I see your point, but this is one of the reasons "mil-spec" parts cost so much. You get what you ordered, not what some vendor decided you might like instead. And those small differences add up. Cable locatons affect air flow and thus cooling. A very minor component change, such as a minor network chipset change, can swap network ports. And failing to notify the customer of the change is a very, very bad practice, especially in a sensitive system, because when a technician opens it up and two otherwise identical systems with the same model number don't match, foul play is going to be suspected.
If the Mac subversion client does so, then the client is not the actual Subversion source code, it's been heavily modified. I've not tested it: is it a command line client? If so, what does "svn --version" say?
Now try it with several hundred passwords. Based on the results of Alec Moffett's old Crack program for DES, I'd suspect that you'd find roughly 10% of all passwords quite easily, with "love" leading the list.
It's no worse than Subversion's insistence on storing user passwords for any protocol but SSH public keys in a local plaintext file.
Do not *EVER* allow a Subversion system to use the same passwords as the user system, and if you have access to the user's accounts, run a check of their stored Subversion passwords to make sure they didn't use their same password somewhere else as for their local user account.
Even if Novell is above board, Microsoft historically and traditionally isn't. Go read their behavior with the ODF and OOXML format lobbying for very recent examples of them deliberately and covertly manipulating standards bodies for intellectual property benefits at the expense of the market and the consumer.
So while Novell doesn't own Samba, they do own a major Linux distribution. Not owning Samba doesn't prevent exactly the kind of manipulation I've described, or even more fascinating ones like integrating other Microsoft patented technologies into it. Mr. Stallman correctly foresaw this kind of behavior: when it got done anyway by Novell and Microsoft in this deal, the FSF correctly strengthened the GPL to prevent a repeat.
Also, Noveel was not based on open source or freeware. While SuSE is a stable commercial release, Novell didn't write it. So their highest management may simply not have accepted the advice, and may still override the beliefs and practices, of their open source developers such as Jeremy. And it's obvious that they did!
Novell wouldn't have had to attack the GPL: they could have, and tried to, follow the letter of the law to protect their customers and promote partnerships with Microsoft for software compatibility, and leave the other developers on whom their work resides out in the cold. Novell tried stunts like this before, with various Netware partnerships designed to freeze out competing projects. It failed in the long run there, and it should fail in the long run here. But in the meantime, we have to be cautious it doesn't screw up other projects.
Jeremy's resignation did that: His presence at Novell would have lent them a lot of sway in integrating new, SuSE-licensed, non-GPL software and packages. And Novell has shown their willingness to ride a lawsuit to death. (Witness the SCO craziness.) If the deal hadn't gotten so much negative attention from people who work as developers, and hadn't obviously pissed off some of their best people, I'm sure we'd see Novell today following a much more patent encumbered path. It wouldn't be a direct attack on the GPL, any more than Tivo's behavior was a direct attack. But the patent poisoning and the careful locking out of the user using their own tools are similar aspects of the same problem.
This helps fix the next generation: dating your average internet addicted geek will likely prevent such a girl (or boy, for that matter!) from ever wanting to have children.
> Heck, I've lived in the mountains of Peru, and I would argue that Peruvians are still designing their road ways more by trial and error than through any sort of rigorous engineering. Seriously, you absolutely wouldn't believe what passes for a road in the Andes.
And yet his subsequent discovery of the Fountain of Youth has permitted him to continue in the role, publishing duplicate stories on Slashdot.
More seriously, it's an interesting story. Teaching a bunch of extremely sharp, motivated people like MIT students to collaborate on a basic physical task that requires high quality control and doesn't have a lot of shortcuts must have been a fascinating task.
Really? When a murder conviction can take years to prosecute, I don't see how you can avoid the publicity without gag orders that would encourage corruption by allowing the prosecutors to drop cases they don't feel like pursuing.
It happens to doctors who perform abortion on a far, far too frequent basis. And it happens to people accused of sexually harassing children, even in cases where the accusation is complete nonsense. (I'm thinking of a particular acquaintance who taught troubled children and was accused of sexual harassment by one of them, and the dates and times claimed by the child were clearly fraudulent.)
Especially because police make mistakes, witnesses can be threatened into silence, and some "crimes" are those of being a bit ahead of the legal curve on human rights, such as providing marijuana for cancer sufferers.
Novell lost one of the industry leaders in file-sharing, one of the core Samba maintainers and developers. You can't replace such a person with some third-world green-card holding MCSE. You need one of the few dozen people in the world who understand CIFS that well, or one of the half-dozen who understand Samba that well. To call such a person just a "laborer" is like my treating your mother to an intimate and sensual evening of champagne, lobster, and dancing, and calling it a "discussion".
I read Groklaw's analyses with an admittedly jaundiced, but that's because I've been keeping an eye on Microsoft's demonstrably larsenous, fraudulent, and criminal behavior for years. They've been caught repeatedly, they've been convicted on rare occasion though they've usually settled out of court, and they've earned the block of salt that any business deal with them should be examined with.
You see, I've actually worked with GPL software and integrating such proprietary 3rd party utilities for employer requirements. The fear of such manipulation is well-founded: Jeremy resigning from Novell over this mess confirms the suspicions that the deal is rotten.
And you've obviously missed the implications of proprietary binary plug-ins for Samba. To pluck a specific example, authentication and account management integration can be coupled as back-end plug-ins to Samba as non-GPL utilities, directly integrated to the GPL-based Samba or other utilities. The result is similar to that of NVidia's binary kernel blobs, and the SuSE tool "YaST" is already designed to handle just that sort of binary only 3rd-party plugin, even when such plugins have no source, require manual EULA sign-up, violate the software bundling standards of RPM, and even overwrite and replace the GPL, RPM published software.
Now, it gets trickier. Microsoft loves, nay courts and ravishes XML patents. They've tried to build a filesystem on it (WinFS, which failed miserably). They've incorporated such patented XML into numerous of their software packages. So, Novell publishes a Samba plug-in (for example) that uses Microsoft patented, or allegedly patented, technology. It's a separate program: it's not part of Samba itself. But Samba winds up relying on it, and integrating it. And the presence of the patent-protected Novell version discourages reverse engineering it: if you want it, no need for a patent fight, just get Novell's servers and clients. This poisons the well for developers, and provides a pleasantly useful cover for Microsoft of "inter-operability" but only with those companies that have drunk the Microsoft patent Kool-aid.
And if you don't think that's likely, take a similarly suspicious look at the history of NTFS drivers and the possibilities for WinFS if Microsoft can ever make it work correctly. Due to WinFS's XML patents, if you're in the US, you won't be able to use open source WinFS drivers without Microsoft permission, at least as near as I can tell from the limited patent history I've seen.
You seem to repeatedly claim that Microsoft and Novell are above board, and we should pay no attention to that man behind the curtain. I've seen that man behind the curtain before: I've even had to fish him out of the gutter, sober him up, and find him a couch to sleep on, and I don't want him at the controls of something as important as GPL licensing.
Whatever happened to fuckedcompany.com, anyway? I rather enjoyed seeing information on that site before I got it under NDA, and pointing it out to the person who insisted on the NDA. (It was particularly sad and funny during layoffs, and an amusing thing to quietly forward the URL to departments about to be visited by the corporate hatchetman.)
You are kidding, right? Even a lawyer or doctor have to testify and turn over records with the right court order. One of the few exceptions is a Catholic priest hearing confession, at least in US law, and the attorneys will pick around the edges of the confessional privileges.
That's funny. In the time I've seen Visual Studio taking to simply start up, especially in a complex environment, the machine might have already been banging into the walls.
More seriously, you've revealed a serious issue, to wit: "Other than the fact that they didn't know how to use the debuggers that come with Visual Studio.net, I really don't think they would of done any better with the same debuggers writing in C under Visual Studio.net."
I agree with you. Actually running the system for a trial run would have revealed the issue. You cannot rely on the cleverest debuggers in the world to replace an actual "smoke test", which clearly didn't happen.
By not wasting their time with 3 times as many lines of C# code? By having a much faster compilation time so tests can be done more quickly? By having free high quality profiling and debugging tools? By being forced to think about what their functions actually pass to each other, instead of waving their hands and saying "then a miracle occurs"? By having such basic errors fail more quickly, rather than after testing and in the field?
Take your pick. I've done some comparisons for simple programs written in C and in C#, though I'm hardly expert in C#, and seen each of these apply at least once.
> It looks like Jeremy is still with Samba so i'm not sure what your attempting to get at with the Google file sharing bit. Novell doesn't own samba.
I was apparently unclear. Novell lost one of their best sources of software to interoperate with Microsoft when they lost Jeremy. They did so when they made a deal that certainly violated the spirit of the GPL which, according to Jeremy, they also violated legally. That actually cost them a lot of their hopes of remaining interoperable.
You've read Groklaw's analyses in a way rather different than I read them. Pamela Jones's offense at the deal has much more to do with the fact that the deal doesn't actually cover *anything* properly, but especially doesn't protect development that may be founded in Novell's work. It's tainted Novell's open source work in a way that makes it potentially unusable without Microsoft persmission.
You also assume that the Novell/Microsoft deal was completely understood by Novell's lawyers in terms of the GPL. Even if they did, nothing prevents Novell from engaging in Microsoft style "embrace and extend", to produce related and patent poisoned programs available as "open source" but encumbered with Microsoft patents. Simply don't release those add-ons under GPL, and create software founded on interactivity with GPL software that is not GPL encumbered. Samba is one of the likeliest candidates for that kind of poisoning: add-on features can be bolted on top of Samba and integrated with it that are not GPL.
Copyright was hardly a fluke. It's a logical outgrowth of the idea that whoever finds something first, owns it and controls its use. Nothing that lasts since the first printings of the Bible more than 500 years ago can be called "a fluke". And the idea that "only physical things have value" is historically nonsensical, given the history of wars of fealty, religious wars, and the common family argument that "mom loved me better than you!".
No, that was "Flying Guy" who wrote the "original, heart-rending scenario".
You seem to keep doing this, citing as fact things that are true only in your memory, and using them as principles for your reasoning. It's a problem I urge you to avoid.
I have kept running off a bit at tangents discussing this with you, because your foundations for your claims are so off-base I can only counter one at a time without running on and on and boring readers. I think it's enough to show that your basic assumptions are off to discount your claim.
The behavior is disturbingly similar to Senator Joe McCarthy's witch hunt for Communists, where he claimed to have a list of Communists working for the State Department but could never produce the list. The senator was eventually discredited, and much of his heroic war background found to be fraudulent as well. But it fanned the flames of anti-communist fervor in the US. And Microsoft today wants to fan the flames of "beware those open source people, they'll steal your work!" fear.
But then Jeremy Allison, the Novell employee who was also one of the primary Samba developers, threw a hissy fit calling the deal illegal and resigned. And Jeremy is a sharp person: I take his opinion seriously.
The end result is that Novell thought they were getting a helpful license deal, and instead shot their own MS services development through the head. I wish I could find Jeremy to buy him the ice cream sundae of his choice. But he's over at Google now: I wonder what they're up to in the filesharing field that we might see in a few years?
Notes about your anonymity discrediting your argument aside, I strongly suggest you look at the Linksys lawsuits, and the results of the SCO lawsuits. The GPL has held up quite well under Microsoft funded attack in SCO's case, where Darl McBride of SCO made exactly your kind of baseless claims about GPL, and the judge repeatedly discarded such claims.
The Linksys lawsuits over their use of a modified glibc without publishing their source code has helped establish the GPL as quite valid in court. The GPL was, in fact, enforced: Linksys now publishes the source code to their modified glibc. So I have no idea where you get the idea that GPL is unenforceable.
And if they're starving or broke, or can't afford the tools of their trade, they produce less art. The goal wasn't to "let the artists starve, they'll produce anyway". It was to "promote progress".
That's not quite fair. Some of us post political opinions here that they dare not voice in their workplace because they're contrary to corporate policy, even posting how to work around such corporate policies. I've done so myself. Anonymity has a long, proud history in American and international politics.
I do agree that it is good to provide sources: I was recently caught myself in a position where I couldn't cite a reference because it would have threatened my anonymity, so it's true that it's sometimes awkward to do.
Slashdot Mirror
I see your point, but this is one of the reasons "mil-spec" parts cost so much. You get what you ordered, not what some vendor decided you might like instead. And those small differences add up. Cable locatons affect air flow and thus cooling. A very minor component change, such as a minor network chipset change, can swap network ports. And failing to notify the customer of the change is a very, very bad practice, especially in a sensitive system, because when a technician opens it up and two otherwise identical systems with the same model number don't match, foul play is going to be suspected.
If the Mac subversion client does so, then the client is not the actual Subversion source code, it's been heavily modified. I've not tested it: is it a command line client? If so, what does "svn --version" say?
Now try it with several hundred passwords. Based on the results of Alec Moffett's old Crack program for DES, I'd suspect that you'd find roughly 10% of all passwords quite easily, with "love" leading the list.
It's no worse than Subversion's insistence on storing user passwords for any protocol but SSH public keys in a local plaintext file.
Do not *EVER* allow a Subversion system to use the same passwords as the user system, and if you have access to the user's accounts, run a check of their stored Subversion passwords to make sure they didn't use their same password somewhere else as for their local user account.
Even if Novell is above board, Microsoft historically and traditionally isn't. Go read their behavior with the ODF and OOXML format lobbying for very recent examples of them deliberately and covertly manipulating standards bodies for intellectual property benefits at the expense of the market and the consumer.
So while Novell doesn't own Samba, they do own a major Linux distribution. Not owning Samba doesn't prevent exactly the kind of manipulation I've described, or even more fascinating ones like integrating other Microsoft patented technologies into it. Mr. Stallman correctly foresaw this kind of behavior: when it got done anyway by Novell and Microsoft in this deal, the FSF correctly strengthened the GPL to prevent a repeat.
Also, Noveel was not based on open source or freeware. While SuSE is a stable commercial release, Novell didn't write it. So their highest management may simply not have accepted the advice, and may still override the beliefs and practices, of their open source developers such as Jeremy. And it's obvious that they did!
Novell wouldn't have had to attack the GPL: they could have, and tried to, follow the letter of the law to protect their customers and promote partnerships with Microsoft for software compatibility, and leave the other developers on whom their work resides out in the cold. Novell tried stunts like this before, with various Netware partnerships designed to freeze out competing projects. It failed in the long run there, and it should fail in the long run here. But in the meantime, we have to be cautious it doesn't screw up other projects.
Jeremy's resignation did that: His presence at Novell would have lent them a lot of sway in integrating new, SuSE-licensed, non-GPL software and packages. And Novell has shown their willingness to ride a lawsuit to death. (Witness the SCO craziness.) If the deal hadn't gotten so much negative attention from people who work as developers, and hadn't obviously pissed off some of their best people, I'm sure we'd see Novell today following a much more patent encumbered path. It wouldn't be a direct attack on the GPL, any more than Tivo's behavior was a direct attack. But the patent poisoning and the careful locking out of the user using their own tools are similar aspects of the same problem.
This helps fix the next generation: dating your average internet addicted geek will likely prevent such a girl (or boy, for that matter!) from ever wanting to have children.
> Heck, I've lived in the mountains of Peru, and I would argue that Peruvians are still designing their road ways more by trial and error than through any sort of rigorous engineering. Seriously, you absolutely wouldn't believe what passes for a road in the Andes.
Kind of like most Perl, eh?
And yet his subsequent discovery of the Fountain of Youth has permitted him to continue in the role, publishing duplicate stories on Slashdot.
More seriously, it's an interesting story. Teaching a bunch of extremely sharp, motivated people like MIT students to collaborate on a basic physical task that requires high quality control and doesn't have a lot of shortcuts must have been a fascinating task.
Really? When a murder conviction can take years to prosecute, I don't see how you can avoid the publicity without gag orders that would encourage corruption by allowing the prosecutors to drop cases they don't feel like pursuing.
It happens to doctors who perform abortion on a far, far too frequent basis. And it happens to people accused of sexually harassing children, even in cases where the accusation is complete nonsense. (I'm thinking of a particular acquaintance who taught troubled children and was accused of sexual harassment by one of them, and the dates and times claimed by the child were clearly fraudulent.)
Especially because police make mistakes, witnesses can be threatened into silence, and some "crimes" are those of being a bit ahead of the legal curve on human rights, such as providing marijuana for cancer sufferers.
And some folks wonder why Slashdot posters do so anonymously, or use aliases.
Novell lost one of the industry leaders in file-sharing, one of the core Samba maintainers and developers. You can't replace such a person with some third-world green-card holding MCSE. You need one of the few dozen people in the world who understand CIFS that well, or one of the half-dozen who understand Samba that well. To call such a person just a "laborer" is like my treating your mother to an intimate and sensual evening of champagne, lobster, and dancing, and calling it a "discussion".
I read Groklaw's analyses with an admittedly jaundiced, but that's because I've been keeping an eye on Microsoft's demonstrably larsenous, fraudulent, and criminal behavior for years. They've been caught repeatedly, they've been convicted on rare occasion though they've usually settled out of court, and they've earned the block of salt that any business deal with them should be examined with.
You see, I've actually worked with GPL software and integrating such proprietary 3rd party utilities for employer requirements. The fear of such manipulation is well-founded: Jeremy resigning from Novell over this mess confirms the suspicions that the deal is rotten.
And you've obviously missed the implications of proprietary binary plug-ins for Samba. To pluck a specific example, authentication and account management integration can be coupled as back-end plug-ins to Samba as non-GPL utilities, directly integrated to the GPL-based Samba or other utilities. The result is similar to that of NVidia's binary kernel blobs, and the SuSE tool "YaST" is already designed to handle just that sort of binary only 3rd-party plugin, even when such plugins have no source, require manual EULA sign-up, violate the software bundling standards of RPM, and even overwrite and replace the GPL, RPM published software.
Now, it gets trickier. Microsoft loves, nay courts and ravishes XML patents. They've tried to build a filesystem on it (WinFS, which failed miserably). They've incorporated such patented XML into numerous of their software packages. So, Novell publishes a Samba plug-in (for example) that uses Microsoft patented, or allegedly patented, technology. It's a separate program: it's not part of Samba itself. But Samba winds up relying on it, and integrating it. And the presence of the patent-protected Novell version discourages reverse engineering it: if you want it, no need for a patent fight, just get Novell's servers and clients. This poisons the well for developers, and provides a pleasantly useful cover for Microsoft of "inter-operability" but only with those companies that have drunk the Microsoft patent Kool-aid.
And if you don't think that's likely, take a similarly suspicious look at the history of NTFS drivers and the possibilities for WinFS if Microsoft can ever make it work correctly. Due to WinFS's XML patents, if you're in the US, you won't be able to use open source WinFS drivers without Microsoft permission, at least as near as I can tell from the limited patent history I've seen.
You seem to repeatedly claim that Microsoft and Novell are above board, and we should pay no attention to that man behind the curtain. I've seen that man behind the curtain before: I've even had to fish him out of the gutter, sober him up, and find him a couch to sleep on, and I don't want him at the controls of something as important as GPL licensing.
Whatever happened to fuckedcompany.com, anyway? I rather enjoyed seeing information on that site before I got it under NDA, and pointing it out to the person who insisted on the NDA. (It was particularly sad and funny during layoffs, and an amusing thing to quietly forward the URL to departments about to be visited by the corporate hatchetman.)
You are kidding, right? Even a lawyer or doctor have to testify and turn over records with the right court order. One of the few exceptions is a Catholic priest hearing confession, at least in US law, and the attorneys will pick around the edges of the confessional privileges.
That's funny. In the time I've seen Visual Studio taking to simply start up, especially in a complex environment, the machine might have already been banging into the walls.
.net, I really don't think they would of done any better with the same debuggers writing in C under Visual Studio .net."
More seriously, you've revealed a serious issue, to wit: "Other than the fact that they didn't know how to use the debuggers that come with Visual Studio
I agree with you. Actually running the system for a trial run would have revealed the issue. You cannot rely on the cleverest debuggers in the world to replace an actual "smoke test", which clearly didn't happen.
By not wasting their time with 3 times as many lines of C# code? By having a much faster compilation time so tests can be done more quickly? By having free high quality profiling and debugging tools? By being forced to think about what their functions actually pass to each other, instead of waving their hands and saying "then a miracle occurs"? By having such basic errors fail more quickly, rather than after testing and in the field?
Take your pick. I've done some comparisons for simple programs written in C and in C#, though I'm hardly expert in C#, and seen each of these apply at least once.
> It looks like Jeremy is still with Samba so i'm not sure what your attempting to get at with the Google file sharing bit. Novell doesn't own samba.
I was apparently unclear. Novell lost one of their best sources of software to interoperate with Microsoft when they lost Jeremy. They did so when they made a deal that certainly violated the spirit of the GPL which, according to Jeremy, they also violated legally. That actually cost them a lot of their hopes of remaining interoperable.
You've read Groklaw's analyses in a way rather different than I read them. Pamela Jones's offense at the deal has much more to do with the fact that the deal doesn't actually cover *anything* properly, but especially doesn't protect development that may be founded in Novell's work. It's tainted Novell's open source work in a way that makes it potentially unusable without Microsoft persmission.
You also assume that the Novell/Microsoft deal was completely understood by Novell's lawyers in terms of the GPL. Even if they did, nothing prevents Novell from engaging in Microsoft style "embrace and extend", to produce related and patent poisoned programs available as "open source" but encumbered with Microsoft patents. Simply don't release those add-ons under GPL, and create software founded on interactivity with GPL software that is not GPL encumbered. Samba is one of the likeliest candidates for that kind of poisoning: add-on features can be bolted on top of Samba and integrated with it that are not GPL.
Copyright was hardly a fluke. It's a logical outgrowth of the idea that whoever finds something first, owns it and controls its use. Nothing that lasts since the first printings of the Bible more than 500 years ago can be called "a fluke". And the idea that "only physical things have value" is historically nonsensical, given the history of wars of fealty, religious wars, and the common family argument that "mom loved me better than you!".
No, that was "Flying Guy" who wrote the "original, heart-rending scenario".
You seem to keep doing this, citing as fact things that are true only in your memory, and using them as principles for your reasoning. It's a problem I urge you to avoid.
I have kept running off a bit at tangents discussing this with you, because your foundations for your claims are so off-base I can only counter one at a time without running on and on and boring readers. I think it's enough to show that your basic assumptions are off to discount your claim.
Look for the keywords "Linux violates Microsoft patents" on Google. You'll find plenty of references like this: http://money.cnn.com/magazines/fortune/fortune_archive/2007/05/28/100033867/.
The behavior is disturbingly similar to Senator Joe McCarthy's witch hunt for Communists, where he claimed to have a list of Communists working for the State Department but could never produce the list. The senator was eventually discredited, and much of his heroic war background found to be fraudulent as well. But it fanned the flames of anti-communist fervor in the US. And Microsoft today wants to fan the flames of "beware those open source people, they'll steal your work!" fear.
But then Jeremy Allison, the Novell employee who was also one of the primary Samba developers, threw a hissy fit calling the deal illegal and resigned. And Jeremy is a sharp person: I take his opinion seriously.
The end result is that Novell thought they were getting a helpful license deal, and instead shot their own MS services development through the head. I wish I could find Jeremy to buy him the ice cream sundae of his choice. But he's over at Google now: I wonder what they're up to in the filesharing field that we might see in a few years?
Notes about your anonymity discrediting your argument aside, I strongly suggest you look at the Linksys lawsuits, and the results of the SCO lawsuits. The GPL has held up quite well under Microsoft funded attack in SCO's case, where Darl McBride of SCO made exactly your kind of baseless claims about GPL, and the judge repeatedly discarded such claims.
The Linksys lawsuits over their use of a modified glibc without publishing their source code has helped establish the GPL as quite valid in court. The GPL was, in fact, enforced: Linksys now publishes the source code to their modified glibc. So I have no idea where you get the idea that GPL is unenforceable.
And if they're starving or broke, or can't afford the tools of their trade, they produce less art. The goal wasn't to "let the artists starve, they'll produce anyway". It was to "promote progress".
That's not quite fair. Some of us post political opinions here that they dare not voice in their workplace because they're contrary to corporate policy, even posting how to work around such corporate policies. I've done so myself. Anonymity has a long, proud history in American and international politics.
I do agree that it is good to provide sources: I was recently caught myself in a position where I couldn't cite a reference because it would have threatened my anonymity, so it's true that it's sometimes awkward to do.