I thought that the compiled Java would remain compatible with the bytecode format used by previous versions.
The bytecode format is still identical. The version number on the file was incremented because classes that depend on assertions could break badly if used on a JVM that doesn't implement them (which is actually a feature of a few classes in java.lang, not a bytecode interpreter feature). Use the '-target' commands recommended by the other posters whenever producing code that doesn't rely on one of the new features.
Carrying on down this path much farther, you become able to argue that C is just syntactic sugar for assembly language, and therefore assembly language is just as good.
grew up in the days before the Internet. Was I exposed to pornography? Yes. In fact most, if not all, of my contemporaries were exposed to pornography. Where did this pornography come from? Well believe it or not, there was a time when pornographic material came in these things called magazines. That's right - magazines!
I don't know about where you come from, but where I am, the magazines tend to be a _lot_ softer than most of what's on the Internet. In fact, my belief is that a _very_ large proportion of Internet porn is not legally publishable in the UK. I don't know what Australian "decency" laws are like, but the situation may be the same there... the Internet may be seen as a method people use to access pornography that is illegal in the country.
IANAL, but I believe if there is no explicit contract, "work for hire" is assumed whenever money changes hands for IP related work.
It would be up to the developer to show a contract (a verbal one would probably be sufficient, as long as Connolly didn't contest it) that stated he retained rights to the work he developed.
The developer in question is not an employee of Furthermore, so this is not "distribution within a single company" which is explicitly allowed. If anything, it was a distribution between a contractor/contractee, which further in the GPL FAQ you will see is considered distribution.
There is a problem with this argument. The developer is not the owner of the copyright in question. This means that his distribution of it to Connoly cannot grant anyone any rights to it under the GPL. Only Connoly has the legal power to do that, as the copyright was assigned to him.
This would be equivalent to me taking a copy of the Windows source code that was leaked, integrating it into a "Windows subsystem" of Linux, then sending a copy of it to someone and claiming that because it was distributed under the GPL, anyone is now allowed to distribute it. I don't own the copyright to Windows, so I cannot grant you that right.
The Mambo developer doesn't own the copyright to the modification he made for Connoly, so nothing he did can grant anyone rights over it. Only things that Connoly does.
As far as the 'validity' of the GPL FAQ, while I'm sure it is not error-free, it comes from the same group of legal scholars (and their assistants) that gave us the GPL. While I am sure not perfect, they know it better then any of us. I readily trust their interpretation better then anybody except Moglen's.
You have to consider, while reading it, that the answers are to questions about generic situations. Any small detail and assumption, which may or may not have been explicitly stated (it's written with the intention of avoiding legalese, therefore a lot of assumptions aren't stated), may change the answer in a specific situation.
This wouldn't actually help a huge amount. It would stop any future distribution from being copyright infringement, but he would still be liable for any infringement the court decided he was responsible for up to the moment he received the copy.
Interesting situation. I think this is acceptable under the GPL, but note:
1. If BigCorp ever distribute a copy of her derivitive work back to her, they are distributing it under the terms of the GPL and are then not allowed to enforce any contracts that further restrict Alice's actions under the GPL, so she can distribute it on in this case.
2. This is all subject to the courts' interpreatation of "within a single organisation". I'm aware that they may extend this to include contractors -- British courts certainly have.
Redundant? I don't see any other posts saying the same thing. While I don't agree with 'troll', it's clearly not 'reduntant' or 'offtopic'. 'Overrated', if you must mod it down, although I'd be inclined to just leave it, if I ever got mod points these days.
If a fellow doesn't know how to drive, and he depends on his GF as a driver, then how could he continue to get around after breaking up with his GF?
To be fair, you'll have to acknowledge that the problem could be fixed by him getting a car that was easier to drive, although maybe not quite as fast, and possibly a tiny amount more expensive.
I'm firmly of the opinion that the difficulty of "rocket science" has been built up in the public consciousness so high that, by now, even building something like SpaceShipOne from first design principles through to succesful flights isn't rocket science.
I'd much rather have an OS in a mission critical environment at least tell me something when it crashes (linux: oops!) than give me a BSOD full of incomprehensible characters.
Windows BSODs contain very similar information to a Linux kernel oops: a brief description of what caused the crash, details of the process that was running at the time, and details of loaded drivers.
The only thing that's missing is the stack trace and disassembly of the executing code presented by Linux. These items are only useful if you have the source code to the module where the crash occurred, which is unlikely with Windows.
I understand there are debugging versions of the kernel available; these might give more information.
I just came to this story from metamoderation. There are a _lot_ of junk moderations of perfectly valid posts as offtopic or troll at the moment. And one with "fp!" as insightful. I think it's a new form of trolling.
By your definition, there have been no innovative operating systems in perhaps 40 years.
Not at all. The last 40 years has seen the advent of many useful OS level features, including the "everything's a file" abstraction (pioneered by Unix and Plan9), standardised security models (Multics, Unix, and others), real time scheduling (not sure where this originated?), journaling file systems, clustering operations, distributed filesystems, instrumentation and access auditing, multithreading with shared memory while maintaining protection between separate applications, and probably a whole host of other such features.
I can't address all of your points, because I don't know enough about all of the areas you touch, but:
Open Mosix Transparent process migration, intended for clustering.
Has been available in various systems since the 1980s. See, for example, "Sprite".
UML Self hosted virtual machines.
A facility that was possible, I believe, with IBM mainframe systems back as far as the late seventies.
RTLinux Realtime microkernel/macrokernel work. Hell, it _is_ patented.
I don't know what the patents are for, but it sure as hell isn't the first kernel (micro or otherwise) to offer real time facilities.
ReiserFS Filesystem based on dancing trees, with a plugin archtecture.
First time I've seen it in a filesystem, but structuring data in trees has been used in database systems since the late sixties. ReiserFS is only an incremental improvement over most of these systems.
ZisoFS Transparant handling of compressed ISO9660 filesystems.
Transparently compressed files have been a feature of many filesystems over the years. ZisoFS is unusual in that it retrospectively adds the feature to an already existing filesystem, but then ISO9660 was designed to be easily extendible, anyway.
InterMezzo Distributed filesystem, with network interrupt transparacy.
I believe this was one of the design goals of the Andrew filesystem, developed at CMU during the early nineties (?).
As to "Adeos" and "Seperate LLC stack", I'm not sure exactly what these achieve, so I can't really comment on them. I truly doubt, however, that they are utterly unique and new, except perhaps in minor implementation details.
that MS employee's were not permitted to participate in OSS projects?
Why do you think that? Have you seen a statement regarding this? I'd like to see what exactly was said, if there is one.
I would personally expect MS to have issues regarding employees contributing to _some_ OSS projects. One particular aspect: I don't suspect they would permit any employees to sign the FSF copyright assignment forms. These grant a lot of rights to the FSF, which might potentially interfere with rights to work the employee does for MS.
I know my company reviewed it when I was considering doing some work on gcj, and decided that we couldn't accept its terms.
What does allegedly "secretly manipulating an industry-wide cooperative standard-setting body" have to do with the patent system?
I haven't read the book, but they're probably arguing for a system where failing to disclose a patent when proposing somebody else uses an invention it covers renders it invalid. This case (if my memory of the details is correct) is a good example of where such a law would help.
Judges aren't lawyers. The two are completely separate professions that happen to involve very similar training and who work together frequently.
That's like saying that IT project managers are programmers. The two clearly have a lot of domain specific knowledge in common, and programmers often become project managers later on in their careers, but they are entirely separate and distinct jobs.
The laws in question are almost certainly "common law"; that is they were never actually written in governmental acts, but arose through the decisions of judges (and juries) in past similar cases.
And, yes, juries are a large part of the problem. They're too easy to sway with emotional arguments, and often award compensation that is substantially too high. They're also often left to decide on matters that they aren't qualified for, particularly when dealing with laws that involve phrases like "a reasonable person".
OK, there is a very important difference between copyright and patents, you know. Copyright only applies to direct copying. Patents apply even to independent reinvention of the same concept, which is a little dubious, morally speaking, if you ask me.
I don't think it does. Linux is not particularly innovative. Almost all of its features were previously included in another operating system. I can't think of a single thing it does that ought to be eligible for a patent, even in a world where software can be patented.
Slashdot Mirror
Open source Java.
Problem solved?
I thought that the compiled Java would remain compatible with the bytecode format used by previous versions.
The bytecode format is still identical. The version number on the file was incremented because classes that depend on assertions could break badly if used on a JVM that doesn't implement them (which is actually a feature of a few classes in java.lang, not a bytecode interpreter feature). Use the '-target' commands recommended by the other posters whenever producing code that doesn't rely on one of the new features.
Carrying on down this path much farther, you become able to argue that C is just syntactic sugar for assembly language, and therefore assembly language is just as good.
Face it. Syntax matters.
grew up in the days before the Internet. Was I exposed to pornography? Yes. In fact most, if not all, of my contemporaries were exposed to pornography. Where did this pornography come from? Well believe it or not, there was a time when pornographic material came in these things called magazines. That's right - magazines!
I don't know about where you come from, but where I am, the magazines tend to be a _lot_ softer than most of what's on the Internet. In fact, my belief is that a _very_ large proportion of Internet porn is not legally publishable in the UK. I don't know what Australian "decency" laws are like, but the situation may be the same there... the Internet may be seen as a method people use to access pornography that is illegal in the country.
Was that his modified version, or the original unmodified one, though?
IANAL, but I believe if there is no explicit contract, "work for hire" is assumed whenever money changes hands for IP related work.
It would be up to the developer to show a contract (a verbal one would probably be sufficient, as long as Connolly didn't contest it) that stated he retained rights to the work he developed.
The developer in question is not an employee of Furthermore, so this is not "distribution within a single company" which is explicitly allowed. If anything, it was a distribution between a contractor/contractee, which further in the GPL FAQ you will see is considered distribution.
There is a problem with this argument. The developer is not the owner of the copyright in question. This means that his distribution of it to Connoly cannot grant anyone any rights to it under the GPL. Only Connoly has the legal power to do that, as the copyright was assigned to him.
This would be equivalent to me taking a copy of the Windows source code that was leaked, integrating it into a "Windows subsystem" of Linux, then sending a copy of it to someone and claiming that because it was distributed under the GPL, anyone is now allowed to distribute it. I don't own the copyright to Windows, so I cannot grant you that right.
The Mambo developer doesn't own the copyright to the modification he made for Connoly, so nothing he did can grant anyone rights over it. Only things that Connoly does.
As far as the 'validity' of the GPL FAQ, while I'm sure it is not error-free, it comes from the same group of legal scholars (and their assistants) that gave us the GPL. While I am sure not perfect, they know it better then any of us. I readily trust their interpretation better then anybody except Moglen's.
You have to consider, while reading it, that the answers are to questions about generic situations. Any small detail and assumption, which may or may not have been explicitly stated (it's written with the intention of avoiding legalese, therefore a lot of assumptions aren't stated), may change the answer in a specific situation.
This wouldn't actually help a huge amount. It would stop any future distribution from being copyright infringement, but he would still be liable for any infringement the court decided he was responsible for up to the moment he received the copy.
Interesting situation. I think this is acceptable under the GPL, but note:
1. If BigCorp ever distribute a copy of her derivitive work back to her, they are distributing it under the terms of the GPL and are then not allowed to enforce any contracts that further restrict Alice's actions under the GPL, so she can distribute it on in this case.
2. This is all subject to the courts' interpreatation of "within a single organisation". I'm aware that they may extend this to include contractors -- British courts certainly have.
Redundant? I don't see any other posts saying the same thing. While I don't agree with 'troll', it's clearly not 'reduntant' or 'offtopic'. 'Overrated', if you must mod it down, although I'd be inclined to just leave it, if I ever got mod points these days.
If a fellow doesn't know how to drive, and he depends on his GF as a driver, then how could he continue to get around after breaking up with his GF?
To be fair, you'll have to acknowledge that the problem could be fixed by him getting a car that was easier to drive, although maybe not quite as fast, and possibly a tiny amount more expensive.
To use an analogy I found in my 'fortunes' file... that's a task very much like herding cats.
I'm firmly of the opinion that the difficulty of "rocket science" has been built up in the public consciousness so high that, by now, even building something like SpaceShipOne from first design principles through to succesful flights isn't rocket science.
the ISS has been orbiting around 250 miles up
Err... yeah, 250 miles ~= 400km, so qualifies as LEO by the definition the parent gave.
I'd much rather have an OS in a mission critical environment at least tell me something when it crashes (linux: oops!) than give me a BSOD full of incomprehensible characters.
Windows BSODs contain very similar information to a Linux kernel oops: a brief description of what caused the crash, details of the process that was running at the time, and details of loaded drivers.
The only thing that's missing is the stack trace and disassembly of the executing code presented by Linux. These items are only useful if you have the source code to the module where the crash occurred, which is unlikely with Windows.
I understand there are debugging versions of the kernel available; these might give more information.
Seriosuly now, wtf mods?
I just came to this story from metamoderation. There are a _lot_ of junk moderations of perfectly valid posts as offtopic or troll at the moment. And one with "fp!" as insightful. I think it's a new form of trolling.
By your definition, there have been no innovative operating systems in perhaps 40 years.
Not at all. The last 40 years has seen the advent of many useful OS level features, including the "everything's a file" abstraction (pioneered by Unix and Plan9), standardised security models (Multics, Unix, and others), real time scheduling (not sure where this originated?), journaling file systems, clustering operations, distributed filesystems, instrumentation and access auditing, multithreading with shared memory while maintaining protection between separate applications, and probably a whole host of other such features.
None of these originated in Linux.
I can't address all of your points, because I don't know enough about all of the areas you touch, but:
Open Mosix Transparent process migration, intended for clustering.
Has been available in various systems since the 1980s. See, for example, "Sprite".
UML Self hosted virtual machines.
A facility that was possible, I believe, with IBM mainframe systems back as far as the late seventies.
RTLinux Realtime microkernel/macrokernel work. Hell, it _is_ patented.
I don't know what the patents are for, but it sure as hell isn't the first kernel (micro or otherwise) to offer real time facilities.
ReiserFS Filesystem based on dancing trees, with a plugin archtecture.
First time I've seen it in a filesystem, but structuring data in trees has been used in database systems since the late sixties. ReiserFS is only an incremental improvement over most of these systems.
ZisoFS Transparant handling of compressed ISO9660 filesystems.
Transparently compressed files have been a feature of many filesystems over the years. ZisoFS is unusual in that it retrospectively adds the feature to an already existing filesystem, but then ISO9660 was designed to be easily extendible, anyway.
InterMezzo Distributed filesystem, with network interrupt transparacy.
I believe this was one of the design goals of the Andrew filesystem, developed at CMU during the early nineties (?).
As to "Adeos" and "Seperate LLC stack", I'm not sure exactly what these achieve, so I can't really comment on them. I truly doubt, however, that they are utterly unique and new, except perhaps in minor implementation details.
that MS employee's were not permitted to participate in OSS projects?
Why do you think that? Have you seen a statement regarding this? I'd like to see what exactly was said, if there is one.
I would personally expect MS to have issues regarding employees contributing to _some_ OSS projects. One particular aspect: I don't suspect they would permit any employees to sign the FSF copyright assignment forms. These grant a lot of rights to the FSF, which might potentially interfere with rights to work the employee does for MS.
I know my company reviewed it when I was considering doing some work on gcj, and decided that we couldn't accept its terms.
The warning isn't the problem. The fact that the company involved feels they have to include the warning for legal reasons is the problem.
What does allegedly "secretly manipulating an industry-wide cooperative standard-setting body" have to do with the patent system?
I haven't read the book, but they're probably arguing for a system where failing to disclose a patent when proposing somebody else uses an invention it covers renders it invalid. This case (if my memory of the details is correct) is a good example of where such a law would help.
Judges aren't lawyers. The two are completely separate professions that happen to involve very similar training and who work together frequently.
That's like saying that IT project managers are programmers. The two clearly have a lot of domain specific knowledge in common, and programmers often become project managers later on in their careers, but they are entirely separate and distinct jobs.
The laws in question are almost certainly "common law"; that is they were never actually written in governmental acts, but arose through the decisions of judges (and juries) in past similar cases.
And, yes, juries are a large part of the problem. They're too easy to sway with emotional arguments, and often award compensation that is substantially too high. They're also often left to decide on matters that they aren't qualified for, particularly when dealing with laws that involve phrases like "a reasonable person".
OK, there is a very important difference between copyright and patents, you know. Copyright only applies to direct copying. Patents apply even to independent reinvention of the same concept, which is a little dubious, morally speaking, if you ask me.
I don't think it does. Linux is not particularly innovative. Almost all of its features were previously included in another operating system. I can't think of a single thing it does that ought to be eligible for a patent, even in a world where software can be patented.