|
|
|
|
|||||||||||||
|
|||||||||||||||
|
|
|
Date: 2002-03-02
Location: Linux Forum 2002, DenmarkAt the Linux Forum 2002 Dr. Karl-Heinz Strassemeyer from IBM in Böblingen, Germany, gave an interview to Ole Tange, SSLUG. The interview was published with live streaming on Internet and can be found in the archive on sslug.mmmanager.org. The interview is 11 minutes long. The original interview in RealPlayer can be found at sslug.mmmanager.org .
In the Open Source community Dr. Karl-Heinz Strassemeyer is probably best known as the man who was responsible for the S/390 patch for the Linux kernel.
Thanks to MMManager and NET-production for recording, streaming and archiving.
KHS: Dr. Karl-Heinz Strassemeyer
OT: Ole Tange
OT: I would like to know how IBM deals with software patents when IBM is developing Linux?
KHS: Okay, I would say it is not my specific theme to talk about, the legal view of IBM, but I am talking as a developer, and about my experiences and how we deal with this.
Now let me give you an interesting story to start off with. When we had done the so called patch for 390, as I already told this morning, I had a talk with Linus Torvalds about it. I think it was early November 1999, I had an appointment with Linus Torvalds and my key engineer Boas Betzler and Boas wanted to show Linus what this patch was all about, so that he could read into it a little bit and give us advices, on how to do it differently or whatever, right?
I had made this appointment. When our lawyers got hold of the fact that I wanted to show running Linux kernel, including our patch to Linus Torvalds they all of sudden got this idea: "What the hell are you doing? You are doing a distribution". I had no idea what a distribution is and I had not the feeling that I was distributing anything when I showed it to Linus Torvalds but I learned about those legal effects later.
Nevertheless, they said "You can not do this". So I said: "You have a problem. I have an appointment with Linus Torvalds in three days and I will hold this appointment so you'd better make sure what I can do."
So we found out that, I could show this nice running code to Linus Torvalds in an executive office, in an IBM lab, which was on IBM premises. And make sure that nobody took anything out so it wasn't a distribution.
So this was what I learned: Basically after that I understood what was critical about a distribution.
When we put source code into the Open Source community, the code we are putting in there, we have clearly patent clearence for in the sense we know what we put out in the Open Source community so we don't claim own patents on anyway. Our patents clearence process makes sure we are not infringing the patents of anybody else. We are doing it with our proprietary code, so we are fine.
What is wrong about this distribution, is basically the millions of lines of code that we never have seen. We don't know if there are any patent infringments [in this code] with somebody we don't know. We don't want to take the risk of being sued for a patent infringement. That is why we don't do distributions, and that's why we have distributors. Because distributors are not so much exposed [to being sued] as we are. So that's the basic deal as I understand it.
Now working through distributors is fine anyway. It is in certain cases found to be a little difficult. Think if you have proprietary code and there is a problem and you want to patch it, you can get the object code patch out to your customer. In this case you have to get the source code to the distributor and the object code from the distributor to your customer.
At IBM typically the guy who does the source code doesn't get the object code patch out either. There are a central department which is doing it. Basically we have to create relationships with distributors like with the internal IBM department who is doing stuff like that.
This is all interesting stuff which has to be done. But I think it is part of one interesting way of looking at the legal implications of this Open Source community. I give you another example.
Most recently we wanted to make good use of Linux, because we wanted to ship some hardware where we said we would put a little operating system kernel into the hardware. This should enable us to do the initial program load specifically of Linux on top it of the small kernel. Flexible from different targets - CD-ROM, network or whatever. The first idea to take Linux was abendend. We didn't want to do a distribution, because we didn't want a patent infringement being detected. If somebody would have taken us to court we might have had to stop shipping our product.
This is silly, so we simply decided to put a little proprietary kernel into the product. That was less risky.
[But] this is an interesting example. If we really want the industry to exploit the value of Linux, those devices are places where it can be exploited. I think we have to do something about giving the guys who are using Linux embedded systems in their product. So they don't sit on a bomb where someone can sue them and then they cannot ship their product any longer.
So I think the industry has to deal with this problem of patents and Open Source. I don't know how, but we were debating a little bit [with Ole Tange], and one of those examples would be, that we make a patent clearence file for Open Source. So we put everything there into the open, that we want to put out. Then we put the burden on the patent holder to screen it. Say, we want to put it out in two or three months. Then we put the burden on the patent holders in any of those areas and say: "Screen it. If you feel an infringement, [but] speak up early enough or shut up forever."
It is like it is done with marriage. You announce them early ahead. If anybody has anything to say anything against it, speak up in time or shut up forever. I don't know [this], I am not a lawyer, Maybe there are a lot of lawers saying: "This crazy stuff that Strassemeyer is saying cannot be made a law at all."
I personally believe in logical concepts something like this or something along those lines. Which gives legal security to those people who are using Open Source code in their products. Specifically Linux kernel is a very prominent piece of that.
I think the business and the industry, if they wan't to capitalize on the benefit of Open Source [they] have to have get the discussion going and agree on some rules which create legal security to the Open Source users.
This was a little bit of discussion we [with Ole Tange] had along these legal lines. I don't know, maybe I get all kind of speak up from lawyers on this.
But I think it is useful that we don't leave it with them.
Basically it comes to business decisions. And we can't leave business decisions to the lawyers alone.
Did I forget anything?
OT: So can we leave the discussion to the developers alone? Do you think that we can leave the legislation to the developers alone and not have the lawyers look at it?
KHS: Come on, bullshit, excuse me for the word. I would not take any responsibility on legal questions, that is what I need the lawyers for. To be very honest, I only can make intelligent proposals and tell them where rules are too complicated for me to follow. So I take the easy way out, let's be very clear about that.
I think, to the developers we can't leave it at all. But the business decision making people should have some influence on the rights, too.
OT: You are not suggesting that proprietary software, if they don't speak up on this Open Source they thereby have lost their rights to sue another proprietary company?
KHS: No, no. I think that is a very important issue. I don't know which holy cow I am hitting here. I think if we get a rule here for the Open Source community only we should leave it to the proprietary world to deal with their patents in a way where they deal with eachother - simply by ignoring that there is something like an Open Source community. If they want to kill themselves with lawsuits I would let them.
OT: I think that about wraps it up. Thank you for your comments on software patents.
KHS: Okay, thank you.
This transcription was made by Hans Schou <chlor.dk>.
|
|
|
||||||||||||
|
||||||||||||||
| ||||||||||||||
|
||||||||||||||
|
|
|
||||||||||||