Slashdot Mirror
FSF updates Free Software definition
Mark Wielaard writes "The FSF has updated their definition of Free Software.
It now says something about the freedom to actually run the program
for any purpose, the freedom to redistribute copies to anyone, anywhere
(export restrictions) and how those freedoms should be irrevocable
(license termination).
"
No, it doesn't dovetail better with the Open Source definition of the OSI. In fact, it points out a failing of the Open Source definition.
Both the Free Software Foundation and the Open Source Initiatives are talking about software licenses. Software licenses are essentially based on copyright law, since in most cases, you have to copy software (i.e. install it) in order to use it, so your rights to copy it must be clearly defined. Copyright law only governs distribution of intellectual property, not use, so when both the FSF and OSI were defining things, nobody thought to point out that the freedom to use the software was important, it just went without saying.
Some recent licenses (such as IBM's Jikes license, and the APSL) have come out with terms saying that the company (IBM or Apple) can under certain circumstances terminate your right to use the software. The OSI has taken the position that such terms are not against the rules. The FSF has taken the position that these terms take away a fundamental freedom, just one they hadn't bothered to write down. Personally, I think such provisions are probably legally unenforcable, but I am not a lawyer, and I don't think either organization wants to sue IBM or Apple. In absence of that, I personally agree with the FSF, such terms make software non-Free.
----
Open mind, insert foot.
I think RMS is (quite rightly) trying to elevate the status of software to that of academic research, in which the different levels of freedom that he outlines are taken for granted, to the benefit of the entire research community.
1) The freedom to run the program, for any purpose (level 0).
Academic research does not come with any prohibitions on its use. If its published, it's usable by anyone for anything.
2) The freedom to study how the program works, and adapt it to your needs (level 1).
This is the entire point of publishing research, so that others may see how it works and possibly modify it.
3) The freedom to redistribute copies so you can help your neighbor (level 2).
Publishing your research means that anyone has access to it at a decent academic library, so copying is not even an issue.
4) The freedom to improve the program, and release your improvements to the public, so that the whole community benefits. (level 3).
Again, this is inherent in academic research. Publishing your results allows others to improve on your ideas or borrow your ideas for another purpose.
Removing any one of the above freedoms from academic research would be a ghastly proposition for all those involved. It would also harm basic research.
I think that developing software should be considered a form of academic research. By realizing that what RMS is trying to do is not so far fetched (i.e., "free research" is already taken for granted), free software can be better appreciated by those who don't understand the model.
The GNU project has never gone into extreme detail on all the criteria software needs to meet in order to be free. Much like the famous Then the Open Source people took the Debian Free Software Guidelines and tried to turn them into a more legalistic framework. Companies now view the Open Source Definition as the "law" they have to comply with, and as with any laws, seek loopholes that will allow them to comply with the letter but not the spirit of the document in a way that gives them an advantage.
Unfortunately, finding loopholes in the OSD doesn't really make the resulting code free. (For example, those with termination clauses). Thus, the explanation of what free software is needs to be tightened. Richard Stallman did not just unilaterally change the definition of free software. Instead, he made explicit something that was true all along, but was never actually stated.
I hate to see these definitions become more legalistic. This is probably somewhat necessary though. Corporations who have no background in the free software community need more detailed information than those of us who have been immersed in it for some time. We just need to make clear that these documents aren't laws to be worked around and that finding a loophole in one of them doesn't mean your software will be considered free.
You could insist that they call you God if they want to interview you or talk to you.
What I'm trying to say is that the "What is Free Software?" essay on the GNU web site is not meant to be a legal style definition of free software. You could (perhaps) devise a license that met the criteria there, but it would still be considered non-free. For example, the termination clause licenses fell into this category until the FSF updated the web site to address that particular problem. The Open Source Definition is presented as "meet these requirements, get our trademark".
So you think that it is "imposing" on someone to insist that they not use a term you find offensive when addressing you. It a black man imposing if he objects to someone saying "Hey, nigger, I've got a question for you?" Stallman probably doesn't feel as strongly about GNU/Linux as that, but the same principle applies. People are asking Stallman questions they want the answer to. If they expect him to answer them, then they have to comply with his wishes in addressing him or he will exercise his free speech right not to answer. If the people asking questions do not wish to use GNU/Linux, that is certainly their right, just as it is Stallman's right to suggest that they use a different term and to refuse to answer if they do not.
You are certainly free to choose another definition and select software based on that or any other criteria. Stallman speaks only for himself and the GNU project that he is the head of. If other people choose to accept his opinions, that is there choice. But he is not forcing his definition on anyone else. If many people choose to agree with him, then that is a tribute to his persuasiveness.
Given the way society has become increasingly dependant on computers and the internet to function, I argue that free software (as defined by FSF) is becoming essential to the stability of our economy and perhaps our society.
If proprietary and IMHO monopoly software is allowed to become essential to our infrastructure, the licensing terms on that software will go from being a simple cost to being more like a tax, or at least a cover charge. We have enough of those from natural monopolies as it is.
By specifying that acts of civil disobediance cannot disallow membership in society and commerce (by terminating the license to use these essential services), the FSF takes an important stance for civil liberty in general.
I guess the factions are finally getting down to cases. The level structure (may be new, but this is the first time I've seen it) is interesting. The highest is improve software and the lowest (ie, most important requirement) is usage.
This ordering, more than any other showcases the difference between FSF/RMS and ESR/OSS. The former sees an egalitarian ability to use software as the key to the Free Software Movement. Source code is the key to using however you want (without source you can't use EMACS on your old Timex/Sinclair) because it gives you platform and site freedom. ESR and his fellow travellers on the other hand, see software as a mertocracy of craftsman. "Let me have the source so I can improve it and we all have better products," is their mantra.
Like most ideology issues, I think the majority of the community is somewhere in the middle. Yeah, we want source to improve and fix, but we also want it to port and steal (oops., borrow). For the the times RMS has been called a hippie/socialist/commie/loser/whatever we still don't like the Apple license.
If their is any single thing from the FSF I'd like to see OSS adopted is the permenance of the license. This to me, even more than source, is key. If I know what the rules will always be I can make a much more imformed and productive choices. I don't so much object to more restrictive source licenses (one time fees and such) as I do to licenses that can be pulled out from under me at any time, after months of work on my part.
One final question for the FSF guys. Given the generally left leaning politics of your staff, have you considered the free to use cause and military usage? What if copylefted imaging aided in missle design or crypto added in war radio traffic? Is the principle of free usage important enough to override other principles. I ask because I've seen a lot of source availible (and source not availible) with free licenses except for the US Military/any military/US Government/any government/the Pentagon/etc. Would such a line violate free software principles in your POV (this is not to be insiteful, but a question of opinion...can we restrict people we don't like...the army, fundies, Green Peace, Discordian, whoever, from usage and still have freeware?).
Herb Nowell
Herb
Again, feel free to sentence me to death if my questions annoy you. I'll come back in 5 minutes anyway. -Sythi
Saying that Red Hat derives a benefit from someone else's voluntary labor is to state that they are benefitting from a positive externality. There are many instances of these. For example, if I paint my house, the value of all my neighbors' houses increases in value. Does that mean they are exploiting my unpaid labor? In general, America does not try to eliminate positive externalities. I do not get paid for everything I do that benefits someone else.
I'd like to draw a distinction between what I think Stallman is trying to do and what Raymond is trying to do. Stallman is trying to explain what free software is. Raymond is trying to define what open source software is in a more rigid sense for his branding campaign. I do not see them as two conflicting defitions.
Restricting software to non-military use would, technically, make the software non-free. The underlying prinicple of free software is that the user is more important than the author. Proprietary software claims the author is more important than the user, and thus gives to the author the right to decide who can use the software and what they can do with it. Putting restrictions on the use of software based on the author's own private morality might seem reasonable (if you agree with the author's moral sense, that is), but is a form of controlling the users and dividing them nevertheless. It is basically saying "I will give you this software, but only if you promise not to share it with some other people I don't like". This is the classic proprietary divide and conquer the users strategy.
Don't forget, one person's no military use is another person's no use on the Sabbath. Or no use by homosexuals. Or no use by the KKK. Or no use by breweries. Once unleashed, where does such a thing stop?