Defining Software Freedom: The Singularity of Free

Reshaping the "free" vs. "open source" arguments.

Last week, I wrote about the somewhat-vague definition of the open core business model, and how it compared to the dual-license business model. Open core, like dual licenses, are all part of the whole "ways-to-make-money-faster-with-open-source-software" genre of business, but they are not the same thing.

It should be noted up front that I may have been a bit presumptuous in assigning the origins of open core to Aaron Fulkerson, the CEO of MindTouch. In a comment on my blog, Fulkerson himself corrected the issue:

"I can't take credit for 'open core.' When we began capitalizing MindTouch I employed a model that made sense to me because it seemed to strike a nice balance between the needs and wants of the community of MindTouch users and the needs (and wants) of our company. I didn't have a name for it until Lampitt coined the term. I simply adopted it. Furthermore, I don't think this is a new model. Indeed, we've been employing this model since the very dawn of software."

Fair enough.

Regardless of its origins, it's clear that the question about what to "do" about open core is still generating a lot of discussion in the free and open source software communities. On Monday, while many of us were watching fireworks and gorging on grilled meats like the carnivores of old, Pamela Jones at Groklaw weighed in with her opinion of open core:

"Simon Phipps is correct: Open Core is Bad for You, the 'you' here being you and me, end users," Jones wrote.

Jones takes an interesting tack on the open core model, though. Rather than pick the model itself apart, she spends quite a bit of time parsing the comments of Phipps and Mark Radcliffe, both of whom are strongly connected to the Open Source Initiative (OSI). Phipps is a Director of the OSI, and Radcliffe is the General Counsel for the organization with the self-assigned mission to define and enforce the Open Source Definition.

While both Phipps and Radcliffe have each come out with opinions that argue against the open core model, they have emphatically not been speaking on behalf of the OSI. Jones (and, indeed, the rest of us) would like to know what the OSI officially has to say about the whole open core farrago. The problem is, I think she (and the rest of us) may be disappointed.

The OSI may not be able to say anything about open core because (as others, like O'Grady and Aslett have pointed out) there isn't anything in the Open Source Definition that runs counter to the open core model. So, while open core may feel skeezy, by the very foundations of open source, it's not.

This leads Jones to wonder if that might be what the OSI intended all along.

What's twigged her to this line of reasoning was this statement from Radcliffe's blog:

"Simon appears to be suggesting that only a 'copyleft' approach in which all of the software must be available under an open source license to meet the Open Source Definition, which is simply incorrect (the Open Source Definition was a reaction to the limitations imposed by the copyleft approach)."

Which leads Jones to make this statement:

"Is Mark suggesting that OSI intended to facilitate less freedom for the code and end users than the GPL offers, that this was an OSI goal, that 'software freedom for the software user' isn't and never was an OSI goal? Does freedom mean only the right to fork the code? If so, I'd like OSI to say so clearly and on the record."

Such a response is appropriate, but only if you consider all software freedom must only be aligned with the copyleft model. And that's where I have to voice my concerns.

The problem is, both communities believe they are more free than the other, and having one community state the other is less free reveals the fundamental misunderstanding between the free software community and the open source software community.

Here's the thing: at a deep and meaningful level, the open source community believes that free software licenses (particularly the copyleft-oriented ones advocated by the Free Software Foundation) are actually more restrictive than open source licenses. Copyleft, the open source camp will argue, imposes, like Radcliffe said, more limitations on software vendors than open source.

Which seems counterintuitive, but it's all about what you believe is important.

To the FSF and its adherents, a free software license is more free because it makes code shareable, keeps it free, and enforces upstream improvements. But the OSI and its followers look at that enforcement and the perpetuity of freedom as a restriction--which is why you hear open sourcers saying open source licenses are more "permissive."

This may seem completely contradictory, but when two worldviews are different like this, it's very hard to apply the same overriding concept of "freedom" equally to each side. But that doesn't seem to stop either side from trying.

Part of the problem, I believe, is that we are trying to apply the notions of freedom along a single spectrum. On one end is software that is Free and software that is Not Free. My use of capitalization is deliberate: the capped words refer to the overarching concepts of Free that so confused us in Philosophy class and continue to puzzle us in this debate.

Spectrum of Free/Not Free

With a spectrum like this, it becomes very tempting to put FSF closer to one end and the OSI closer to another--or vice versa. Everybody, it seems, has an opinion on which group belongs where, and where other companies, licenses, and software belong.

But I'd like to think of the positioning as more of an inclusive circle. Outside the circle, bounded by a line that demarcates the notion of free (small caps), are things that are proprietary. These are not-free licenses, software, and companies, which definitely reside outside the circle.

The event horizon/singularity of Free

You can argue what (and where) the actual line is--and I'm sure some people will--but let's for now stipulate that there is a free/not-free "event horizon" that separates the two concepts. At the center of the circle is the absolute concept of Free. The closer you are to the center, the closer to the "singularity" of Free you are.

Inside the circle, then, are things that are free. All the free software licenses are in there. So are all the open source licenses. Or, if you apply it organizations, the FSF and the OSI are both inside the circle. But they are at different places in the circle. The arrived in the circle from similar, but different directions, taking different paths. They are each moving toward the center of the circle at different rates. And--this is the important thing--because they are occupying different positions within the circle, the absolute center of the circle will always look different to them.

This is not hard to intuit. Put an object in the middle of the field, and scatter a bunch of observers around it at different angles and distances from the object. Based on their individual perspectives, the appearance of the object will vary significantly. And that's just a material object in three-dimensional space; don't get me started on the concept of Free.

From this, it's not hard to reason the OSI is just as committed to software freedom as any free software advocate, and vice versa. But the OSI's perspective of Free is different than the FSF's. I am not trying to push anyone towards one side of the fence or the other: I believe anyone within this circle has their good points and their flaws. They are all striving for the same thing, the best way they know how.

Neither free or open source software is "more free" than the other, because Free means different things to different people.

As long as these groups are moving towards the Singularity of Free, then any differences they might have on the route taken are just details.

ITWorld DealPost: The best in tech deals and discounts.
Shop Tech Products at Amazon