact that this code was in Itanium specific code and was probably never run by more than a handful of people doesn't help SCO's quest for damages either.
Indeed, SCOX is up 10% today as of 3PM EDT under heavy trading. Not quite as much as I had expected, though. I figured it would go up $2 after the release of this "letter", but it's only up a $1.50.
But if this were true, they would not have to spin it quite so hard. "Books and Internet sites intended and authorized for the purpose of teaching and other non-commercial use cannot be copied for commercial use." Why does he use the word "intendend"? Intention has nothing to do with it. What matters is the copyright notices and any authorization. You can't legally copy an entire book that is copyrighted, but what about code snippets in a book? Numerical Recipes for example claims that that each routine is individually covered by copyrights, but are the rights under fair use different in the absence of such specific claims?
It sounds as if they are going to say that a simple statement like "This material was made available in hopes that someone could learn something from it" is equivalent to "Any commercial use of any part of the code shown here is forbidden." They could also be planning to claim copyright to bits of code that are in SysV but also present in ancient UNIX no longer under copyright, for which they need to argue that the code was never put in the public domain and the copyrights to the pieces has never lapsed. Time will tell.
Which are of little value if you care about security. Security is what drives the upgrade train. Old programs have known security holes. New programs require new compilers and libraries. If you use your machine on a network or over a modem, you can't go with old distributions.
"It's simply not feasible to hold back development just for the sake of old machines running at 1/20 or even 1/50 the clock speed of today. Such machines can't handle demanding modern applications."
Like emacs, LaTeX, X, squid, ppp, ftpd, sshd, xfig, xv, and mozilla? Sorry, they run great on a P200 (well, the lizard is marginal).
I am sure that most people would agree with this statement. But my point is that they are not honestly making their case about this, they are throwing out true but irrelevent legalistic statements to make their argument appear more convincing to the uncareful reader. They know that transfer of copyright has nothing to do with putting code under the GPL. They want people to think, "Hmm, SCO never signed anything transfering their copyrights, so their code can't be under the GPL," despite the lack of legal or logical connection between these two statements.
This is why we hate them. Well, this is one reason why we hate them.
They now admit that SGI was responsible for one example of "stolen" code they have been touting, but forget to mention that was removed from the kernel and was only compiled into intanium kernels.
"To date, we claim that more than one million lines of Unix System V protected code have been contributed to Linux through this model."
This is much stronger than the "derivative code" claim, because they are saying one million actual specific lines of code were taken, not just ideas and algorithms, and that they are from SysV code. IBM/Sequent AIX stuff is not System V code. They are going to get in trouble for telling such a blatant lie.
They are specifically claiming that SysV code was legally put into books and posted on public web sites but for non-commercial use only, and that Linux programmers illegally copied this code. This is new. And weakens their claims by making the chain from SysV code to my Redhat CD one link longer. It also sounds like they are claiming that any ancient Unix code that is still in their SysV codebase is SysV code.
"Some have claimed that, because SCO software code was present in software distributed under the GPL, SCO has forfeited its rights to this code. Not so - SCO never gave permission, or granted rights, for this to happen."
Once again a big lie: No one says the code is GPL because it was distributed under the GPL, they say that is GPL because SCO distributed it under the GPL. They can't claim they never gave permission to distribute the kernel under the GPL with their code in it, when they themselves knowingly did it. It's the distribution that's key, as they admit here. Doesn't matter who fired up emacs and typed the code in.
"Transfer of copyright ownership without express written authority of all proper parties is null and void."
A true, but irrelevent statement, as GPLed code does not generally involve the "transfer" of copyrights. Sounds good, though, doesn't it? Also ignoring that they don't own copyright to code written by IBM. Think about it. Does SCO own the copyrights to AIX? If they do, then they can sell it themselves without giving a penny to IBM. They can't sell AIX themselves? Then they don't own the copyright to the code.
Blah blah. More of the same legalistic, but not legally specific, mumbo jumbo about liability of Linux users and SCO's "IP". But the comment about programmers popping in code that is publically available but not public domain is new, and I expect to hear a lot more about it from them.
It occurs to me that IBM would not want to own control of the UNIX(TM)* sysV codebase. It would put IBM in too powerful a position with respect to other vendors, and cause them to gang up on IBM. There could be lawsuits about IBM abusing its monopolistic control of the source code and licenses that its competitors need to operate. Such accusations would be a problem even if they are false.
* UNIX(TM) is a trademark of The Open Group, and it is not the name of a single operating system no matter what SCO puts in its court documents.
With Redhat 9, the sound card and video card in my desktop don't work as well as they did with my RH4 distribution. Some days I am this close -->||<-- to reinstalling it.
I have seen complaints that Linux users are too resistant to change. Well, it worked, and now it doesn't, so why did somebody change it? And yes, I have looked at some of the source code to try to find the problem, but there is no documentation in the code or explication of the code structure, so there is no way someone who isn't already up to his eyeballs in X could make sense of it.
Not many sensible investors looking for long-term profit would invest in SCO. But they aren't the only ones who buy stock. As an example, say some company or person wanted to reward the SCO execs for their litigious nonsense. This company could do so legally and largely unnoticeably by buying SCO stock. With the execs dumping stock which is really worthless, money is just going from the purchasers to the officers' pockets.
After all, SCO is being sued by IBM for patent infringement, and they will almost certainly lose. They are going down. It is merely a question of when.