Linus Torvalds, the founder and lead developer of the Linux open-source operating system, has some strong views about the , which he shared with eWEEK Senior Editor Peter Galli in an e-mail exchange last week. Torvalds also last week announced he was taking a leave of absence from Transmeta Corp. and becoming the first full-time fellow at the Open Source Development Lab, where he will continue to drive the next version of the Linux kernel, 2.6, due later this summer.
Do you expect anything to change now that you are working for the OSDL in terms of your focus around Linux?
I dont foresee any particular changes. That said, I remember when I first joined Transmeta, and some issues Transmeta ended up having with SMP ended up being how I started getting into Linux SMP [symmetric multiprocessing] development—not because Transmeta asked me to per se, but because the situation was just different enough from my situation in Helsinki that my priorities shifted. In other words, were all creatures of our environment, and in that sense any change will obviously reflect some way in what I do.
There has been some talk that the members of the OSDL, like IBM, HP or others, may try and sway your focus and get you to include technologies they want to see in the kernel. Are you concerned about this?
No. But part of the reason Im not concerned about it is that we were pretty proactive about it. Exactly to not raise these kinds of concerns, my contract says that I have final word on the kernel, and the copyright remains with me personally. Ive always felt it was important to let people know that there arent any direct commercial influences on the maintainership of the kernel, and that the maintainership is done on purely technical grounds.
Your current focus is obviously on the 2.5 kernel and bug fixing so it can become 2.6. Are you still on track for a release this summer?
Im never on track, and maybe Ill have to move to Australia to make good on it, but on the whole Im actually pretty happy with where we are. Delayed (as usual), but there are no big show-stoppers. Were getting to the point where Ill start doing the so-called “pre-kernels” to encourage more people to start testing stuff heavily.
Where are you at with the kernel and what are you currently concentrating on?
Right now theres nothing particularly worrying going on, and its mostly a lot of “locking down the hatches.” Sometimes too much of it, since some of the bug-fixing has degenerated into “cleanups” again, so Ill have to try to convince people to let it go and just fix bugs.
The OSDLs focus has been to drive Linux further into the enterprise, yet yours has always been more broad than that. Is the enterprise focus something you are going to push more and focus greater effort on?
Ive always focused on the desktop (a fairly high-end desktop admittedly). I think thats still where the most interesting stuff is, and the reason is largely still the same: its the area that sees the most varied usage patterns. Obviously, scalability is always sexy, so I enjoy that part too, but on the whole I think there are enough people looking at the high end, that for the good of Linux we should still concentrate on the “lower” end of desktops. The high end, to a large degree, is the easy part. The problems are well known, the solutions are out there too, and the workloads tend to be well-behaved. So its not something I worry about.
Are you seeing/hearing of any existing or potential Linux users pushing back or delaying rollout plans as a result of the SCO/IBM court case?
No, Ive not seen any customer/user respond in any way to the SCO/IBM spat, but thats partly because I tend to work only on the technical side, and the lawsuit hasnt affected that part.
What is your position on SCOs IP claims and its allegations that code from Unix System V like its non-uniform memory access [NUMA] technologies have been incorporated into Linux?
As far as I can tell, SCO doesnt have any IP claims. Their lawsuit isnt about IP claims; its about some contract dispute with IBM. The only IP issues they have brought up in a verifiable way has been the RCU [Read Copy Update, a way to access data structures that may be changing on multiple CPUs with less locking than normal] work that IBM did, and that SCO doesnt have any IP rights to that I can see: the patents are all IBM, and the code was written by (and thus copyrighted by) IBM too. Well, it was Sequent at the time, but theyre all IBM now.
SCO alleges that you need to focus more on getting clarification as to where the code that goes in the Linux kernel comes from. Do you have any plans to change the current Linux development model?
No. I allege that SCO is full of it, and that the Linux process is already the most transparent process in the whole industry. Lets face it, nobody else even comes close to being as good at showing the evolution and source of every single line of code out there. The only party that has had serious problems clarifying what they are talking about is SCO, and now when details start emerging like with RCU, its clearly about IP that they had nothing to do with, and dont even own. Im sure that they are confident that they own the collective work of Unix, but thats a separate thing entirely legally from being the actual copyright owner of any specific section of code.
As the founder and lead developer of Linux and the copyright and IP holder, do you intend to get involved in the IBM/SCO legal matter?
The less I have to do with suits, the happier I am. That said, if Im called as an expert witness, Ill go if only because Id be curious about the process.
Some people in the open source community say SCO may itself have taken Linux code and included it in an unauthorized way in Unix System V, while others say its use of and contribution to Linux means it essentially open sourced those products. What is your opinion on these claims?
I personally think its a lot easier (and thus more likely) to integrate open source code into a proprietary platform than the other way around. That said, I dont like the SCO FUD, and I dont have any huge reason to spread FUD around myself. In other words, I dont personally know of any such code.
SCO has said you need to sign the non-disclosure agreement (NDA) in order to look at the alleged copied code in question. I assume you are not going to sign it.
Thats a fairly safe assumption. I dont generally sign NDAs even with friendly companies, because it can hinder my work. Id be crazy to sign one with SCO. Especially as signing an NDA would make the act of then seeing their claims totally useless, since I couldnt then go out and search the public for the sources. However, now that SCO is starting to talk a bit more about what they seem to object to, I have less and less interest in seeing the code. As mentioned, the stuff they seem to be complaining about they have absolutely no IP rights to that I can tell.
Has this legal matter affected in any way how you or any of the kernel developers are working on the current kernel?
Not that I know of, although I do suspect a fair amount of time has been wasted just worrying about it.