In addition, new customer sales indicate that there is still no better option for rock-solid, dependable technology for their core businesses than our SCO UNIX solutions.
Yeah, right. I hope SCO sober up before they go to court.
If you didn't know your code was in there while you were selling copies of the book, perhaps you'd have a claim to ask for all the copies back. Perhaps.
But if you kept selling the book, in the full knowledge of what it contained, can you then renege on your contract with the purchasers and demand the books back? Hell no.
This is the situation SCO are now in. Even after their initial allegations against IBM, they continued both selling copies of Linux, and offering it for free download from their web site.
Clearly they were entering into a contract with those people, with the knowledge that Linux contains RCU, SMP, and whatever else. That contract is the GPL. SCO has no grounds to revoke their customers contracts.
*If* (and this seems awfully unlikely) IBM breached their contract with SCO, then SCO can sue IBM. That doesn't make a whit of difference to people who got a GPL licence to the kernel from SCO.
There's a HP rx2600 Itanium2 desktop machine upstairs. It's running Linux and GNOME, looks very nice, is damn fast and reasonably quiet and would make a studly Linux machine for somebody who needed a large VM space.
Obviously Apple are using a "words mean what I want them to mean" definition of "64-bit desktop". To me, it means "something small and quiet enough to put on a desk, and designed to be used from the console by a single user", and the rx2600 qualifies. There's no single line between workstation and pc anymore.
Only thing more fanatical than a linux user... is a Macintosh user.
That's so true. I just happened to leaf through a Mac magazine (forget the name) and the degree of persecution/delusion expressed just took my breath away. I had trouble believing it wasn't a spoof. Just about every page had one or more of these:
* "Mac is so much better."
* "Even though we're slower in MHz and benchmarks it's really much faster"
* "Everyone (or Microsoft) is out to get us."
* "Only dumb people like other systems."
* "Why won't hardware vendors support us? We should all write and complain."
* "Steve is infallible"
* "Steve is ruining Apple"
It reminded me so vividly of Amiga or OS/2 users a few years ago. But the Amiga users had a better sense of humour.
It made Slashdot look like a paragon of selfconfidence and maturity. I'm sure not every Apple user (or publication) is like this, but some of them are pretty loopy.
But does usefulness necessarily imply marketability, as you originally said? No, of course not. Useful is valuable, but not everything valuable is marketable, as any undergrad economics student ought to know. For example
The thing is useful, but not at such a price and to enough people to justify for-profit production.
The costs of marketing and sales outweigh it's nonzero usefulness.
You cannot make a market because there is no way for users and producers to communicate effectively.
There is no way to charge for the thing. (e.g. simple un-patented ideas.)
Economics of production or sale don't allow a market to form; e.g. negative economies of scale.
Air is useful, but there is no market in it. The laws of calculus are useful but there is no market in them.
One theory for the rise of open source software is that some software is not appropriate for a per-user-licenced market.
Plenty of people are posting that this shows something about the difficulty of open source or Linux development. It really doesn't.
Projects die and people burn out on all platforms.
It's bitter when it happens to you, but it's part of the game.
80% of small businesses fold without the first two years. It's even higher in IT. I suspect the numbers are similar for projects inside big companies, though the failure can be covered up. Even within Microsoft, over 50% of projects are reported to be cancelled before release, and many people burn out after a few years. It might not make Slashdot headlines but dig around enough in people's blogs and you'll find all the same depression and disillusionment and sorrow.
Hell, it could have been even worse if it was a commercial/closed source project. The guy might have lost a lot of money, rather than just feeling he wasted his time.
The one good thing about open source is that when a project shuts down, it doesn't have to die. Other people can restart it or fork it perhaps some time later. I think this is some consolation.
That means nothing. IBM paid for a permanent, irrevocable licence too, that allowed them to relicence their own code however they wanted. (Yes, I've read the licence.) SCO are suing them anyhow.
There is no reason whatsoever that SCO could not turnaround and say that Sun has leaked code into Linux, regardless of what promises they've made to Sun.
Re:Be afraid "linux crunchies"
on
My Visit to SCO
·
· Score: 1
The message of that article is that sometimes unscrupulous people can abuse the court system to make a lot of money when they don't really have a moral case. OK, I knew that. I suppose it's worth being reminded every so often. It's hardly unique to Linux or even IT.
What is that meant to make me do? Sit around being depressed? Withdraw from society to a mountain fortress? Cave in to every troll with a lawyer?
What's with the title, What SCO Wants SCO Gets? Does this Daniel Lyons wanker seriously believe that IBM will just hand over $3,000,000,000 because SCO's management have stood over other companies in the past?
In other words, like many religious folk, the Linux-loving crunchies in the open-source movement are a) convinced of their own righteousness, and b) sure the whole world, including judges, will agree.
My observation (as an atheist, fwiw) is that practically everybody adopts that position ahead of a court case. I don't see IBM wondering out loud whether they might lose, and they know a thing or two about the law.
Anyhow, Linux people have been busy researching SCO and building extensive documentation about the case. I can't think of much more that could usefully be done at this stage. Can you?
My understanding is that BSA searches without a warrant are notionally "voluntary", because the BSA promises lower payments if they don't have to get a court order. But it is pure bluff to say that they any legal authority to do a search without a court order.
If SCO did try to go to court, the company ought to get in touch with IBM. Obviously it's not in IBM's interest to have any FUD cast on AIX's legal standing.
I'm not saying that nobody would be stupid enough to cave in, but I kind of doubt it.
I'm glad you liked them. (And I thought you were trolling, shame on me.:-)
I don't usually count clusters because that's more "cooperative computing" than a single kernel handling a large number of procs.
Well, single images are much harder to design and build and therefore make for better dick-size-wars, I agree. Big SMP machines sell low-end hardware the same way top-end cars sell mass-market models.
But clusters are far better for bang/buck, and for top-end scalability on problems that can handle it. Nobody could build a scalable 4000-CPU (or 50Tflop) shared-everything image. Doing it from a bunch of Linux boxes is not quite simple, but it is feasible. There will always be a few applications that need collosal single machines, but it seems like most high-power machines in the future will be clusters.
My guess is they're smoking angel dust, and you're far better off keeping away from it:
Effects include:
Death from stopped breathing, convulsions, brain hemorrhage, kidney failure, drug combinations, fatal accidents. Murder, suicide, self mutilation, and drowning from swimming under the influence to enhance floating sensation.
No, the contract is not necessarily terminated, and I don't think your phrasing is correct. One party unilaterally backing out of a contract does not necessarily terminate it.
(Surely this is obvious? I can't just wake up and decide to nullify my mortgage.)
Whether the contract is in fact terminated depends on whether it has a clause saying that it can be terminated by a party, or under certain conditions. This is the point at question at the moment: did the SCO-IBM have such a clause? SCO says yes, IBM says no.
Neither is it necessarily the case that SCO is in breach, unless the contract contains a "you must not act like a dumbass" clause. Making random press releases has no effect on the contract.
By default contracts can only either be completed, or all parties can mutually dissolve it, or one party can renege, or the contract can be frustrated (impossible to fulfil.) (IIRC, law class was a long time ago.)
Even if the contract was terminated, or SCO was in breach, that does not necessarily mean that IBM has lost the right to use SCO's material. In general when a contract is terminated or breached a court will have to decide on an appropriate outcome.
They can demand their own code be returned or destroyed which would leave gaps in the AIX source, but there's nothing at all to stop IBM from continuing to support the existing AIX installed base. IBM may be prohibited from selling any more AIX, but the existing copies will continue to exist.
McBride's "tree and branches" analogy seems to be saying that *any* Unix *by definition* is SCO property. If that is true, which seems incredibly unlikely, then nothing IBM could do to AIX would remove the taint of SCO.
Read-Copy-Update for example never shipped in either OpenServer or UnixWare, but McBride is claiming rights to it because it was in some unixes.
But this just makes no sense in terms of copyright: if you didn't write it (or acquire the rights) then it's not yours. SCO has no rights to code that was added by other people based on ancient SysV.
It's possible that IBM signed such a contract, but it seems unlikely. Even if they did, that would only be a problem for IBM, not for other Linux users.
Much as SCO would like people to imagine that, it is pure FUD, and even the most risk-averse manager shouldn't give it any credence.
For SCO to turn up with a sheriff, they'd have to get some kind of court order or injunction against anyone using AIX. IANAL, but I understand that this requires them to prove 1- that it's likely their suit will succeed, and 2- that SCO will suffer irreperable harm if people continue using AIX.
Obviously the first is pretty flaky, since SCO don't seem to have any hard evidence that there is infringement or that they have the right to pull customer licences.
Secondly, SCO is not suffering any irreparable harm. At the absolute peak of Darl's fantasies, they are missing out on licence revenue they are entitled to. If they won, they could be "made whole" by a licence payment in arrears from IBM.
Informed sources say that for SCO to get an injunction, they'd have to post a cash bond proportional to the possible harm caused by the injunction. IBM's Unix/Linux sales are a few billion dollars per year. Do you suppose SCO has that kind of cash?
(I have a nice image of Darl's friends going to a bail bondsman to try to raise $5,000,000,000, but of course they couldn't even make the 10% fee bondsmen typically charge.)
I'd hazard a guess that some IBM lawyers would turn up if SCO tried to get such an injunction... for values of "some" approaching "filling the sky and blotting out the sun". IBM simply can't afford to lose that one.
Re:I Have To Agree With Some Points Made Here
on
Settling SCOres
·
· Score: 1
If SCO faked it up, then putting the infringements in the comments would be half clever as you said. I'm not sure it would stand up in court though -- having comments that bear no relation to the code they're describing would raise eyebrows.
Pure genius!
Well, the real question is whether there are *no* unmolested copies of the SCO code around. I'd want to be awfully sure of that if I were SCO, because if it turns out that they did fake the evidence then they're up for perjury or contempt of court, to say nothing of a countersuit from IBM.
I suspect a thorough discovery process might find either an old backup of the source at SCO showing those comments weren't there, or perhaps some SCO source licensee has a copy.
If SCO claimed *not* to have any backups from a few years ago that'd have to look pretty suspicious.
Other possibilities are that the comments came from some earlier source, like the Lions book or old BSD. Comments in the kernel can be kind of patchy, so I can imagine some old ones hanging around for a long time. Or perhaps, as another poster said, they came from a standard specification.
The agreement as written is extremely generous to SCO, potentially including as Confidential Information items that are already public knowledge.
Sometimes companies propose overbroad NDAs as ambit claims just to see if you'll sign. If you really wanted to see their stinking source you might by able persuade them to strike some of the more onerous clauses. Or perhaps not.
I would be *extremely* leery of signing an NDA with a desparate and litigious company that has shown so much willingness to screw its previous friends.
If I really wanted to do this, I would get rock-solid legal advice (and possibly a second legal opinion) beforehand.
You really have to ask yourself, though, what are you going to gain from examining it? The community panel is not legally binding, and SCO has the freedom to show only part of the evidence. At best it will be inconclusive, which allows SCO to keep spreading FUD. At worst, you get sued as well.
SCO have made very serious accusations against the developer community. The onus is on them to substantiate those claims with more than a few vague sentences.
If SCO want to try this in the court of public opinion then they need to produce some evidence. If they want to hold onto the evidence until discovery in court then they should shut up and wait for the trial.
It's unlikely (not impossible) because code going into Linux is reviewed to a greater extent than I have seen or heard of on commercial projects. Have you ever worked on a project where you can't even commit code, let along ship it, without it going through at least two levels of review? Perhaps you have, but it's the exception not the rule.
It seems likely to me that SCO code would be caught by review if somebody tried to submit it to Linus, because it would "feel wrong" for Linux, or not look like it was original work. Things in the kernel work in particular patterns, and old code looks different to new code. See Hellwig's comment about how messy SCO code is. If it was copied, it would have to be rewritten so much that people might not bother in the first place.
Of couse it's not possible to catch all plagiarism by review, particularly when the purportedly original document is not available for comparison. It does at least seem like the kernel team have made a good-faith effort to prevent it, which ought to count for something.
There is also the question of whether anyone would *want* to copy from SCO. It's a pretty crappy OS by modern standards after all. Copying from BSD I can imagine, but SCO?
In any case, SCO have admitted today that they don't own the copyrights, and they only have a breach-of-contract case. So users who don't have contracts with SCO should be safe.
So at the point where SCO make a concrete and verifiable allegation, rather than vague FUD, then it's worth taking it seriously. SCO don't even have to reveal their code -- just saying where it is in Linux would allow the developers to know whether it was new or not. Up until then, SCO are just pissing in the wind and deserve all the crap they get.
Linus has better things to do than respond to trolls. SCO is a troll. A troll with a lawyer is still a troll.
They're not even a particularly *good* troll, at least by Slashdot standards. The original complaint contained really bizarre claims like the "Linux is a bicycle" paragraph, and while that can be amusing it does kind of tell people that you're not serious.
And then SCO started changing their story halfway through -- "it's just IBM", "it's not in the kernel", "it's in the kernel", "it's Linus". That kind of inconsistency always gives an amateur troll away. Look at the Subject Line Troll or the Recipe Troll -- they pick a good thing, and they stick to it. SCO could learn from them.
I think Linus was probably pretty sure that there was no infringing code in the kernel, and so he decided not to waste time or breath on the matter. If only the rest of us were so smart.;-)
It turns out today that SCO don't own the copyright or patents after all! So this is all pure FUD and bullshit on McBride's part. Not only are end users safe from copyright infringement, but so are distributors.
SCO *may* have a breach-of-contract tort against IBM, if IBM was stupid enough to sign a contract promising not to work on competitors to Monterey, which I doubt.
http://www.perens.com/Articles/SCO/BigLie.html This is the coolest thing Novell has done for themselves and for the rest of the world in years.
SCO are going to be torn limb-from-limb by countersuits. Bring popcorn!
Ah, interesting. That's more or less the reply I expected.
Saving bandwidth is the least important reason why open source people use diffs. It's more about being able to see what the intended change to the file is, being able to quote and discuss it in email client, and being able to merge simultaneous changes easily (or know that they conflict).
I suppose if everyone had a sufficiently good SCM system then you might be able to check your thing in on a private branch (or whatever) and allow the other guy to see it that way. I think open source projects refrain from that because it would generate too many little branches that don't really need to be stored in the long term, but with a system good enough to hide them except from the people who need them it'd be OK. Come to think of it Subversion would support that way of working pretty well.
Slashdot Mirror
Yeah, right. I hope SCO sober up before they go to court.
If you didn't know your code was in there while you were selling copies of the book, perhaps you'd have a claim to ask for all the copies back. Perhaps.
But if you kept selling the book, in the full knowledge of what it contained, can you then renege on your contract with the purchasers and demand the books back? Hell no.
This is the situation SCO are now in. Even after their initial allegations against IBM, they continued both selling copies of Linux, and offering it for free download from their web site.
Clearly they were entering into a contract with those people, with the knowledge that Linux contains RCU, SMP, and whatever else. That contract is the GPL. SCO has no grounds to revoke their customers contracts.
*If* (and this seems awfully unlikely) IBM breached their contract with SCO, then SCO can sue IBM. That doesn't make a whit of difference to people who got a GPL licence to the kernel from SCO.
There's a HP rx2600 Itanium2 desktop machine upstairs. It's running Linux and GNOME, looks very nice, is damn fast and reasonably quiet and would make a studly Linux machine for somebody who needed a large VM space.
Obviously Apple are using a "words mean what I want them to mean" definition of "64-bit desktop". To me, it means "something small and quiet enough to put on a desk, and designed to be used from the console by a single user", and the rx2600 qualifies. There's no single line between workstation and pc anymore.
Only thing more fanatical than a linux user... is a Macintosh user.
That's so true. I just happened to leaf through a Mac magazine (forget the name) and the degree of persecution/delusion expressed just took my breath away. I had trouble believing it wasn't a spoof. Just about every page had one or more of these:
* "Mac is so much better."
* "Even though we're slower in MHz and benchmarks it's really much faster"
* "Everyone (or Microsoft) is out to get us."
* "Only dumb people like other systems."
* "Why won't hardware vendors support us? We should all write and complain."
* "Steve is infallible"
* "Steve is ruining Apple"
It reminded me so vividly of Amiga or OS/2 users a few years ago. But the Amiga users had a better sense of humour.
It made Slashdot look like a paragon of selfconfidence and maturity. I'm sure not every Apple user (or publication) is like this, but some of them are pretty loopy.
But does usefulness necessarily imply marketability, as you originally said? No, of course not. Useful is valuable, but not everything valuable is marketable, as any undergrad economics student ought to know. For example
Air is useful, but there is no market in it. The laws of calculus are useful but there is no market in them.
One theory for the rise of open source software is that some software is not appropriate for a per-user-licenced market.
Plenty of people are posting that this shows something about the difficulty of open source or Linux development. It really doesn't.
Projects die and people burn out on all platforms.
It's bitter when it happens to you, but it's part of the game.
80% of small businesses fold without the first two years. It's even higher in IT. I suspect the numbers are similar for projects inside big companies, though the failure can be covered up. Even within Microsoft, over 50% of projects are reported to be cancelled before release, and many people burn out after a few years. It might not make Slashdot headlines but dig around enough in people's blogs and you'll find all the same depression and disillusionment and sorrow.
Hell, it could have been even worse if it was a commercial/closed source project. The guy might have lost a lot of money, rather than just feeling he wasted his time.
The one good thing about open source is that when a project shuts down, it doesn't have to die. Other people can restart it or fork it perhaps some time later. I think this is some consolation.
That means nothing. IBM paid for a permanent, irrevocable licence too, that allowed them to relicence their own code however they wanted. (Yes, I've read the licence.) SCO are suing them anyhow.
There is no reason whatsoever that SCO could not turnaround and say that Sun has leaked code into Linux, regardless of what promises they've made to Sun.
The message of that article is that sometimes unscrupulous people can abuse the court system to make a lot of money when they don't really have a moral case. OK, I knew that. I suppose it's worth being reminded every so often. It's hardly unique to Linux or even IT.
What is that meant to make me do? Sit around being depressed? Withdraw from society to a mountain fortress? Cave in to every troll with a lawyer?
What's with the title, What SCO Wants SCO Gets? Does this Daniel Lyons wanker seriously believe that IBM will just hand over $3,000,000,000 because SCO's management have stood over other companies in the past?
In other words, like many religious folk, the Linux-loving crunchies in the open-source movement are a) convinced of their own righteousness, and b) sure the whole world, including judges, will agree.
My observation (as an atheist, fwiw) is that practically everybody adopts that position ahead of a court case. I don't see IBM wondering out loud whether they might lose, and they know a thing or two about the law.
Anyhow, Linux people have been busy researching SCO and building extensive documentation about the case. I can't think of much more that could usefully be done at this stage. Can you?
I would be surprised if that happens.
My understanding is that BSA searches without a warrant are notionally "voluntary", because the BSA promises lower payments if they don't have to get a court order. But it is pure bluff to say that they any legal authority to do a search without a court order.
If SCO did try to go to court, the company ought to get in touch with IBM. Obviously it's not in IBM's interest to have any FUD cast on AIX's legal standing.
I'm not saying that nobody would be stupid enough to cave in, but I kind of doubt it.
I'm glad you liked them. (And I thought you were trolling, shame on me. :-)
I don't usually count clusters because that's more "cooperative computing" than a single kernel handling a large number of procs.
Well, single images are much harder to design and build and therefore make for better dick-size-wars, I agree. Big SMP machines sell low-end hardware the same way top-end cars sell mass-market models.
But clusters are far better for bang/buck, and for top-end scalability on problems that can handle it. Nobody could build a scalable 4000-CPU (or 50Tflop) shared-everything image. Doing it from a bunch of Linux boxes is not quite simple, but it is feasible. There will always be a few applications that need collosal single machines, but it seems like most high-power machines in the future will be clusters.
I'd love to see an E-10000 or a pSeries690 with a full spread of processors running Linux.
Your wish is granted:
1000-CPU Itanium HP Linux cluster
Linux on 32-way POWER4, 60GB RAM
Linux on pre-production POWER5
Linux on 24CPU E10000
Fast enough?
Ask Google and you can find similar numbers for 64-CPU ia64 machines from SGI and 100-CPU clusters from HP.
Stump up a few million bucks and HP, IBM, SGI et al will be beating down your door to sell you a big Linux machine.
Buy a regular 2GHz PC, put Linux and distcc and a ibm-aix-ppc cross compiler on it, and enjoy faster compiles.
Effects include:
Sounds about right.
No, the contract is not necessarily terminated, and I don't think your phrasing is correct. One party unilaterally backing out of a contract does not necessarily terminate it.
(Surely this is obvious? I can't just wake up and decide to nullify my mortgage.)
Whether the contract is in fact terminated depends on whether it has a clause saying that it can be terminated by a party, or under certain conditions. This is the point at question at the moment: did the SCO-IBM have such a clause? SCO says yes, IBM says no.
Neither is it necessarily the case that SCO is in breach, unless the contract contains a "you must not act like a dumbass" clause. Making random press releases has no effect on the contract.
By default contracts can only either be completed, or all parties can mutually dissolve it, or one party can renege, or the contract can be frustrated (impossible to fulfil.) (IIRC, law class was a long time ago.)
Even if the contract was terminated, or SCO was in breach, that does not necessarily mean that IBM has lost the right to use SCO's material. In general when a contract is terminated or breached a court will have to decide on an appropriate outcome.
I realize you're trolling, but: what makes you think BSD is any safer?
As soon as SCO notice that BSD is not quite dead yet, they'll be on your ass as well. The same bogus claims of code theft can be applied equally.
McBride's "tree and branches" analogy seems to be saying that *any* Unix *by definition* is SCO property. If that is true, which seems incredibly unlikely, then nothing IBM could do to AIX would remove the taint of SCO.
Read-Copy-Update for example never shipped in either OpenServer or UnixWare, but McBride is claiming rights to it because it was in some unixes.
But this just makes no sense in terms of copyright: if you didn't write it (or acquire the rights) then it's not yours. SCO has no rights to code that was added by other people based on ancient SysV.
It's possible that IBM signed such a contract, but it seems unlikely. Even if they did, that would only be a problem for IBM, not for other Linux users.
Much as SCO would like people to imagine that, it is pure FUD, and even the most risk-averse manager shouldn't give it any credence.
For SCO to turn up with a sheriff, they'd have to get some kind of court order or injunction against anyone using AIX. IANAL, but I understand that this requires them to prove 1- that it's likely their suit will succeed, and 2- that SCO will suffer irreperable harm if people continue using AIX.
Obviously the first is pretty flaky, since SCO don't seem to have any hard evidence that there is infringement or that they have the right to pull customer licences.
Secondly, SCO is not suffering any irreparable harm. At the absolute peak of Darl's fantasies, they are missing out on licence revenue they are entitled to. If they won, they could be "made whole" by a licence payment in arrears from IBM.
Informed sources say that for SCO to get an injunction, they'd have to post a cash bond proportional to the possible harm caused by the injunction. IBM's Unix/Linux sales are a few billion dollars per year. Do you suppose SCO has that kind of cash?
(I have a nice image of Darl's friends going to a bail bondsman to try to raise $5,000,000,000, but of course they couldn't even make the 10% fee bondsmen typically charge.)
I'd hazard a guess that some IBM lawyers would turn up if SCO tried to get such an injunction... for values of "some" approaching "filling the sky and blotting out the sun". IBM simply can't afford to lose that one.
If SCO faked it up, then putting the infringements in the comments would be half clever as you said. I'm not sure it would stand up in court though -- having comments that bear no relation to the code they're describing would raise eyebrows.
Pure genius!
Well, the real question is whether there are *no* unmolested copies of the SCO code around. I'd want to be awfully sure of that if I were SCO, because if it turns out that they did fake the evidence then they're up for perjury or contempt of court, to say nothing of a countersuit from IBM.
I suspect a thorough discovery process might find either an old backup of the source at SCO showing those comments weren't there, or perhaps some SCO source licensee has a copy.
If SCO claimed *not* to have any backups from a few years ago that'd have to look pretty suspicious.
Other possibilities are that the comments came from some earlier source, like the Lions book or old BSD. Comments in the kernel can be kind of patchy, so I can imagine some old ones hanging around for a long time. Or perhaps, as another poster said, they came from a standard specification.
Yes, here it is.
The agreement as written is extremely generous to SCO, potentially including as Confidential Information items that are already public knowledge.
Sometimes companies propose overbroad NDAs as ambit claims just to see if you'll sign. If you really wanted to see their stinking source you might by able persuade them to strike some of the more onerous clauses. Or perhaps not.
I would be *extremely* leery of signing an NDA with a desparate and litigious company that has shown so much willingness to screw its previous friends.
If I really wanted to do this, I would get rock-solid legal advice (and possibly a second legal opinion) beforehand.
You really have to ask yourself, though, what are you going to gain from examining it? The community panel is not legally binding, and SCO has the freedom to show only part of the evidence. At best it will be inconclusive, which allows SCO to keep spreading FUD. At worst, you get sued as well.
SCO have made very serious accusations against the developer community. The onus is on them to substantiate those claims with more than a few vague sentences.
If SCO want to try this in the court of public opinion then they need to produce some evidence. If they want to hold onto the evidence until discovery in court then they should shut up and wait for the trial.
Good point.
"Whoa... nasty infestation you've got there. Still but, a bit of hydrogen cyanide ought to clear them out."
It's unlikely (not impossible) because code going into Linux is reviewed to a greater extent than I have seen or heard of on commercial projects. Have you ever worked on a project where you can't even commit code, let along ship it, without it going through at least two levels of review? Perhaps you have, but it's the exception not the rule.
It seems likely to me that SCO code would be caught by review if somebody tried to submit it to Linus, because it would "feel wrong" for Linux, or not look like it was original work. Things in the kernel work in particular patterns, and old code looks different to new code. See Hellwig's comment about how messy SCO code is. If it was copied, it would have to be rewritten so much that people might not bother in the first place.
Of couse it's not possible to catch all plagiarism by review, particularly when the purportedly original document is not available for comparison. It does at least seem like the kernel team have made a good-faith effort to prevent it, which ought to count for something.
There is also the question of whether anyone would *want* to copy from SCO. It's a pretty crappy OS by modern standards after all. Copying from BSD I can imagine, but SCO?
In any case, SCO have admitted today that they don't own the copyrights, and they only have a breach-of-contract case. So users who don't have contracts with SCO should be safe.
So at the point where SCO make a concrete and verifiable allegation, rather than vague FUD, then it's worth taking it seriously. SCO don't even have to reveal their code -- just saying where it is in Linux would allow the developers to know whether it was new or not. Up until then, SCO are just pissing in the wind and deserve all the crap they get.
Linus has better things to do than respond to trolls. SCO is a troll. A troll with a lawyer is still a troll.
;-)
They're not even a particularly *good* troll, at least by Slashdot standards. The original complaint contained really bizarre claims like the "Linux is a bicycle" paragraph, and while that can be amusing it does kind of tell people that you're not serious.
And then SCO started changing their story halfway through -- "it's just IBM", "it's not in the kernel", "it's in the kernel", "it's Linus". That kind of inconsistency always gives an amateur troll away. Look at the Subject Line Troll or the Recipe Troll -- they pick a good thing, and they stick to it. SCO could learn from them.
I think Linus was probably pretty sure that there was no infringing code in the kernel, and so he decided not to waste time or breath on the matter. If only the rest of us were so smart.
It turns out today that SCO don't own the copyright or patents after all! So this is all pure FUD and bullshit on McBride's part. Not only are end users safe from copyright infringement, but so are distributors.
SCO *may* have a breach-of-contract tort against IBM, if IBM was stupid enough to sign a contract promising not to work on competitors to Monterey, which I doubt.
http://www.perens.com/Articles/SCO/BigLie.html
This is the coolest thing Novell has done for themselves and for the rest of the world in years.
SCO are going to be torn limb-from-limb by countersuits. Bring popcorn!
Ah, interesting. That's more or less the reply I expected.
Saving bandwidth is the least important reason why open source people use diffs. It's more about being able to see what the intended change to the file is, being able to quote and discuss it in email client, and being able to merge simultaneous changes easily (or know that they conflict).
I suppose if everyone had a sufficiently good SCM system then you might be able to check your thing in on a private branch (or whatever) and allow the other guy to see it that way. I think open source projects refrain from that because it would generate too many little branches that don't really need to be stored in the long term, but with a system good enough to hide them except from the people who need them it'd be OK. Come to think of it Subversion would support that way of working pretty well.
According to a recent Economist article, X-Men 2 was nearly simultaneously released around the world to prevent this problem. Perhaps they're wrong?