Tuesday, July 30, 2013

Links for 07-30-2013

Podcast: Simon Phipps talks licenses, patents, and the future of open


Simon weighs in on the recent debate over whether we live in a "post open source" world in which, for many projects, some will argue that picking a license doesn't really matter or even be actively harmful. Simon then goes on to offer his views on how the principles behind open source are taking root in many areas beyond software. Finally, how are software patents going to play out?

Listen to MP3 (0:15:01)
Listen to OGG (0:15:01)

[Transcript]

Gordon Haff:  Hi everyone. This is Gordon Haff, Cloud Evangelist with Red Hat, and I'm here on the floor at OSCON with the inimitable Simon Phipps, who is the president of the Open Source Initiative as well as wearing quite a few other hats. Welcome, Simon.
Simon Phipps:  So great to be with you again, Gordon.
Gordon: Simon, as there often is at OSCON, the subject of licensing has come up. I know not everyone's a license geek, but I know you in particular feel that this is kind of important, what's going with the debate around GitHub and this debate about whether we live in this post‑open source world where, hey, who cares about copyright? This business about having open source licenses is really just catering to a bad system. Could you give your perspective in this?
Simon: I believe that open source is about bringing together many parties with overlapping vested interests to collaborate on the overlapping part of their interests. For that to work, you have to create a safe space where people don't have to ask permission to innovate and to participate.
To create that safe space, somebody somewhere has to take a minimum amount of bureaucratic action to create the safe space, because unfortunately, or maybe even fortunately, under world copyright rules, when you create a piece of software, you automatically have a copyright over it, and it's just not safe to use software from somebody else if they haven't given you permission to use their copyrighted work.
As a consequence, open source is sometimes framed in this post‑open source world as needless bureaucracy, but actually I believe it's the minimum amount of needed bureaucracy to avoid needless bureaucracy. It's the open licenses that create that permission‑less space, where you now no longer need a whole lot of bureaucracy in order to collaborate with others, and I think that's the perspective that's being missed in the discussion about GitHub.
The question is not about the needless bureaucracy of copyright. Rather it's about saying that the way of law is forces you to take some action over copyright, and it's irresponsible not to act. You're putting your collaborators at risk. You are endangering the companies that you want to collaborate with, and you know what? Most of them are smart enough to simply not do it.
If you don't put a license on your code, you've not created the safe space for collaboration, and anyone that shows up is a fool.
Gordon: What GitHub has done is that basically they have, at some level, picked some licenses for you, and put some layman's explanation of them. Is that a good approach?
Simon: Yeah, I think it's a great approach. I wrote an article about this last year, about GitHub, and I commented that what was needed was some recognition that a license was required. Now, what GitHub has done is they've added into the process of creating a repository, a step where you either say, "I am knowingly not putting a license on this code," or you say, "I'm going to pick a license."
Then if you don't understand which license you should pick, they've built a very cool website that helps you understand the implications of the licenses you could pick, and it's biased towards their worldview, which is a very liberal licensing worldview that views the MIT license as modern, but nonetheless they take the step of forcing you to think about it, and they give you the tools to think about it.
I think what they've done is great. I wish it had been a part of the original design, but it's there now and I think it's a very good step. The problem they've got though is that there were a very large number of projects created before that came about, and depending on how you view the statistics, there are at least 25 percent of the projects on GitHub that don't have a copyright license, so I believe GitHub needs to go further.
They need to actively flag the projects that were created before this tool showed up, and even if it's just a matter of putting a little red disc at the top saying, "This project may not come with permission to use it, and if you're the project owner, click here to pick a license." That's the thing they need to do.
Once they've done that, flagged all the existing projects, forced people to think about licenses on the way through into new projects, then I think they've gone far enough.
Gordon: That's probably enough license geekery for one podcast. You and I could probably talk about this for a while, and we've got a couple other folks we both know who are here. We could probably go on for about half a day on it, but let's talk about the bigger picture here.
Open source has really, I think, been widely recognized at this point as being about collaboration, innovation, and people not just being consumers, and we're really starting to see this moving beyond the software world.
Simon: Yeah, I think we are. We're seeing the concepts behind open source moving beyond the software world into open data, into civic activities, into hardware. For me, this is all a reflection of a change that's happened in society. I think that we came into the beginning of the 1990s with a world that was the epitome of the industrial society.
The industrial society works by finding natural contours of the landscape where there are control points, where there are pinch points in the flow of goods or the flow of wealth, and corporations then occupy those control points and monetize, and most of the law has been about deciding arbitrating between competing requests to monetize a given control point.
When the Internet came along, and in particular when it became something that everyone had access to, it turned society into a huge peer‑to‑peer society, where instead of just having a consumer role, citizens now potentially have the role of a creator as well, or the role of a repurposer, or they can play all three roles in a given relationship.
Consequently, I think we're beginning to discover that the same principles that drove open source, discovering that you could collaborate with people all over the world without needing to ask permission, and without going through a middleman, I think we're beginning to find those same things beginning to apply in other areas as well, to making things.
3D printers open up manufacturing to everyone, or at least prototyping. Microfinance initiatives open up financing to people. We're going to see a revolution in finance when it becomes possible to crowd found rather than just crowd fund, when it becomes for communities of developers to buy a stake in a project and benefit from its success.
That whole landscape is changing, and I believe it's all down to the Internet turning society from a set of linked control points into a mesh of peer relationships.
Gordon: Clearly the software world's been really eaten in many respects by open source. There are currently some very large successful proprietary software companies out there, but things like cloud computing would clearly not be possible without open source. That would never have happened if it had to buy proprietary software licenses for every scaling cloud instance.
What are some of the particular areas, maybe including some of the ones you just mentioned, where you really see the next one or two things where we'll look back in, I don't know, 10 years or whatever, and go, "The world has changed"?
Simon: For me, the thing that most needs addressing is how legislation gets created. I think that when I look at bad legislation that's around at the moment, like bad copyright law or bad patent law, the difficulty that we have is legislators are still acting as a grand jury, impartially deciding between who gets which control point, and we need to change our legislative process so that our representatives actually are our representatives, and are not just our representatives on the grand jury, arbitrating between the commercial concerns.
I think if we don't make that change over the next 10 years, we'll see society spiral into a pit of despair over the way that large corporations are able to claim control over new innovations, just like we saw many years ago, mp3.com being snuffed out by the music industry, despite Michael Robinson doing all the things you would ever want him to do in terms of fairness and the protection of copyright.
I think that's going to be one big area that's got to change over the next 10 years, legislation. I think that the expectations about salary are going to have to change, not that we are going to have to be on a lower salary, but many of us, either knowingly or unknowingly, enter the stock lottery, where we assume that if we work for a startup company, we're going to make a huge fortune, so we accept our working conditions on that premise, or where we assume that the stock market is going to pay our retirement.
I think that's going to change, because I think that the opportunity to create large amounts of wealth relied on there being control points in the hands of very few companies, and I think we're going to see the opportunity to create wealth distributed more evenly across society. We're going to see people needing more to directly create their wealth, and possibly not have the expectation of being the lottery winner who gets the billion dollar buyout.
New startups, I think, are going to have to scale down their expectations of success, because I don't think the great big pots of gold are going to be there in the mesh society anymore.
Gordon: Certainly software startups have much lower capital requirements, or arguably they did. You didn't have to buy boatloads of your former employers very expensive servers and operating systems, but look at something like Tesla, for example. Manufacturing concerns still have very large capital requirements.
Simon: They are living at a natural contour of the landscape. They're monetizing a natural scarcity in the ability to manufacture huge electric vehicles. There's very little chance of somebody who doesn't heavily invest in that being able to work on it, so there's a natural fold in the landscape that provides a control point that Tesla is able to sit on.
A software startup, if it tries to artificially hide its source code, it will discover that more agile competitors will quickly work in an open source collaborative world and produce competing products that do the same thing, and no amount of pleading that there should be patents or copyrights protecting them is going to be seen sympathetically.
Really, what needs to happen is that software startup companies need to expect to exercise influence over open source communities to create their business advantage, rather than expecting to create artificial scarcities to monetize.
I think there will still be the Teslas of this world, where there are natural hills where you can build a hill fort, where there are natural valleys that the rivers flow through where you can put a turnpike, but for the software world, I think that the expectation of being able to make huge profits out of controlling software needs to go away.
Not because you couldn't do it if you could keep a secret, but because software is too easy for really clever people to write, and you can't hire all the smart people, as Bill Joy said, but you can collaborate with most of the smart people, and you can influence most of the smart people, and naturally, as a consequence, open source projects will dominate in any innovation area.
Gordon: One last question, real easy one. What's going to happen with patents?
Simon: The question with patents, although I would love to see the law change so that software patents were impossible, I think what we're more likely to see is some course corrections in the landscape of software patents. I very much agree with Mark Lemley about software patents, that the problem with software patents is functional claiming is taking place. Functional claiming is where you say that you're claiming a patent on an effect rather than on a thing.
For example, in the world of drugs, lets say that you were going to produce an aspirin headache tablet. You could get a patent on the aspirin for curing headache, and then later you could get a second patent for the aspirin to control blood clots, and the function there is to cure headache, to cure blood clot. You can't patent the idea of stopping blood clots, you can't patent the idea of stopping headaches, but you can patent aspirin to stop headaches, or aspirin to stop blood clots.
What's happened in the software world is the courts have been treating computers as if they were a thing rather than as if they were a concept, so people have been able to write patents that say, "A computer which has a one‑click button on the screen," or, "A computer which play back an MP3 file," and because you're able to combine a Turing machine with a functional claim, people have been able to effectively make patents that patent functions rather than patent inventions.
The most obvious thing that I think will change in patent law is that the courts will wake up and discover that Turing machines are not devices, that Turing machines are concepts, and that most patents that say, "A machine which," are actually invalid. I think when that happens, there will be a great big shakeout of most of the software patenting world.
Exactly how it looks, I'm not sure. Maybe you'll be able to make a functional claim, "A computer with an Intel processor running Linux which." Maybe that will be specific enough to say, "A device which," to make a functional claim.
I think that Mark Lemley's right. The sweet spot to fix is functional claiming. Once that's done, the number of bad software patents will fall away dramatically, because you'll no longer be able to patent an idea just because a Turing machine will execute it.
Gordon: Thank you, Simon. Anything you'd like to add?
Simon: No. It's been great to see you, and I'd be very pleased to chat about some of the things that we've been talking about today in the future, if we can find each other again.

Gordon: Great. Thank you, Simon.

Wednesday, July 17, 2013

Links for 07-17-2013