Torvalds Says Linux IP Is Sound 336
An anonymous reader submits: "In an interview with CRN, Linus Torvalds says he's confident there won't be any IP problems discovered in Linux. In fact, Torvalds, says he was extra careful with issues like the IBM Read Copy Update code."
Say what? (Score:5, Interesting)
I thought he didn't care? As in "I don't want to know what we're putting in, don't tell me"? And now he was "extra careful"? Or is this some other type of IP he's referring to?
RCU code (Score:5, Interesting)
2. RCU is a patented technology
3. Patents are publicly viewable
4. Therefore, RCU cannot be a trade secret
I don't see any way SCO can have a claim unless the RCU code that IBM donated contained SysV code or code derived from SysV. I seriously doubt IBM would be stupid enough to do that.
Anyway, since trade secrets are no longer protected once they are publicly revealed, no one should have anything to worry about except possibly IBM.
Linus being naive? (Score:4, Interesting)
Open Source (Score:2, Interesting)
I know things like GPL try to address IP with open source but have their been any big court cases concerning IP on open source Software to test thing out? (forgive me if I an not aware of some big case, that 99% of everyone always knows about)
Re:The circle is complete (Score:1, Interesting)
Re:The circle is complete (Score:1, Interesting)
Re:Open Source (Score:2, Interesting)
Someone - anyone - can say that they hacked trade secret/patent protected/copyright code into, say, the Linux kernel.
Unless these claims are matched in the version control and change logs, they'll have a hard time proving it.
T&K.
Re:Outcomes of the SCO trial (Score:5, Interesting)
What?
I thought that linux would simply have the offending code removed - the rest of the code is GPL's and I could release "isoroxix" tomorrow with the same non-sco code. Linux wont go away, and even if it did there's always HURD
Re:Say what? (Score:5, Interesting)
Re:I haven't done more than peek at the kernel sou (Score:3, Interesting)
ate_utils is/was part of the NUMA code for IA64. It has been removed from the latest development and prepatch kernels (it's obsolete, from what I've read.)
There are some less interesting similarities between the signal handling code, but that's hardly remarkable. How many ways can you write a switch statement?
The key thing to remember though is that SCO doesn't actually claim they wrote the code. They claim to own rights in code that IBM wrote. I would bet that it's safe to assume that a detailed inspection of AIX and Linux 2.4.21 will reveal similarities. The question is, does IBM have the right to relicense code they created?
IP is ImPrecise (Score:2, Interesting)
Please, please, please can we avoid using the acronym IP? It is not at all a valid concept - these are all completely disparate areas of law and should never be referred to in the same breath without clarification.
It is often used as a term that encompasses patents, trademarks and copyright - but if you ever hire an "IP lawyer"(sic) you will nearly invariably find that they only specialise in one or two of the above areas.
Q.
Re:Say what? (Score:4, Interesting)
He was referring to two different things. The "I don't want to know" bit is about patents, which is a totally different animal from IP and copyright. Knowing about a patent ahead of time is tantamount to being tainted by the patent, and if the patented technology shows up in Linux, you're liable for triple damages because you knew about the patent and it went in anyway (under the US system anyway). You cannot check for patents because of this problem.
IP and copyright are treated differently under the law. There is a little more leeway in these cases. Simple copyright infringement can be handled by stopping the infringement, especially when it can be shown via an audit trail like the kinds that exist for the Linux kernel that the infringement was not malicious or with forethought on the kernel maintainers (i.e. they thought they were getting code whose use was free and clear).