Who needs a cup? A little tin, a little lead, pound it between a couple of rocks, put it in the palm of your hand and hold your hand out over a fire for long enough. The creases in your palm form the guid channels for the liquid alloy; you just drip it through them.
Although I agree that the copyright issue is really weak for SCO, I think that some creative lawyering could make it a lot more dangerous.
The Supreme Court has a case on the docket this session about whether copyright applies to an aggregation. If the Court decides that it does, and I expect them to do so, then I would play up Torvalds' statement about how he rewrote errno.h. Remember the e-mail where Linus talked about how he'd originally used a different set of constants, but had had to shift back to the Sys V #defines so that other programs would work? If an aggregation is copyrightable, that consititutes a clear admission of wilful infringement.
In fact, I wouldn't rule out a successful line of arugment like that if the Court rules that an aggregation isn't copyrightable. I think a good attorney -- and David Boies is a really good attorney -- could make a solid case that there was protected intellectual content in the pattern of choices of the constants, based on the grouping of constants with similar meaning together. If he can establish that point, then Torvalds' e-mail might as well talk about "cutting off their air supply".
Well, I'm not sure that I agree about what the documents appear to show, but I'll concede that for the sake of argument. Even were I to grant that, it's far from clear that Novell wins.
Your argument hangs on the meaning of "specificity", and in this case, that hangs on exactly what the parties involved said to one another. "Specificity" certainly can mean "by name and as listed", but it need not mean that. If SCO has depositions in which the witnesses say, 'Yes, in discussions, we agreed that "as necessary" meant "all Unix copyrights plus anything else needed, effective on the date on which we signed the amendment"', and Novell can't refute them, then the case is over in SCO's favor. If Novell has depositions which say "Yeah, weren't we clever? By fudging on the wording, we transferred nothing, but made it seem like we were giving them something. Sharp practice maybe, but that's their loss", and SCO can't refute that, then Novell wins.
For my part, I expect that there is going to conflicting testimony by various parties, and it'll come down to who the jury believes. It's going to be a very nasty trial: current Novell employees will be asked about how much they'll lose in stock options if Novell loses the case, and former Novell employees will be asked "why they ever-so-conveniently left the company", with insinuaiotns that they were about to be fired.
I think we agree about the key thing: this is a routine motion, which won't be granted immediately. I doubt that the judge will deny it outright.
Folks, this filing is a "Duh" -- not because it's valid, but because any competent attorney will file for a declaratory judgement. After all, the judge might grant it, and there's no harm in trying.
I don't think that they expect the judge to grant the motion, though. IBM would have to show that there's no controversy about SCO holding the copyrights to the UNIX code in question, and that's patently false. SCO claims the copyrights and has documents the appear to show that. Novell has documents which might impose some limitations on the transfer, but it isn't clear that they apply in this case, or that they're valid. SCO has affadavits from the people who signed the original documents which appear to show that they intended to transfer the copyrights in full. That matter is currently under judicial review, and until it's resolved, that controversy prevents a declaratory judgement.
I expect the judge to deline to rule on the motion in the interests of allowing Novell v. SCO to proceed unimpeded. IBM's attorneys have to try, both because they might win, and because if the Novell case goes against SCO, then IBM will have it in the record that they asked for declaratory judgement earlier, and so can forestall some procedural objections from SCO in that case.
The first time, yes, and there's almost never a second time for a mentally ill and dangerous person. The standard for determination of "guilty but insane" or "not guilty by reason of insanity" is that the person was unable to realize that his or her act was wrong at the time the act was committed. Such individuals, however, tend to wind up in mental institutions for long periods of time: think Theodore Koszinski.
This is based on two major factors, one legal and the other medical.
First, the right to refust treatment is deeply welded with the right to seek it. Treatment can only be required when a person would be a treat to himself or others if he withdrew from treatment. Medical confinement is even harder to require, as it should be. That's particularly true in the case of psychological disorders. Remember that even though _One Flew Over the Cuckoo's Nest_ was fiction, the kinds of abuses upon which it was based weren't.
Second, it's important to know that anti-psychotic and anti-manic drugs are not panaceas. People taking high doses of these drugs are more functional than they would be without them, but do not necessarily feel better, and often feel worse. In the absence of a threat to others, medical treatment is provided for the benefit of the individual being treated, not for the benefit of society. That's why we let cancer patients die when they say they don't want treatment any more, and that's why we let mentally ill people go off their meds.
Whadaya mean "need a fusion reactor"? Our design plan is that the CPU will be so hot and so compressed that it will serve as its own fusion reactor! Intel and AMD are already on board with this. If it weren't for Transmeta, we'd have a perfect plan for world domination! (Think of the games, too! Microsoft Reactor Simulator 2000 Rad...)
Actually, you're wrong. India is going through a huge period of economic growth throughout its economy. In this, it is replaying a pattern very like the other industrializing countries of the world. It appears to you and me that India is a shambles, but that isn't because the economy is doing poorly, but because it started out doing so much worse.
Most countries go through an extended mercantilist period during their early mass industrialization. During that period, wages in the industrializing country are typically quite low becuase the coutry's currency is artificially depressed. During that period, the country's industrial production skyrockets. Since consumers in the country buy their own products with their own currency, the irrational pricing structure of their industry's exports doesn't affect them, and they act as an internal gate which forces the quality of their exports up.
Eventually, however, growth leads to major industries being unable to provide for their own production with local acquired raw materials. At that point, prices of locally produced products start to reflect the relative level of the currency: foreign raw materials must be bought in foreign currency, which raises the prices of the finished goods into which they are made. That triggers a sharp round of inflation, which leads to a more restrictive currency policy. The price difference between finished good produced in country and those produced abroad gradually shrinks, due to this pressure.
To see this pattern in action, you can go back to Japan in the fifties through the eighties, S. Korea since the eighties, and India now. Alternatively, you can go back to the United State in the late nineteenth century, or to the great European powers in the early nineteenth century.
Europe and the United States managed to extend the period during which they could pursue a mercantilist policy somewhat longer by maintaining a captive market to which finished goods could be exported and from which raw materials could be imported in the local currency. The European powers did this by maintaining colonial markets in Asia, Africa, and, to a lesser extent, the Americas. The Americans settled our West, which became a huge source of raw materials for our East coast industries. The captive markets allowed the industrial base to continue to acquire raw materials at a disproportionately low price.
Schumpeterian equilibrium may well apply to an economy which is dependent on a influx of externally produced raw materials balanced by an egress of internally produced finished goods. That's not the case for economies in their earlier stages of industrialization and development. I don't know how long it will take for India to reach that state, but given the combination of destitution and size of her population, I wouldn't be inclined to expect her government to adopt less mercantilist policies any time soon. It's not rational to do so.
I call bullshit on this, Chumley. Permissions and account handling didn't change at all from NT to XP. So either you're lying or you confused the changes in the domain forest structure when you moved to AD from the NT domain auth model.
I wish I could believe the latter was more likely, but I regret to say that I consider the former more probable. You see, you confuse an Office major vesion migration with an NT major version migration, and any sysadmin worth his or her salt would know that those are very different beasts. Thus, I conclude that you're making your story up.
Never served on a jury, have you? In fact, I'd wager you've never even been into a courthouse. I think that you'd rethink that statement the same way Avi Rubin rethought what he thought about poll watchers if you ever had.
But what you're ignoring is that the CSS providers spent money to create those proprietary standards. If the OSS community wants access to them, then somebody is going to have convince the CSS vendors to yield them up.
Want full access to the Windows source for all people? Microsoft is a business, and, for a suitable sum of money, could be convinced to part with it. (Under US law, the power of eminent domain would certainly suffice to require that.) Go to your friendly local government and convince them to buy the Windows source from Microsoft. At nominal rates of return on investment, I figure that it will only cost on the order of US$250,000,000,000. Maybe a little more...say a round trillion dollars US.
Then again, maybe that isn't feasible. But that's what asking for a "fair shake" would require. If it isn't going to happen, that is not the fault of the court. It's the fault of the polity that it doesn't want to spend that kind of money.
Actually, one of the key points where EU antitrust law is unsettled is what the purpose of antitrust law is.
The US antitrust system is older, and so there are a bunch of court cases which establish that the Sherman Act protects consumers, but makes no attempt to protect competitors. It only provides a means for competitors to sue after the fact and recover triple damages. The courts have consistently decided that there is no direct protection provided to competitors or competition; if the monopolist can show a benefit to consumers, then the court supports the right of the monopolist to sell.
EU law is much less well-established, and it isn't clear what the law means. In fact, the biggest single point which I expect to see argued in front of the Court of the First Instance and the ECJ is whether only consumers are protected by the act, or if competitors are protected as well. If the former, then the EC will lose this case, fine or no fine, and Microsoft will set a huge precedent protecting the company's ability to bundle new features into Windows. If the latter, then the feature extension Microsoft has engaged in over the years will be blocked in Europe.
Uhh...no. Although I'm very angry with the Slashdot editors for posting this particular lead, this is blatant infringement of the most egregious form. Fair use only defends people who violate copyright in small (e.g. by printing a limited passage with full attribution) or who have some other significant difference from the original (e.g. a parody is almost be definition infringing, but protected by fair use.) Including a whole text verbatim for the purpose of avoiding the tariff a copyright holder has set is not fair use, whether or not the author benefits from it.
The GP is correct: ALL GSM phones have a SIM. It is a requirement of the spec for GSM that a subcriber be able to remove and replace his or her subscriber information module (SIM) and transfer it. That creates and fosters competition among the carriers, because once you have a phone, you can use it to access the network of any GSM carrier.
Well, no. Although your particular machanism may remain private -- although I wouldn't, and don't, bet on it -- the odds that any given dictionary attack will steal some user's password are very high. And the odds that some dictionary attack will eventually steal yours are also very high.
Another alternative is to use a random pronounceable word generator for your language of choice. Entropy/character is lower than that of the first character of random free text, but higher than that of frequently memorized text. (How many Americans who use first letters of a phrase come up with "Fsasya,ofbfutcann" or some initial segment thereof? Answer: lots of people, since they're using the first letters of the Gettysburg Address, which they were required to memorize at school.) Problem is, first letters of the words in a phrase is vulnerable to dictionary attacks against the dictionary of lyrics of popular songs, etc.
Now that is a statement of intent offered by a person who had both the authority and knowledge to make it. Unfortunately, it doesn't help IBM's case very much, and may, in fact, hurt it.
First, that's in a letter from AT&T to IBM. It's already established that IBM's terms were somewhat less stringent than everybody else's, and the stuff I've been talking about is RCU, which was taken from Sequent's Dynix, not IBM's AIX. I've always assumed that the sentence supposedly added to clause 2.01 clarified that the IBM interpretation applied to all licensees, but I've never seen it, and, without seeing it, can't tell.
Even if it does say that, the IBM interpretation is much less friendly to IBM in the SCO case than many people seem to think. AT&T disclaims ownership of modifications and derivative works prepared for IBM, but doesn't disclaim all ownership of those works. It only says that they are owned as a whole by IBM. That would mean that IBM didn't need AT&T's permission to disseminate, for instance, AIX as a whole to a third party who also had a license to use System V Unix. It says nothing about whether IBM had the right to disseminate a piece of AIX to a third party who had NO license to use Unix.
Even more troubling, the second sentence clearly retains ownership of any code taken from Unix itself. It would not be unreasonable to take that to mean that if any of the allegedly novel Dynix source was taken letter for letter from Unix (e.g. if it contained the line for (; *name; name++)), then AT&T still held copyright to that line. In that case, dissemenation of that line would constitute an infringement of AT&T's rights.
What clear statement of intent? The $echo article is no statement of intent: it is not a deposition; it is not presented over the signature of a corporate officer, nor on the letterhead of an attorney representing AT&T. It has exactly as much legal force as my trolling here does: none.
But that doesn't say it's wrong, either, merely that it isn't binding. If you want to know why I'm particularly skeptical about it, here's a cool test for you. Run down to Novell's site, and troll through the documents they ever so helpfully put up there. Look for the magical "last sentence added to section 2.01". Tell me if you find it, because I didn't. I looked for it, because that was the first piece of evidence that Novell had ever presented which in any way called into question SCO's claim to ownership. I find no references to it except in the pdf that was sent to SCO to intimidate them.
I think that Novell is playing the same kind of game for PR the SCO is. I don't trust them, and I don't believe that their contract says what they would like you to believe it says. Otherwise, it would ahve leaked, and it has not done so.
No, I'm aware of the $echo magazine article. It's irrelevant.
First, it is only a magazine article, and is worth as much as the paper that the electronic version is printed on. It's not a contract; it refers to a clause in the contract. It doesn't even quote the clause.
Second, the most plausible interpretation of "no ownership" is just that: "no ownership" -- you have the right to distribute the software without paying us. That says nothing about other contractual limitations on your right to distribute the software. Just because AT&T claimed no ownership does not mean that Sequent got an exception from the derivative work license terms from AT&T.
Nonsense. They list files and lines of files, and show, via the RCU example, exactly how the content of two of those files directly increased the value of Linux to the detriment of their own business. There are only two questions: whether Sequent got an exception from the derivative work license terms from AT&T, and whether Novell's case holds. I've seen no evidence from IBM to support the former, and what I've seen out of Novell is not at all convincing on the latter. That will hang on what the various people involved remember meaning when they signed the contract.
Engineering is all about tolerances and modes of failure. If I design my car to be able to take a fifteen mph front end collision, and you drive into a wall at thirty, I'm not responsible, and my E&O won't wind up paying out.
Currently, software is built in a craft/guild model: senior developers (masters) teach junior developers (journeymen) who've reached a certain level of expertise. Interns (apprentices) are drafted into the profession and groomed into junior devs. There is a widely held notion of subjective quality, and we can recognize a masterwork, but we can't quantify what it takes to generate one.
Software engineering will become a true engineering discipline only when there is an objective measure of defect level and an objective notion of what constitutes an adequately circumscribed operating environment. Once we have adequate definitions of those things, though, software production will become industrialized almost immediately.
Did you read the Security Focus article? It explicitly stated both that Blaster was not related to the blackout and that SF had been one of the first publications to extend the hypothesis that they had been related.
In short, the Microsoft bashers were wrong -- and at least Security Focus had the guts to acknowledge it.
Slashdot Mirror
Who needs a cup? A little tin, a little lead, pound it between a couple of rocks, put it in the palm of your hand and hold your hand out over a fire for long enough. The creases in your palm form the guid channels for the liquid alloy; you just drip it through them.
I didn't read the article, but, come on, how useful can the book really be? If I wanted to date design patterns I'd have become a computer programmer!
Although I agree that the copyright issue is really weak for SCO, I think that some creative lawyering could make it a lot more dangerous.
The Supreme Court has a case on the docket this session about whether copyright applies to an aggregation. If the Court decides that it does, and I expect them to do so, then I would play up Torvalds' statement about how he rewrote errno.h. Remember the e-mail where Linus talked about how he'd originally used a different set of constants, but had had to shift back to the Sys V #defines so that other programs would work? If an aggregation is copyrightable, that consititutes a clear admission of wilful infringement.
In fact, I wouldn't rule out a successful line of arugment like that if the Court rules that an aggregation isn't copyrightable. I think a good attorney -- and David Boies is a really good attorney -- could make a solid case that there was protected intellectual content in the pattern of choices of the constants, based on the grouping of constants with similar meaning together. If he can establish that point, then Torvalds' e-mail might as well talk about "cutting off their air supply".
Well, I'm not sure that I agree about what the documents appear to show, but I'll concede that for the sake of argument. Even were I to grant that, it's far from clear that Novell wins.
Your argument hangs on the meaning of "specificity", and in this case, that hangs on exactly what the parties involved said to one another. "Specificity" certainly can mean "by name and as listed", but it need not mean that. If SCO has depositions in which the witnesses say, 'Yes, in discussions, we agreed that "as necessary" meant "all Unix copyrights plus anything else needed, effective on the date on which we signed the amendment"', and Novell can't refute them, then the case is over in SCO's favor. If Novell has depositions which say "Yeah, weren't we clever? By fudging on the wording, we transferred nothing, but made it seem like we were giving them something. Sharp practice maybe, but that's their loss", and SCO can't refute that, then Novell wins.
For my part, I expect that there is going to conflicting testimony by various parties, and it'll come down to who the jury believes. It's going to be a very nasty trial: current Novell employees will be asked about how much they'll lose in stock options if Novell loses the case, and former Novell employees will be asked "why they ever-so-conveniently left the company", with insinuaiotns that they were about to be fired.
I think we agree about the key thing: this is a routine motion, which won't be granted immediately. I doubt that the judge will deny it outright.
Folks, this filing is a "Duh" -- not because it's valid, but because any competent attorney will file for a declaratory judgement. After all, the judge might grant it, and there's no harm in trying.
I don't think that they expect the judge to grant the motion, though. IBM would have to show that there's no controversy about SCO holding the copyrights to the UNIX code in question, and that's patently false. SCO claims the copyrights and has documents the appear to show that. Novell has documents which might impose some limitations on the transfer, but it isn't clear that they apply in this case, or that they're valid. SCO has affadavits from the people who signed the original documents which appear to show that they intended to transfer the copyrights in full. That matter is currently under judicial review, and until it's resolved, that controversy prevents a declaratory judgement.
I expect the judge to deline to rule on the motion in the interests of allowing Novell v. SCO to proceed unimpeded. IBM's attorneys have to try, both because they might win, and because if the Novell case goes against SCO, then IBM will have it in the record that they asked for declaratory judgement earlier, and so can forestall some procedural objections from SCO in that case.
The first time, yes, and there's almost never a second time for a mentally ill and dangerous person. The standard for determination of "guilty but insane" or "not guilty by reason of insanity" is that the person was unable to realize that his or her act was wrong at the time the act was committed. Such individuals, however, tend to wind up in mental institutions for long periods of time: think Theodore Koszinski.
This is based on two major factors, one legal and the other medical.
First, the right to refust treatment is deeply welded with the right to seek it. Treatment can only be required when a person would be a treat to himself or others if he withdrew from treatment. Medical confinement is even harder to require, as it should be. That's particularly true in the case of psychological disorders. Remember that even though _One Flew Over the Cuckoo's Nest_ was fiction, the kinds of abuses upon which it was based weren't.
Second, it's important to know that anti-psychotic and anti-manic drugs are not panaceas. People taking high doses of these drugs are more functional than they would be without them, but do not necessarily feel better, and often feel worse. In the absence of a threat to others, medical treatment is provided for the benefit of the individual being treated, not for the benefit of society. That's why we let cancer patients die when they say they don't want treatment any more, and that's why we let mentally ill people go off their meds.
Oh, yeah. This one's got false arrest written all over it, particularly if the young man was wearing a medicalert bracelet or carry medical id tags.
First thing the parents have got to do is...find a lawyer...hmmm...
Whadaya mean "need a fusion reactor"? Our design plan is that the CPU will be so hot and so compressed that it will serve as its own fusion reactor! Intel and AMD are already on board with this. If it weren't for Transmeta, we'd have a perfect plan for world domination! (Think of the games, too! Microsoft Reactor Simulator 2000 Rad...)
Actually, you're wrong. India is going through a huge period of economic growth throughout its economy. In this, it is replaying a pattern very like the other industrializing countries of the world. It appears to you and me that India is a shambles, but that isn't because the economy is doing poorly, but because it started out doing so much worse.
Most countries go through an extended mercantilist period during their early mass industrialization. During that period, wages in the industrializing country are typically quite low becuase the coutry's currency is artificially depressed. During that period, the country's industrial production skyrockets. Since consumers in the country buy their own products with their own currency, the irrational pricing structure of their industry's exports doesn't affect them, and they act as an internal gate which forces the quality of their exports up.
Eventually, however, growth leads to major industries being unable to provide for their own production with local acquired raw materials. At that point, prices of locally produced products start to reflect the relative level of the currency: foreign raw materials must be bought in foreign currency, which raises the prices of the finished goods into which they are made. That triggers a sharp round of inflation, which leads to a more restrictive currency policy. The price difference between finished good produced in country and those produced abroad gradually shrinks, due to this pressure.
To see this pattern in action, you can go back to Japan in the fifties through the eighties, S. Korea since the eighties, and India now. Alternatively, you can go back to the United State in the late nineteenth century, or to the great European powers in the early nineteenth century.
Europe and the United States managed to extend the period during which they could pursue a mercantilist policy somewhat longer by maintaining a captive market to which finished goods could be exported and from which raw materials could be imported in the local currency. The European powers did this by maintaining colonial markets in Asia, Africa, and, to a lesser extent, the Americas. The Americans settled our West, which became a huge source of raw materials for our East coast industries. The captive markets allowed the industrial base to continue to acquire raw materials at a disproportionately low price.
Schumpeterian equilibrium may well apply to an economy which is dependent on a influx of externally produced raw materials balanced by an egress of internally produced finished goods. That's not the case for economies in their earlier stages of industrialization and development. I don't know how long it will take for India to reach that state, but given the combination of destitution and size of her population, I wouldn't be inclined to expect her government to adopt less mercantilist policies any time soon. It's not rational to do so.
I call bullshit on this, Chumley. Permissions and account handling didn't change at all from NT to XP. So either you're lying or you confused the changes in the domain forest structure when you moved to AD from the NT domain auth model.
I wish I could believe the latter was more likely, but I regret to say that I consider the former more probable. You see, you confuse an Office major vesion migration with an NT major version migration, and any sysadmin worth his or her salt would know that those are very different beasts. Thus, I conclude that you're making your story up.
Never served on a jury, have you? In fact, I'd wager you've never even been into a courthouse. I think that you'd rethink that statement the same way Avi Rubin rethought what he thought about poll watchers if you ever had.
But what you're ignoring is that the CSS providers spent money to create those proprietary standards. If the OSS community wants access to them, then somebody is going to have convince the CSS vendors to yield them up.
Want full access to the Windows source for all people? Microsoft is a business, and, for a suitable sum of money, could be convinced to part with it. (Under US law, the power of eminent domain would certainly suffice to require that.) Go to your friendly local government and convince them to buy the Windows source from Microsoft. At nominal rates of return on investment, I figure that it will only cost on the order of US$250,000,000,000. Maybe a little more...say a round trillion dollars US.
Then again, maybe that isn't feasible. But that's what asking for a "fair shake" would require. If it isn't going to happen, that is not the fault of the court. It's the fault of the polity that it doesn't want to spend that kind of money.
Actually, one of the key points where EU antitrust law is unsettled is what the purpose of antitrust law is.
The US antitrust system is older, and so there are a bunch of court cases which establish that the Sherman Act protects consumers, but makes no attempt to protect competitors. It only provides a means for competitors to sue after the fact and recover triple damages. The courts have consistently decided that there is no direct protection provided to competitors or competition; if the monopolist can show a benefit to consumers, then the court supports the right of the monopolist to sell.
EU law is much less well-established, and it isn't clear what the law means. In fact, the biggest single point which I expect to see argued in front of the Court of the First Instance and the ECJ is whether only consumers are protected by the act, or if competitors are protected as well. If the former, then the EC will lose this case, fine or no fine, and Microsoft will set a huge precedent protecting the company's ability to bundle new features into Windows. If the latter, then the feature extension Microsoft has engaged in over the years will be blocked in Europe.
Uhh...no. Although I'm very angry with the Slashdot editors for posting this particular lead, this is blatant infringement of the most egregious form. Fair use only defends people who violate copyright in small (e.g. by printing a limited passage with full attribution) or who have some other significant difference from the original (e.g. a parody is almost be definition infringing, but protected by fair use.) Including a whole text verbatim for the purpose of avoiding the tariff a copyright holder has set is not fair use, whether or not the author benefits from it.
The GP is correct: ALL GSM phones have a SIM. It is a requirement of the spec for GSM that a subcriber be able to remove and replace his or her subscriber information module (SIM) and transfer it. That creates and fosters competition among the carriers, because once you have a phone, you can use it to access the network of any GSM carrier.
Well, no. Although your particular machanism may remain private -- although I wouldn't, and don't, bet on it -- the odds that any given dictionary attack will steal some user's password are very high. And the odds that some dictionary attack will eventually steal yours are also very high.
Another alternative is to use a random pronounceable word generator for your language of choice. Entropy/character is lower than that of the first character of random free text, but higher than that of frequently memorized text. (How many Americans who use first letters of a phrase come up with "Fsasya,ofbfutcann" or some initial segment thereof? Answer: lots of people, since they're using the first letters of the Gettysburg Address, which they were required to memorize at school.) Problem is,
first letters of the words in a phrase is vulnerable to dictionary attacks against the dictionary of lyrics of popular songs, etc.
Treason! There is no disciplinary termination -- that would imply that the Computer was not properly managing the affairs of the citizens it serves!
Report to the Bright Future Reeducation immeidately for Preventive Clone Activation Sequence initiation.
The Computer is your friend, and only seeks the best interests of you and Alpha Complex. Have a nice day!
Now that is a statement of intent offered by a person who had both the authority and knowledge to make it. Unfortunately, it doesn't help IBM's case very much, and may, in fact, hurt it.
First, that's in a letter from AT&T to IBM. It's already established that IBM's terms were somewhat less stringent than everybody else's, and the stuff I've been talking about is RCU, which was taken from Sequent's Dynix, not IBM's AIX. I've always assumed that the sentence supposedly added to clause 2.01 clarified that the IBM interpretation applied to all licensees, but I've never seen it, and, without seeing it, can't tell.
Even if it does say that, the IBM interpretation is much less friendly to IBM in the SCO case than many people seem to think. AT&T disclaims ownership of modifications and derivative works prepared for IBM, but doesn't disclaim all ownership of those works. It only says that they are owned as a whole by IBM. That would mean that IBM didn't need AT&T's permission to disseminate, for instance, AIX as a whole to a third party who also had a license to use System V Unix. It says nothing about whether IBM had the right to disseminate a piece of AIX to a third party who had NO license to use Unix.
Even more troubling, the second sentence clearly retains ownership of any code taken from Unix itself. It would not be unreasonable to take that to mean that if any of the allegedly novel Dynix source was taken letter for letter from Unix (e.g. if it contained the line for (; *name; name++)), then AT&T still held copyright to that line. In that case, dissemenation of that line would constitute an infringement of AT&T's rights.
What clear statement of intent? The $echo article is no statement of intent: it is not a deposition; it is not presented over the signature of a corporate officer, nor on the letterhead of an attorney representing AT&T. It has exactly as much legal force as my trolling here does: none.
But that doesn't say it's wrong, either, merely that it isn't binding. If you want to know why I'm particularly skeptical about it, here's a cool test for you. Run down to Novell's site, and troll through the documents they ever so helpfully put up there. Look for the magical "last sentence added to section 2.01". Tell me if you find it, because I didn't. I looked for it, because that was the first piece of evidence that Novell had ever presented which in any way called into question SCO's claim to ownership. I find no references to it except in the pdf that was sent to SCO to intimidate them.
I think that Novell is playing the same kind of game for PR the SCO is. I don't trust them, and I don't believe that their contract says what they would like you to believe it says. Otherwise, it would ahve leaked, and it has not done so.
No, I'm aware of the $echo magazine article. It's irrelevant.
First, it is only a magazine article, and is worth as much as the paper that the electronic version is printed on. It's not a contract; it refers to a clause in the contract. It doesn't even quote the clause.
Second, the most plausible interpretation of "no ownership" is just that: "no ownership" -- you have the right to distribute the software without paying us. That says nothing about other contractual limitations on your right to distribute the software. Just because AT&T claimed no ownership does not mean that Sequent got an exception from the derivative work license terms from AT&T.
Nonsense. They list files and lines of files, and show, via the RCU example, exactly how the content of two of those files directly increased the value of Linux to the detriment of their own business. There are only two questions: whether Sequent got an exception from the derivative work license terms from AT&T, and whether Novell's case holds. I've seen no evidence from IBM to support the former, and what I've seen out of Novell is not at all convincing on the latter. That will hang on what the various people involved remember meaning when they signed the contract.
HTML breaks
Necessary. Sufficient?
Oops! Missed a close tag
Engineering is all about tolerances and modes of failure. If I design my car to be able to take a fifteen mph front end collision, and you drive into a wall at thirty, I'm not responsible, and my E&O won't wind up paying out.
Currently, software is built in a craft/guild model: senior developers (masters) teach junior developers (journeymen) who've reached a certain level of expertise. Interns (apprentices) are drafted into the profession and groomed into junior devs. There is a widely held notion of subjective quality, and we can recognize a masterwork, but we can't quantify what it takes to generate one.
Software engineering will become a true engineering discipline only when there is an objective measure of defect level and an objective notion of what constitutes an adequately circumscribed operating environment. Once we have adequate definitions of those things, though, software production will become industrialized almost immediately.
Did you read the Security Focus article? It explicitly stated both that Blaster was not related to the blackout and that SF had been one of the first publications to extend the hypothesis that they had been related.
In short, the Microsoft bashers were wrong -- and at least Security Focus had the guts to acknowledge it.