When “Barter” Is Not, Or Why GPL Is Actually Epiphenominal

Eliot Frick
in open source

The open source phenomenon rolls on. Its power and value are significant. And yet, proponents would seem to be guilty of overreach. This overreach may not be such a good thing for the open source movement in the long run.

A "barter" cycle, but only if you ignore one of the most important inputs.

The good people at Groklaw have a post up from yesterday entitled, “The GPL Barter Cycle — A Graphic” which endeavors to elucidate how open source software is a self-contained barter cycle. The process diagram is lucid and the thesis it explains is similarly clear. But, there appears to me to be something missing.

For those that don’t know, GPL is a “copyleft” license. It provides a framework by which software source code can be freely distributed, and ensures that any derivative work carries the same expectation of free distribution. Which, I think this is a great thing, isn’t it? Because of GPL, we have innovations that wouldn’t have existed otherwise. Mobile devices, embedded devices, start-up businesses; a great many things have been the beneficiary of the open source movement.

There’s just one little problem. Many proponents of open source seem to be suggesting that all software should be free. What’s more, they spare no sophistry to make their convoluted case for how software should somehow transcend the property rights that made its existence possible in the first place. This strikes me as not such a good thing for open source.

If you click through to the Groklaw post, you’ll see their diagram. I’ve altered it above. You can click the image in this post to see it at full size. You see, I think they’ve left out the most important input. Without this input, open source goes away. The input is the livelihood made by open source contributors at their day jobs. Without proprietary software firms, these open source contributers would have no income, and would be too invested in finding other employment to be donating their time to this ostensible barter cycle.

My point is that GPL is an epiphenominon of the marketplace. Take away property rights wholesale, and you kill open source. In the vacuum left by any hypothetical abolition of proprietary software, there is no mechanism by which developers can be compensated for their work. A working copy of Linux doesn’t pay your mortgage or put food on the table.

A sober analysis would seem to suggest that open source advocates would appreciate the existence of the proprietary software market. As it stands, open source ideologues often gesticulate grandly so as to distract their proselytes from the basic economic facts. <yoda>Cutting off their noses to spite their faces they are.</yoda>

Props to PO1R, who created the diagram at Groklaw, and Larry Ewing in turn for the Tux logo used therein.

  • Brian Schwartz

    I just read the GNU post about all software being free and it's pretty absurd. I cannot trade my little bit of code for food or mortgage payments and without subsidies from universities (which is one of the ways they suggest you could make money in this freeconomy) there would cease to be software. There isn't that much education / grant money floating around to maintain software.

    Your adjustment to the diagram and the term epiphenomenon seem correct, however as a result of this phenomenon over the last 10 years we now have entire eco-systems of open sources projects, usually managed by for-profit companies (think wordpress and automattic and dotnetnuke and dotnetnuke corp) that support a large number of other companies and freelancers.

    The entire LAMP group of software while probably originally developed by people who were getting paid by proprietary software company are now probably close to self-sustaining (people who work on them get paid to work on them by companies who get paid to do LAMP consulting).

    It's an interesting paradigm shift as a new generation of programmers now expect IDE's to be free and open source. (And while Visual Studio isn't expensive, per se, it's still not free, nor will it likely ever be). I continue to be amused at the die-hards (Linus Torvalds of the world) champion free while still somehow making money on it (paid speaking and consulting engagements) and there are continual arguments within these ecosystems about something not being free or “true to the open source community”. The zealots reeks of hypocrisy.

    Somewhere someone will need to get paid or innovation will cease. Programming for fun can be fun, but fun won't feed my family.

  • Eliot Frick

    Agreed re: ecosystems managed by for-profit companies. That's really the ultimate point I think; leaderless really means first mob rule and then rule by might. That companies like automattic step up and take a leadership role around the open source tool they champion means there's still a business model, which means wealth is generated, which means we still get to have software and not some kind of Mad Max software scenario.

  • Jillian Ada Burrows

    If you go to the heart of the matter, it's really about a person's (or corporation's) right to help increase innovation by building on a base of work that has few restrictions. I personally like the MIT or BSD type licenses. They allow one to derive a product and sell it. The BSD license allows one to keep parts proprietary and closed sourced (like Apple's OS X, which really is derived from an open source project the XNU kernel and Darwin projects — derived from FreeBSD + Mach Kernel + NextStep).

    I believe that the end goal of the FOSS (Free and Open Source Software) movement is to help keep software innovation accelerating. Unfortunately, the movement can't agree on certain ideals. A number of them went up in arms when they found out that Oracle was going to buy MySql. They wanted to keep the code base from ever going out of the public domain. I agree with that idea. I also think that companies can prosper with open source licensing. In the field of Game Theory, it is common that cooperative strategies end up benefiting each party more than competitive strategies. In this light, the cooperation that the FOSS movement nourishes has had and will continue to have many benefits to the economy (not necessarily to the individual contributor).

    What I've heard and have seen is the software economy shifting to a service economy. In it, all parties benefit from collectively working a project that can be implemented by themselves and people with the appropriate knowledge. This is because they can change not just for the service of setting it up, but also their development efforts to further the project and/or adapt it to their clients needs.

    On the other hand, this “democratizing innovation” lowers the bar of entry. It then becomes the consumers responsibility to manage the quality of work they pay for, but this is often impractical. They often don't have the knowledge to know how good of work they may be getting. However, this may be akin to Microsoft getting the biggest market share even though it has been the most buggy OS since Win 95. (Win 7 might be changing that.) The consumers didn't know how crappy it was, because it seemed easy enough to use. They also couldn't (wouldn't) try out other OSes because of their investment. They never go the chance to see that having to restart your computer every few days is an inconvenience. They were content with it because the interface was easy enough.

    You argument is based on the assumption that companies that win in the market create good, desirable, thoughtful code that is the pinnacle of software innovation. This is often not the case. I just hope that companies learn the cooperating on software is the best way to foster the acceleration of technology. It would be great if more companies formed around open source projects that functioned exactly like their non-FOSS counterparts. They might be able to create viable profits to pay all of the contributors for their work. Other companies will have different approaches to how the code should evolve, the code tree will be forked, and the different teams will compete for the best product. At least, it's an interesting experiment.

  • Eliot Frick

    @Jillian – I appreciate you taking the time to respond. I'm sorry I haven't replied until now. Your thoughts make me wish to expand in another post, which I shall endeavor to do when I have the time. Cheers!

  • Eliot Frick

    @Jillian – I appreciate you taking the time to respond. I'm sorry I haven't replied until now. Your thoughts make me wish to expand in another post, which I shall endeavor to do when I have the time. Cheers!

  • Thomas Leith

    OK, but read Stallman. The end-game is “the software itself is free, but the services provided via the software won't be.” Consider how IBM pays lots of Linux Contributors quite directly to be Linux Contributors, not something else.

    A working copy of Linux doesn’t pay your mortgage or put food on the table.

    Stallman's point, almost exactly — he thinks that a working copy of Linux or Windows or Solaris or Photoshop or Ilustrator for that matter SHOULDN'T pay your mortgage or put food on the table. For him it is fine for a programmer to be paid to implement a feature but not fine for him to sell the same feature over and over again having done no additional work. Stallman has a of labor theory of value and looks at a stream of income derived from IP law as a kind of usury.

    Stallman and GNU (emphatically) do not make a case that software somehow transcends the property rights that made its existence possible in the first place. Far from “transcending” property rights, Stallman and the Free Software Movement are insisting they be respected.

    This is a long conversation and what's here is imperfect. We can talk about it sometime if you like,