What exactly does it mean when Richard Stallman says that the Creative Commons' Attribution-ShareAlike license has a "Weak Copyleft"? Why exactly is it that "Freeware" and "Non-Free Software" mean the same thing, while "Free Software" is something else entirely? And what is this business with "Free Beer", and where can I get some? If you've asked yourself these questions, this column is for you.
There's a great deal of jargon surrounding the various cultures and ideas around software freedom. I'm going to take you on a brief tour to try to make some sense out of the language. Each set of terms arises from a particular philosophy and movement of people, so it'll also be necessary to introduce some of those movements.
The term "software" used to mean any kind of recorded material: video or audio recordings, photographic imagery, and computer data tapes. Somehow, the term has come to specifically mean "computer programs."
The term "software" used to mean any kind of recorded material
The term itself carries some interesting ideological baggage since the original meaning of a "ware" is "something you sell" -- and usually a tangible object, while "soft" arises from a usage of the word to mean "intangible" (this is perhaps more obvious through the opposite usage of "hard" as in "hard copy" or "hardware"). In other words, the logical meaning of "software" is "an intangible object you can sell", which entails the whole idea of "intellectual property."
This combination clearly sets up the expectation of treating the data of computer programs as a material objects to be sold, which reflects the mentality of usage at the time the term was coined. Today, we are pretty much stuck with the term "software" as a term for computer programs -- so much so that the old meaning of "software" has had to be subsumed by new terms, such as "content."
In any case, the word "software" has been combined with a number of terms to reflect alternative business models to the copy-sales model of the early proprietary software culture.
The word "software" has been combined with a number of terms to reflect alternative business models
The first of these is "shareware" which was software you were allowed to share freely with other users. But, following the pattern of previous proprietary software, the source code was not revealed, being regarded as a trade secret, and you were meant to pay for the shareware if you used it seriously. Shareware was proprietary software culture on the honor system -- not a fundamental change in the basic manufacturing model of creating and selling software.
"Freeware" was a tiny variation on this. It indicated an ethical shift of focus in which the author had no expectation of you needing to pay for the software. It still generally means software released as a binary with the source code being kept secret. Reasons for releasing "freeware" ranged from promoting other, expensive licensed software, to the output of hobbyists.
The term "freeware" is often incorrectly used as a synonym for "free software" by people who don't appreciate the difference implied by the lack of source code or of the license to adapt and modify the software.
The term "freeware" is often incorrectly used as a synonym for "free software" by people who don't appreciate the difference implied by the lack of source code
Another variation on "shareware" was the value-added idea: to add to the guilt motivation of asking people to be honorable and pay up on shareware, authors would often offer enhanced features to people who paid to register their software.
But of course, taken to an extreme, this could also be understood as disabling functional software to create a "demo" which showed what the software could do. Then, when the software was registered, the user would receive a key, or possibly an entire new version of the software that was fully working. This kind of shareware came to be known as "demoware."
Yet another variation which occurred much later was "adware." This is software of the "freeware" or "demoware" variety which actually displays advertisements on the user's computer while it is operating. This is sometimes advertising the more advanced versions of the same software, and sometimes simply sending targeted ads to users.
When Richard Stallman started the GNU project, he described what he was doing as creating "free software." He actually meant both free as in zero-cost and free as in free for the user to use in any desired way. It was somewhat later that he came to see this distinction as so fundamental. In expressing the distinction, he rather famously said that he meant "free as in 'free speech', not free as in 'free beer.'"
Sadly, many people didn't understand him, as evidenced by the scads of people who have interpreted "free beer" to mean "any kind of software other than the kind the GNU project produces." Somehow, they've managed to lump all kinds of free software into this category, which is odd, since it's such an obvious statement.
What Stallman dismissed as "free beer" software was the "freeware" of the previous section
What Stallman dismissed as "free beer" software was the "freeware" of the previous section: software that was without cost and could be distributed verbatim, but which did not come with either the legal permission nor the practical possibility of revising and improving it. It's actually kind of a subtle distinction for anyone who isn't a programmer, but for Stallman and his colleagues, who are programmers, it is a critical difference.
In another attempt to relate this distinction, Stallman (and others) have noted that in many languages, freedom from monetary cost and freedom from other restrictions have different words. In French, for example, the former is "gratis" and the latter is "libre." This frequently goes with some very oddly chauvinistic sentiments about the superiority of such languages.
From it, however, we've seen the borrowing of the terms from French to give us "gratis software" (synonymous with "freeware") and "libre software" (synonymous with "free software").
We've seen the borrowing of the terms from French to give us "gratis software" and "libre software"
From Stallman's writings we also get yet another term which is particularly ironic: "non-free software." This actually includes software which may or may not be sold with a license fee, but which either does not grant modification privileges or does not provide access to the source code from which to make modifications.
Of course, few people are confused about "shrink-wrapped" software sold in a store -- it's very obvious that that isn't free. So most of the time when "non-free" is used it's used for software distributed alongside free software, which can only happen if the software allows verbatim re-distribution without paying license fees. In other words, it really only comes up if it is freeware.
So in practice, "non-free software" usually means "freeware", which is pretty confusing!
"Software freedom" is the thing "free software" has which makes it "free." It came to us mainly through the writings of Eben Moglen, a lawyer who was instrumental in revising the GNU General Public License (GPL). He also organized the "Software Freedom Law Center," which does a lot of good for the free software and free culture community, especially in working with the Free Software Foundation.
"Free software" includes both "copyleft" and "non-copyleft" software
Now "free software" includes both "copyleft" and "non-copyleft" software, which raises the question, "What is a copyleft?" In short, it's this: with a copyleft, if you publish an altered versions of the software, you have to make the source code available to anyone who gets the software, and you have to make it all available under the same terms you received the software. It's a nice quid pro quo arrangement -- we give you software that is free, and in exchange you must agree to keep it that way. The definitive copyleft license is the "GNU General Public License", so copyleft software is sometimes called "gpl software" (occasionally even if the license is not actually the GPL, but that should really be regarded as incorrect).
Why do guys in suits hate "free"? There's plenty of conspiracy theories around this, but in the end it comes down to something very simple: people who are in the business of making money by producing products are very quick to ask how the production side of the business works, and being told that it runs on magic pixie dust doesn't sit very well with them.
People who are in the business of making money by producing products are very quick to ask how the production side of the business works, and being told that it runs on magic pixie dust doesn't sit very well with them
While the ambiguity in the term "free software" might well be a positive marketing device for the average end-user ("Hey, I can get this for free!"), it is extremely negative for people who sympathize with producers and expect to pay for what they use. They're much more likely to immediately ask questions about sustainability.
Also, being busy people, they don't always sit to listen to the circumlocutious explanations we like to give for how it all works, they just want it in an nutshell. Their theoretical understanding and their experience with "freeware" and "shareware" suggests that "free" software is probably a bad thing. So calling it that immediately sets up an obstacle.
So one group of enterprising people in the free software community who did get how free software sustainability worked decided they'd had enough with playing these pedantic games and that substance was more important than words. They decided to start emphasizing the key aspect of free software that distinguished it from the "freeware" which was causing all the brand confusion. This was the Open Source Initiative.
That distinction is that with free software, you have the source code. So they called it "open source software." This has been an extraordinarily successful campaign, and accounts for a great deal of the penetration of free software into the business world, and through them, into society at large.
I frequently encounter clueless people who think that "open source" licenses are different than free software licenses. Particularly common is the belief that "open source" licenses lack a copyleft. These are myths. Both "free software" and "open source software" mean software under a free license, with or without a copyleft.
Both "free software" and "open source software" mean software under a free license, with or without a copyleft
A more sophisticated criticism suggests that "open source" is more subject to brand confusion than is "free software," although the evidence is mostly against this.
However, there is one category of twisted rip-off of the "open source" label which is Microsoft's "shared source" concept. This is meant to sound like "open source," but isn't. "Shared source" does make source code available, but the license is highly restrictive as to what you can do with it.
Although some people love to nitpick over the motivational differences between free-software and open-source software, most people in the community recognize that they are the same thing, at least in terms of results. So, a number of phrases and abbreviations have been coined to try to express neutrality on this occasionally hot-button issue. Some come with popular abbreviations you may have seen as well:
They all mean the same thing: software which comes with a license which allows you to do what you like with it, and which comes with source code so you can.
In 2002, a whole new set of jargon was introduced by a new organization started by Lawrence Lessig, called Creative Commons.
"The Commons" is a concept from a specific metaphorical essay by Garrett Hardin, titled "The Tragedy of the Commons." The essays is about the sharing of physical public resources -- grazing land, in particular, and how it fails, requiring either property rights enforcement or regulatory enforcement.
"The Commons" is a concept from a specific metaphorical essay by Garrett Hardin, titled "The Tragedy of the Commons"
Later writers have pointed out that in the intellectual regime of "Creative" works, this idea no longer makes sense. The "Tragedy" of the commons is replaced with a "Comedy of the Commons". The "Creative Commons" takes its name from this metaphor, and proponents like to describe intellectual production theory in these terms (sometimes to a fault).
The Commons or The Intellectual Commons, or the Creative Commons is therefore very similar to the idea of the "Public Domain".
"Commons-based peer production" describes the way in which people can work together better as equals (peers) if they use "commons-based" licenses (in other words, "free licenses") in their work. This takes the focus away from "who must I seek permission from to work on this?" and puts it on "what am I already allowed to do with this work?"
Extending this idea to the large scale, especially when there is money involved, is the idea behind "commons-based enterprise", which suggests using this kind of production to displace tasks we have previously relied on tightly-bound "corporate enterprise" to accomplish.
The Creative Commons licenses are modular, at least in concept, and the modules describe what each license is meant to ensure:
The most basic module is "Attribution." This means telling who made the work. It's intended as a defense against plagiarism although it has some more subtle uses. It lets you insist that credit be given to a particular entity (usually yourself) for creating the work.
"ShareAlike" is the essence of the "copyleft" idea, and they are basically synonyms. The ideas is that you must "share alike" -- that is, "share the work in the same was as it was shared with you." The CC licenses don't have anything analogous to the "source code" requirement of the GPL, mainly because defining what constitutes source code doesn't make much sense outside of software.
The "NonCommercial" module is highly controversial. The idea was that people would share works for non-commercial use, but still be able to demand payment from commercial users. This sounds sensible, but is actually full of holes, because what constitutes "commercial use" is very hard to quantify. It also makes the work incompatible with real "free culture" works.
The "NoDerivatives" module is only slightly less controversial. The idea was to protect political speech from being altered to change its meaning. The name can be confusing, because of course, the ND modules does not restrict any of the pre-existing fair use rights to quote from a work.
From these modules come the six core Creative Commons licenses:
...plus some extra weird licenses:
These allow derivatives based on "samples" of the work to be reused according to one of the other license terms, while not allowing verbatim use. They are not licenses for collections of samples or individual works intended to be used as samples (because they would actually prohibit the obvious use of such works).
...and finally, there is:
CC0 is simply a way of expressing that a work is in the "public domain." It either verifies an existing state of affairs or it creates such a broad license (not even requiring attribution) that the work is effectively in the public domain.
Of course the "public domain" is the collective body of works that aren't affected by copyright law.
The term "Free Culture" was made famous by Lawrence Lessig's book of the same name, in which he promoted the ideas behind the Creative Commons. He considers it to include all of the kinds of sharing enabled by the CC licenses.
The word "free commons" has therefore usually been restricted to those works falling into the By or By-SA terms, or under a variety of licenses from other parties which provide similar terms
However, there is opposition to this idea by those who feel that the NonCommercial and NoDerivatives modules are antithetical to the creation of a true "free commons", and the word "free commons" has therefore usually been restricted to those works falling into the By or By-SA terms, or under a variety of licenses from other parties which provide similar terms: the most prominent of these being the License Art Libre (or "Free Art License" in English).
They sometimes refer to the other CC licenses as "semi free" and the "commons" they create as either a "weak commons" or as "no commons at all." Either way, they are referring to the way that the NC and ND modules block the production of derivative works and the sharing of intellectual production with those who are creating "truly free" works.
These are sometimes called "free cultural works" as promoted by an organization (and wiki) calling itself Freedom Defined.
There are also a few people (not so many) who refer to "open source culture," and for example, the Blender movies which are released under the Creative Commons Attribution license are described as "Open Movies."
Occasionally, you'll also see "copyleft culture," though that should strictly be applied only to those promoting the use of copyleft licenses (like the Creative Commons Attribution-ShareAlike) to these works.
A closely related idea to sharing expressions through various copyright licenses is sharing the information contained in the expressions, that is to say, the data. This is an idea preferred by scientists, who are not so much interested in keeping their texts free as in keeping their scientific data free and accessible -- leading to the term "open access."
Which means the policy of making scientific data available to the public for free. Since copyright protection for scientific papers is largely regarded as unimportant by scientists (who don't make money by selling these rights), the scientific community has been very favorable to simply placing their content into the public domain, using the CC0 assertion. So "open access" is closely related to CC0 in practice.
Some people thought that this free software thing was so cool that we ought to do this with everything, even material production. Of course, in reality that's a little complicated because matter isn't so easy to replicate, but proponents quickly came to realize that it was really the design for the material objects that could be open.
Some people thought that this free software thing was so cool that we ought to do this with everything, even material production
By now, however, the language of freedom was so completely confusing, that proponents had a hard time coming up with a name for themselves and their products. This has resulted in a plethora of confusing, related, and often synonymous or contradictory names:
Free hardware was quickly rejected, because while "free" is slightly confusing when applied to software, it makes no sense at all when applied to hardware.
Open hardware is becoming the most common way to refer to hardware with a free-licensed design.
However, it is also still widely used to mean hardware whose design is secret or proprietary, but whose operation is well-documented so that it is "open" for the purposes of creating free software drivers for it. Due to the fact that most variations have failed to resolve the ambiguity, the term "open hardware" has come to be associated with the former definition.
In the hopes of resolving this conflict, some groups have proposed the following variations:
Open design hardware is suggested as the name for hardware whose design documents are free-licensed so that anyone can build it. Except by those who think that it means hardware who design is open for people to write drivers. Oops.
free design hardware Was therefore suggested as a way to emphasize that the hardware design documents were free-licensed. Still, though, many people dislike the "free" ambiguity with anything relating to hardware
Open source hardware has therefore also been suggested for hardware whose design is free-licensed, but free software advocates who hate open source advocates don't like it. Nevertheless, this has been the most unambiguous way to refer to this definition, and is the term preferred by Wikipedia.
Open specification hardware was suggested as the term for hardware whose operation is specified for the purposed of writing drivers, though some people think it means a "full specification" of the hardware, which would amount to a free-licensed design. So that's still confusing.
Free-software-friendly hardware is the most unambiguous way to say that the hardware provides the necessary operational information necessary to write software drivers for it.
As you can see, not every battle has been rendered history here. Some of these names are still in flux. Others are associated with definite movements in the community and may be hot button issues for colleagues. But I hope that after reading this, you'll at least have an idea of what people are talking about!
This work may be distributed under the terms of the Creative Commons Attribution-ShareAlike License, version 3.0, with attribution to "Terry Hancock, first published in Free Software Magazine". Illustrations and modifications to illustrations are under the same license and attribution, except as noted in their captions (all images in this article are CC By-SA 3.0 compatible).