Tell me about the GNU GPL (General Public License). Are you still working on it?
Yes, we're slowly working on thinking about what to put into GPL version three, but not with a great sense of urgency.
What's not clear about the GPL now?
It's not so much as in things that are not clear as a matter of trying to handle some issues differently. One issue that we might want to handle differently is the question of using a modified version publicly.
Usually, when you install a program on your computer and you run it, you're using it privately. There are certain kinds of use that are public, for instance, putting it on a server and letting the public get access to that server and run that program. So our idea is that, it's legitimate for a developer to say, if you make a modified version and you put it into public use, then you have to make the source codes available to those who are using it.
There was some confusion around that, wasn't there?
There is no confusion about that. GPL version two doesn't require this, it's very clear. But we're thinking that it might be a useful requirement to have for some programs so the idea is to design a way that a developer can optionally require this when using GPL version three.
The way we're thinking of is, if you design the program so that it has a command that the user can use to download the source, then whoever makes a modified version and installs it publicly has to keep that working and provide the source code of the version in use.
And you won't make an umbrella rule that requires everyone to do that?
No. This is because that would mean a big change for existing programs and we don't want to do that.
We might also put in some kind of additional clause to take away rights from those who attack the community with patents. But we're not sure what kind of clause would be legitimate and effective. So we haven't done it yet.
We might put in something that says you can't put the software into a kind of DRM signature system that won't let the users run modified version. So, if you're distributing it in a machine that requires signatures to run the program, then you've got to give the signature keys to the users so they can store their modified versions so it would run. This would essentially make such a scheme futile, but that's the whole point: people should be free. But again, this is a possibility and we're not sure we're going to do this.
I should point out that the GNU GPL is just one of many free-software licences. Any licence that says you have these four freedoms is a free-software licence. Now, there are two large categories of free-software licences: the copyleft and the non-copyleft licences. The copyleft licences say that all modified versions must also be free and they must be under the same licence. The GNU GPL is an example of a copyleft licence.
There are also free-software licences that are not copyleft, and those permit non-free versions of the program. Now, I think it's a mistake to permit that in most cases. Why should we cooperate in any way with resistance of non-free software? Why should we let our code be used in any non-free software? The way I see it, if someone else is not going to share with you and me, then why should he be able to use our code if we're not going to be able to use his codes? The GPL says, you're welcome to join us in sharing… but you can't just take and privatise, and then use our code to subjugate people.
Talkback
Richard Stallman is not responsible for Linux which was created by Linus Torvalds and contributed to by many people across the world. According to Stallman, Linux should be called GNU/Linux because it consists of the Linux kernel plus GNU software. This is simply not true - if that was all Linux consisted of it would be largely useless. A Linux distribution consists of a lot of open source software which is not exclusive to Linux but which is key to its success and has nothing to do with the GNU project. Software which is not completed in a timely fashion, such as the GNU kernel (called the Hurd) may be of interest to programmers but is of no real value. Linus Torvalds' genious was to invent a method of working which produced such remarkable results.
Software was shared long before Richard Stallman created the Free Software Foundation. Much of it came out of academic and research institutions and was distributed by user groups. The narrow focus of the GNU project excludes the Internet - without which Linux would just be another student project. The X Window System, which is a key part of desktop Linux, was produced by a consortium of computer manufacturers and academic researhers and ows its success to being an open source project (well before the term was invented).
Today the work of the Mozilla Foundation promises to enable millions of users to reclaim the iternet from the virus and spyware writers. The success of the Firefox project shows just what open source development can do.
I gratefully run Mr. Stallman's software every day, but on the topic of "open source software" he is confused. When a bunch of us met in Palo Alto in 1998 and decided to call it "open source" rather than "free", it was to be inclusive of the BSD license. Under a BSD style license, software is usable by anyone for any purpose, even if they don't share their subsequent improvements with the community. Mr. Stallman's "GPL" is far more restrictive, and the Open Source community is far larger than Mr. Stallman's important corner of it.
At ISC we publish BIND and other software under a BSD style license, and I hope that Mr. Stallman agrees that our work has been helpful in creating and expanding the industry, beneficial to the public, and will be relevant going forward -- in spite of our unwillingness to follow the FSF/GPL ideology.
I started using freely distributable software in 1969, through the facilities of the Digital Equipment Corporation's User's Society (DECUS). It was kept in a library, listed in a catalog, and delivered for a small copying fee on paper tape. Most software in those days was owned by either the person or company who wrote it or who paid to have it written. Most software was very expensive, and as a college student I was grateful for having access to software and its source code that was contributed by DECUS members.
In the 1980s the "production software market" started up, and people began to produce software for the masses. By this time the Unix operating system had a great following in engineering, scientific and academic circles, with the University of California hosting a huge effort to create what became known as the Berkeley Software Distribution (BSD). But this distribution's license allowed vendors to create different "spin-offs" of code, and did not require them to ship the source code, so their customers could not (for the most part) change the code to meet their needs or even fix bugs they found. They were dependent on the manufacturer's programmers, who were often overloaded with work.
Lots of people continued to write software that they distributed without charge, and with the source code available. But Richard Stallman decided to write an entire operating system called "GNU" (for "GNU is Not Unix"). And he wrote a software license (GPL) that guaranteed the freedoms that he enjoyed:
o Freedom to read the source code
o Freedom to change the source code
o Freedom to distribute the source code
would be available to anyone who used or modified his code.
Richard, as the evangelist of Free Software, build an army of followers and believers. He then founded the Free Software Foundation.
The software tools, compilers, libraries, and other things that the Free Software Foundation wrote was not what most people would consider a complete system, but it was headed that way. And along the way millions of people benefited from using the exisiting GNU code on top of other operating systems.
In 1991 Linus Torvalds started his kernel project. It was the right project at the right time, and mated with the GNU code and other free software code such as sendmail, BIND, the X Window System sparked people's imaginations as to what could be done.
At first I did not understand Richard's insistance on inserting the word "Free" in the equation. The Open Source people felt that the word "Free" was confusing business people, since they thought that the software was gratis, and should always be gratis. Somehow the software would appear by "magic". However, in the twenty years I have known Richard, I have never heard him say that you should not get paid for writing software. He only feels that after the software is written, the source code should be available to those who want it.
After ten years as executive director of Linux International, and thirty-five years in the computer industry, I agree with Richard that not having the source code freely available to the customer is an impediment to doing business. Not everyone will agree with the GPL license, just as not everyone agrees with BSD. That is their right. That is their "freedom".
We do not live in the same environment of the 1980s that spawned the proprietary system market. Software then was expensive and the people who wrote software for those expensive computers were few. Knowledge about how to write whole systems was scarce and communications were much slower than today.
Now there are too many customers with too many unique needs for any company to meet those needs. the companies' "telephone support lines" today just get busy signals, dropped lines and intollerable waits. Free Software's insistance on the availability of source code offers competition and choice that proprietary software does not allow. It offers freedom. When p