I’ve been researching and musing over how open source and business go together like two slices of bread. The contraption tastes good only if the stuff in between meshes well. Over the next few days I want to cover three main themes:
- What is Free and Open Source Software all about, where did it come from, when were the conditions right , why did it come about, and how is it relevant to open source hardware and the work going on at FeF?
- What is the current state of open source hardware, who is doing it, to what degree are they successful, and how do they differ from each other?
- Given the current state of the art in open source business models, how can it be improved on – and what is the actual business plan going to look like?
I look forward to sharing on this topic and hearing feedback.
Thanks Lawrence! This is an interesting study indeed, and I’m looking forward to learning more and contributing some of my own ideas as well. Specifically, I’d like to figure out if the unmodified GPL (per the license itself), can be effectively applicable to liberated hardware, or if important optimizations in licensing approach are needed for resilience.
>when were the conditions right
IMHO, the problem has to be parallelizable, and appropriate norms need to be set.
The task of writing novel or game is not as parallelizable as writing an encyclopedia or operating system.
Even still, the difficulty of coordinating an effort means that it looks very important to set healthy norms early on, and vigorously refine them to keep the community happy/productive/aligned. An existing culture of computer-users helped shape Gnu, and Nupedia brought the culture of encyclopedia professionals to the fledgeling Wikipedia project.
>how do they differ from each other
Two often overlooked differences are the structure of property law, and the ease of un-doing.
Copyrights are gratis, permanent, and by default restrict non-commercial use. Patents automatically cost money, often require a lawyer even to obtain, confer fewer rights, and expire. The patent system doesn’t seem open to the sort of reversals that free software licenses are designed to carry out on the copyright system.
Second, as Wikipedia has shown us, anarchy works when there’s an “undo” button. It’s OK to replace the article on Red Winter Wheat with the word “penis” and nothing else, as long as a revision history is available. But playing with hardware usually has irreversable consequences.
I think these two differences will contribute to the difficulty of open-source hardware.
Lawrence,
My observations are that projects which are aided to their endpoint by the open contributions of many remain open source projects, but that projects which are largely the sole work of a few individuals, end up returning to a proprietary model.
There are perhaps three main costs for a product: design, promotion, and production – if Open Source can mitigate one or two of these costs centers, than the project might be viable as open source; however, if individuals bear these costs alone, then giving away the product for free starts looking like a fools errand.
So the question might better be asked – how can Open Source mitigate the design, production, and/or promotional costs for a Torch Table? Let us assume that a better design could reduce production costs – if say self-rep parts are either lower in cost or faster to assemble. It would be asking a lot it seems to lower the costs of a large xy table substantially by simply redesigning it as it’s mostly a material problem. And the less room for savings-by-design, the less room to share the burden of development.
But that is the physical hardware; what if we look at the electronics instead. Here, there is a clear opportunity to clip into the rep-rap project, and benefit from their momentum.
Say the cost is $1,600 in parts, and that can be reduced by $200 through Open Source – is that a good fit for the Open Source business model? – It seems problematic, and illustrates the challenge of moving the Open Source metaphor into large hardware.
Perhaps the solution is to assert Openness, not merely on the tool, but on all the tool creates. In essence, that any user is expected to contribute their designs for things back to the project, then the number of objects the project can produce increases, and a community with network effect is born.
Good luck either way…
Lightbulbs are, for all practical purposes, open source. People make a living making them, selling them, placing or replacing them, maybe even recycling them. Not a lot to be done with designing them.
So I’m guessing design will have to be part-time, together with the rest. That way you can wait for parts, ideas, and help.
How close are you to being able to sell CEB machines? LifeTracs? MiniTracs? Maybe even simpler things which you can make at FeF? It can be really simple if there’s a market; it just has to be simple to make for you, and difficult for others, cos that’s where the money is.
Part of what you can sell is service, teaching, tuition. Can you go somewhere to make a CEB machine for them? Somewhere where they have specific resources? Hey, folks out there, maybe it’s time to hire Lawrence to make a CEB machine for you, or for your research institution, so he can level his own finances and then go back to FeF a month or three later? Or are we not there yet?
I think the open source business models will develop themselves. Now we’re looking for practical steps, right? So it’s offers versus needs, both yours and other’s.
One has to remember that the hardware in itself will never be free, since we actually have to get something physical: plastics, metal, leather, lumber, you name it. Even if you literally pull these things out of the soil, you originally payed something for the land, be it glass pearls or green paper bills.
What is free, taking the lifetrac as an example, is the description on how to make one, i.e. its another form of “software”. This becomes especially true when you load that description into the cnc table and produce the parts letting the computer do most of the time-consuming work.
Otherwise I think “free hardware” will have basically the same problems and benefits as the free software. Patents, copyright fights etc.
Remember that ordinary computer software can be associated with both patents and copyrights.
I think it is just a matter of broadening the definition of software.
Perhaps, this might be interesting for you:
Open Source Clean Technology Initiative
“Perhaps the solution is to assert Openness, not merely on the tool, but on all the tool creates. In essence, that any user is expected to contribute their designs for things back to the project”
That is definitely a faborable design principle for the project in general. Tell us more about this. In particular, how do we implement a license for this, or how do we enforce compliance with that license?
With respect to asserting Openness on all the tool creates:
One does this by including specific language in the license under which the design, and or device is delivered.
There would be two options, to purchase or make a proprietary machine – which would require a commercial license from FeF, or to agree to the fully open version, which comes with the obligation to contribute useful design back to a library of useful designs.
It would be helpful if the supporting software were able to automatically upload the designs, and more important to browse and download the designs.
This would need to be at the 3d design phase, which is complicated at this point; but that can wait, what is critical is to put into place now, the legal structures under which both of these options are preserved.
There will be overhead requirements for policing functions with respect to the commercial licensing. Are there any easy solutions for this issue?
[…] Open Source Business Models Essays | Open Source Ecology […]