Aug 20

Tim O'Reilly

Tim O'Reilly

My Tongue-Lashing from Eben Moglen

It created a bit of a stir at the O'Reilly Radar Executive Briefing on Open Source a few weeks ago when Eben Moglen, who'd been invited to speak with me about free software licensing in the era of Web 2.0, chose instead to take me to task for talking about open source rather than free software for the past ten years, and for "wasting time promoting commercial products." A number of people asked to see the video from the session. Even though we hadn't planned to release the video from the executive briefing, we were able to get a copy. Here it is:

And here is the file in ogg format for those of you who, like Eben, prefer not to use Flash.

Some quick takeaways: both Eben and I feel a bit uncomfortable with how this session turned out. I'd like to say that we both demonstrated the truth of Oscar Wilde's dictum that "only the mediocre man is always at his best." I believe that Eben regrets his confrontational tone, and I regret that I wasn't quick enough on my feet or forceful enough to steer the conversation onto the topics I wanted to talk with Eben about. We ended up mostly talking past each other. I do believe that the issues that I invited Eben to talk about are among the most urgent facing free software advocates today, so it was disappointing to me to have my position that Web 2.0 provides some fundamental challenges to free software characterized by Eben only as self-promotional hype, and Google and other centralized data services dismissed as "thermal noise" in the long term trend of the computer industry towards decentralization and freedom.

The closest Eben came to acknowledging my position that the fundamental challenge of the Web 2.0 era may not be free software but free data, and the right of users to view, delete, modify, or freely transfer to a competing service the data that is stored about them in centralized databases, was when he said:

"You used to have Fourth Amendment rights in the United States, which meant that police had to get a warrant after convincing a judge they had probable cause. Now they fill out a subpoena blank and send it to Google or EZpass or whoever it is. I know federal prosecutors who would entirely agree that the constitutional landscape has completely shifted under them in the past decade. They no longer have to worry about the judicial restraints on what government is able to discover, because they have a private surveillance system now functioning that they can subpoena at convenience. That's a deeply important problem for anyone who cares about constitutional freedom in the United States.... I absolutely agree the serious public policy issues of all this centralized information in private hands that can be pushed around by governments is a uniquely serious difficulty for people in the 21st century who care about individual freedom."

I completely agree with Eben about the constitutional and civil liberties risks of the new computing paradigm. But I'm perplexed that that is the only implication he sees. It's a bit hard to reconcile the forceful statement quoted above with the notion that Web 2.0 is thermal noise.

I also struggle with Eben's notion that "we" (the open source movement) missed the opportunity to engage with these issues for the past ten years, when Richard Stallman told me in 1999 that the issue "didn't matter," [pdf], and that even in this session, Eben spent much of his time arguing that the computing paradigm hasn't changed significantly in the past decade. Eben says at the end of the session that he was inviting me to a conversation I've been ignoring. I unfortunately felt the same way, and it's a conversation I've been asking the industry to engage with since my very first public talk on open source in 1997.

Eben also did ultimately get around to acknowledging some of the dilemmas provided by software as a service. He made a strong statement about why the Free Software Foundation ultimately chose not to close the "SaaS loophole" in GPLv3:

"We've got to conclude that what Google does, they have a right to do in freedom. They shouldn't need anyone's permission to run programs. Stallman was right about that at the very beginning. If you have to ask other people's permission to run a program, you don't have adequate freedom. And that means you've got to have the right to run programs for someone else. And what's more, you have to have the right to make private modifications. Because if you don't have a right to make private modifications, and keep them to yourself whenever you want to, then the principle of freedom of thought is being rudely disrupted by a required responsibility to disclose what you are thinking to someone else... So if we take the philosophical responsibility to provide freedom seriously, we're going to have to say ... their rights, properly protected, may conflict with other people's rights, properly protected. The solution isn't to reduce anyone's rights."

I was impressed by Eben's insightful framing of the conflict and the decision that was made. However, he seemed to have no answer, besides negotiation, and perhaps legislation, to address this conflict of rights. It will be interesting to see how the situation evolves. And that's actually the right answer. As Larry Lessig argues in a different context, for many situations triggered by technological change, it's better to wait for the dust to clear before pushing new laws (or licenses.)

Meanwhile, I continue to feel that the focus of the free software movement on "software" rather than on "freedom" is the real lost opportunity. In the first era of the computer industry, lock-in was provided by hardware; in the second era, it was provided by software; today, it is provided by centralized databases driven by winner-takes-all network effects. Focusing only on free software is as limiting as focusing on free hardware. It's freedom that matters. I would have thought that Eben and I could have found common cause there, and would love to have a real conversation about these issues.

tags: open source, web 2.0  | comments: 83   | Sphere It

Previous  |  Next

0 TrackBacks

TrackBack URL for this entry:

Comments: 83

  Swashbuckler [08.20.07 06:48 AM]

I think there's a minor inconsistency between what Mr. Moglen said and something RMS wrote a while back.

Moglen says "And what's more, you have to have the right to make private modifications. Because if you don't have a right to make private modifications, and keep them to yourself whenever you want to, then the principle of freedom of thought is being rudely disrupted by a required responsibility to disclose what you are thinking to someone else."

However, RMS wrote "As a computer user today, you may find yourself using a proprietary program. If your friend asks to make a copy, it would be wrong to refuse."

Those two statements seem at odds with one another. Perhaps RMS's thinking on the subject has evolved.

And regarding freedom of thought, if EB really values that why cannot he accept that there are those that are more concerned about open source than about free software?

  anonymous [08.20.07 07:22 AM]

If you really wonder what's going on here, go back and read "Animal Farm" by George Orwell. That should answer all your questions about "Mr. Moglen" and the MIT squatter Richard Stallman.

  Search Engines Web ‚ñÑ‚ñÄ‚ñÑ‚ñÄ [08.20.07 08:18 AM]

It is refreshing to see such passion and spontaneity in interviews. Sometimes the conscious concern about civility and politeness can dilute the spark and drive from many debates.

Whether of not one agrees with his reactions concerning the downsides of commercialization and marketing - it presents a opportunity for a future spirited debate and challenge with others just a forceful in their opposing ideologies and with equally driven personalities.

  mpg [08.20.07 08:24 AM]

I'm certainly not an apologist for RMS, but let's put the two sentences quoted above quote in full context -- it's not as extreme as it might sound otherwise, I think...

"As a computer user today, you may find yourself using a proprietary program. If your friend asks to make a copy, it would be wrong to refuse. Cooperation is more important than copyright. But underground, closet cooperation does not make for a good society. A person should aspire to live an upright life openly with pride, and this means saying “No” to proprietary software."

  Paul Walsh [08.20.07 08:37 AM]

You may or may not remember Tim, but I questioned why people referred to you when defining Web 2.0, to which you kindly responded by providing your definition again in a new blog post.

I like to see honesty in the public domain like this. But, I can't help but feel Eben was a little brash, even though I usually like the blunt approach. His sarcasm didn't go down well with me on this occasion either as it countered your attempts to respond to his lengthy and articulate comments.

  Swashbuckler [08.20.07 08:41 AM]

I don't think the extra text provides any more useful context.

EB says that you should have the freedom to make private mods that you don't have to share. RMS said that you should always share (refusing to share is wrong).

No matter how you slice it, those two statements are in conflict. As I said before, RMS's views may have evolved - and if they have that's fine.

  Jon [08.20.07 10:04 AM]

Thanks for posting this.

  Alex Tolley [08.20.07 10:24 AM]

I think that part of the problem in the conversation was the possible misunderstanding of definitions.

When I read Eben's article [
it seems that he is including data and other forms of information in the wider meaning of software. In turn, it may be that he sees Web 2.0 as a set of technologies and architectures du jour, rather than the larger concept you have in mind. And of course he is targeting the open source focus of the discussion, rather than the Web 2.0 angle.

If this was a religious argument, Eben would be arguing about the philosophy of beliefs and their transmission, whilst you would be arguing about organizations and whether to build cathedrals versus local churches. To use a secular metaphor, Eben would be talking about the meaning of the enlightment and why it is important, you perhaps more about organizing the school system to support the various curricula.

To extend the metaphor again, open source in his terms just means being able to read and write the texts in a common language, whereas he is arguing that this is a side show compared to who gets to decide what is in the texts and how they are used.

This is a shame, because both sides of the conversation are needed.

  niyam bhushan [08.20.07 10:35 AM]

Tim, thanks for the 'openness' of your response. the debate you raise about users' data versus software is not only valid, but the inevitable next step in this digital-culture journey. some may soon see software as data too, with its own authoring tools... while others may see the lines blurring between software and data. the debate is and will grow in decibels.
Tim, maybe Eben's been a bit brash, but i was actually looking forward to your response on whether you agree or disagree with eben, and whether there are things you've learnt from this that you wish to implement in your evolution of the O'Reilly universe.

niyam bhushan

  JP [08.20.07 10:45 AM]

It seems like this guy suffers from the classic computer science view of "we *should* be worrying about X" when in fact, the market determines and shapes what direction computers and networks like a unconscious organism. its no different than an ant in a colony "thinking" 'hey, we should be gathering more food', when it has little to do with the direction of the organism/colony other than to provide actions, take input, and push output back out.

/goes back to gathering food

  chromatic [08.20.07 11:03 AM]

Meanwhile, I continue to feel that the focus of the free software movement on "software" rather than on "freedom" is the real lost opportunity.

I agree... but wasn't the point of defining the term "open source" to avoid having to explain what freedom meant in the context of software?

  Thomas Lord [08.20.07 11:31 AM]

Meanwhile, I continue to feel that the focus of the free software movement on "software" rather than on "freedom" is the real lost opportunity.

Code is law.


  JJS [08.20.07 12:29 PM]

Code is so much easier than social relationships.

The FSF and proponents of Open Source, Web 2.0, etc are to each other as the developers of the Linux kernel are to those of Apache, MySQL, etc. There are similarities and areas of overlap, but the focus of each group is correctly on its own project.

The FSF has always carried the GPL banner and, IMHO, that license has had the most influence on the FOSS community. It has certainly set the terms of the debate, considering that there is Free Software licensed exclusively under the GPL, and Open Source Software licensed under many licenses with a variety of conditions.

RMS and Eben have always shied away from making public statements about issues outside the realm of software copyright and patents. Thus, I am not surprised by Eben's actions. He may have a personal opinion about 'data freedom', but it is understandable that he would take advantage of a public forum to promote the subject in which he has expertise. Representatives from the EFF would be much more interested, and knowledgeable, in the discussion about data.

Later . . . Jim

RenaissanceCore IDS, check it out at:

  Anon [08.20.07 12:32 PM]

O'Reilly and Moglen are both reasonably bright, thoughtful people, who happen to be approaching an issue from different directions. Moglen is more interested in individual freedoms; O'Reilly is more interested in commercial freedoms.

O'Reilly's view is rather shallow and limited. Commercial freedom is a poor substitute for the real thing. Moglen's view is slightly lacking - the GNU people are missing a small slice of the overall picture in their devotion to the individual. But overall their approach is much better than O'Reilly's.

  Richard Steven Hack [08.20.07 01:00 PM]

There's really only ONE freedom: public domain.

You want freedom? Dump ALL the licenses.

Any license that says you MUST do something is not freedom.

  Tim O'Reilly [08.20.07 01:10 PM]

Anon --

I think you misunderstand my position if you think I'm mainly interested in commercial freedoms. I do believe in the marketplace, and the role of commercial freedom in helping create a vibrant marketplace. But I also spend a lot of time thinking about what kind of freedom users need.

And what I've been saying, which I think Eben misses, is that the rules of the game are changing, and if you care about freedom at any level, you need to adapt.

As Tom Lord said above, quoting Lessig, "code is law." I think Tom meant it to say I was wrong in my statement that the FSF was focusing too much on software and too little on freedom, but the quote captures exactly what I mean. The structure of the code we rely on is changing from "distributed and run on an individual computer" to "run on a server, which much of the value provided by shared data."

How does this change in the way computer applications are delivered change what we need to do to preserve user freedom?

I don't think that most free software advocates have come to grips with this architectural change in the computer industry. And that's why I celebrate real game-changing free software like Brad Fitzpatrick's latest effort. Brad is thinking about free software in the Web 2.0 era. Eben Moglen dismisses Web 2.0 as "thermal noise" and is focused too much on the way that software used to work.

That's what I'm talking about.

  Jose [08.20.07 02:09 PM]

Swashbuckler, Moglen is talking about your right to privacy, that you should be able to make a rought draft (put some thoughts down) and not share it, not distribute it. RMS was talking about the properness of sharing third party software that has already been distributed, that has already been made public (voluntarily presumably). These two views are completely consistent. One applies to something that hasn't been distributed, and the other applies to something that has been distributed. [Note that the modifications Moglen talks about have not been distributed even if the base software upon which it is based has; we are talking about keeping the modifications private.]

  Michael R. Bernstein [08.20.07 03:21 PM]

Tim, if you get a chance to engage with Eben again on this topic, I think a question that might cut to the heart of the matter would be "Do we need a 'Free Data' movement (perhaps leveraging the same or similar copyleft mechanisms) to complement the 'Free Software' movement?"

I personally think that the answer is 'yes'. That is what it is going to take in order to take advantage of the diffuse mass of CPU cycles and data storage that he noted was spread on everyone's lap and cooperatively wrest away control from the overall much smaller but much more concentrated proprietary data silos.

In fact, let me coin a phrase (with apologies to Bill Joy): "Most of the world's computers don't belong to your company".

Centralization still will have scalability advantages, of course, but 'Free Data' avoids the creation of new monopolies regardless of whether the technical challenges of ad-hoc distributed virtual server farms (which is what I suspect Eben was hinting at) are met or even tackled. Or at least, whatever 'natural' monopolies arise will be less able to leverage themselves into positions of control.

  abhishek [08.20.07 03:22 PM]

>> In the first era of the computer industry, lock-in was provided by hardware; in the second era, it was provided by software; today, it is provided by centralized databases driven by winner-takes-all network effects.

We could generalise that. In all these eras, lock-in is provided by broken "intellectual property" laws enshrined in a culture of restriction. To say it is provided mainly by some particular application of these laws is to miss the point.

It seems to me the FSF has concerned itself with freedom in the bigger picture -- whether it's software, hardware, patents, DRM, legislation, etc. -- while you seem focused on the relative minutiae of technological patterns.

I suspect, as it seems Moglen does, that this latest challenge is only another configuration of the same old philosophical and legal structures. I can't see how anyone isn't focusing on the issue of freedom to think so.

But "Web 2.0" (whatever it may be) is interesting in that "restriction" maps more appreciably than before into a cost per person per unit of time. It could expose the value of protecting freedom to a lot of individuals who don't otherwise care.

However, this is incidental, and de-emphasising the fundamental philosophy would be a remarkable misrepresentation. (Kind of like "open source"?)

Anyway, thanks for posting the video and driving the discourse.


  Tim O'Reilly [08.20.07 03:39 PM]

abhishek, you say "It seems to me the FSF has concerned itself with freedom in the bigger picture -- whether it's software, hardware, patents, DRM, legislation, etc. -- while you seem focused on the relative minutiae of technological patterns."

Technological patterns may seem like minutiae to you, but understanding them is critical to strategy. Bill Gates understood where the IBM PC was taking the industry better than IBM did, and the world of proprietary software that Richard and Eben lament was the result. I've been saying for a long time that the web is taking the industry in a new direction that both companies and free software advocates don't seem to understand.

If you want to get ahead of the game, you need to think about where technology is taking us, not just where it's been.

When you say that the FSF has concerned itself with the bigger picture of freedom -- what do you make of Eben's refusal to consider the move to Web 2.0 centralized data-driven applications as even relevant to the cause of freedom? "Thermal noise" is a pretty strong dismissal of something that appears to be a fairly fundamental architectural change in how software is deployed...and thus, how freedom can be either preserved or lost.

  Michael R. Bernstein [08.20.07 03:41 PM]

Small correction to my phrasing above. I meant "[...]overall much smaller but much more concentrated infrastructure of the proprietary data silos."

  ricardo galli [08.20.07 03:44 PM]

I think you are still missing Eben's points in your answers and afterthoughts.

He complained that now is the first time you refer to "freedom" after neglecting it for years while you choose to go to the "pragmatic" open source proposition. You did not answer this specific question.

Eben said that if you instead had chosen the "free software" you could understand better that is _not_ a software license problem but a conflict of rights that can be solved only at a higher level (diplomacy, negotiations, laws...).

So I don't see any big unconsistency in his words and the "implications" he sees. You try to bring a software license debate while he thinks is not --anymore-- a software or license problem.

I also agree with the core message that during last years you didn't exploit your popularity and influence in the hacker community and politicians (and money) to promote the freedom values --instead of the "pragmatic" open source--, although suddenly you want to move the "freedom debate" into your own frame and conference about the much hyped web 2.0 concept (and trademark).

What I saw in the debate is that Eben tried to move the discussion to a higher level --a compromise with individual freedoms-- but you still insist he had to answer in a lower level, which affect only to a very small part of the society and computer industry and cannot be solved with software license --at least according to his principles--.

Now that you seem to be thinking hard about individual freedoms, did you change your stance towards the "free software" and "open source" propositions?


  Mathieu Gauthier-Pilote [08.20.07 04:14 PM]

Regarding the "conflict of rights", my point of view is that the conflict isn't really between the right of people to run an execute programs + the right to make private modifications vs the right of the same people to study the code and share it, now that Web services are prevalent.

The problem is that, somehow, that a lot of people have gotten to think that the rights of individual human beings are to be those of corporations, i.e., social institutions, organizations, groups.

Allowing for-profit social institutions to invoke the right to the privacy of their personal lives makes absolutely no sense whatsoever. There obviously has to be a way for the community of equal citizens that we are to control what groups claiming to render commercial services do with our private information.

The documentary "The Corporation" made a point of the ridiculous notion that corporations should have the protection that individual, mortal, human beings, united as a political body, gave themselves in their national constitution.

The US constitution is broken, as is that of so many other States. It is more than time for the citizens of the most powerful nation on Earth to start the discussion on the propositions they intend to put forward in the next Constitutional Convention.

  abhishek [08.20.07 05:36 PM]

>> what do you make of Eben's refusal to consider the move to Web 2.0 centralized data-driven applications as even relevant to the cause of freedom?

I didn't get the impression that it wasn't relevant to the cause of freedom; only that it was no more relevant than the status quo. Especially considering the reasoning behind the SaaS decisions in the GPLv3, it hardly seems like the FSF is ignoring new trends.

>> Technological patterns may seem like minutiae to you, but understanding them is critical to strategy. [...] the web is taking the industry in a new direction that both companies and free software advocates don't seem to understand.

Absolutely. However, the fundamental philosophy of freedom is invariant, regardless of the industry's whims. This is the bigger picture.

Strategically, I agree that it's important to constantly re-examine the implementation of that philosophy. But again, what leads you to believe the free software movement is failing to do so? Moglen suggested he pragmatically bought time with the GPLv3 to do exactly that.

Nevertheless, the functional difference in this discussion amounts to: in terms of freedom, is Web 2.0 as significant a paradigm shift as, say, the democratisation of hardware? I can only imagine the case for Web 2.0 being all that special is not convincing enough. It's probably not that the free software community is misplacing focus.


  Tim O'Reilly [08.20.07 05:53 PM]

Abhishek -- you say:

"what leads you to believe the free software movement is failing to do so? Moglen suggested he
pragmatically bought time with the GPLv3 to do exactly that."

Well, he made that assertion, but failed to tell us how he'd bought that time. He mainly said that there was a conflict of rights that wasn't soluble in the license, and would have to be worked out between the parties.

And in the rest of his comments, he basically said that Web 2.0 doesn't matter. That doesn't seem to me to be a strong sign that the FSF has come to grips with these issues.

And as I said, what was disturbing to me was that instead of responding to my questions, Eben chose instead to make me a target. I think that this is an important issue, and to me, he didn't appear to be willing to tackle it head on.

But opininons may vary.

  Matthew Sporleder [08.20.07 05:54 PM]

Ha! What an absolute joke that was! The gplv3 has "bought time"? O'Reilly "ignored" the real problems of client's not knowing what was going on behind the scenes on servers? "We've known about this for ten years" I don't even need to reply to those ridiculous statements. :)

Public domain is the most free. If you want credit, try BSD. It allows for commercial interests and private interest.

If you want access to what a server is doing, ask for it. If you don't get it, replace it. It's exactly like conventional software vs open source software. You can even start a little business with your feature-copying-wares. (see zimbra)

  Anonymous [08.20.07 08:23 PM]

How about a low-bandwidth or audio-only version?

  Me [08.20.07 09:23 PM]

flash8 .. bleh...

i'll go cry on my linux powerpc

  Thomas Lord [08.20.07 09:45 PM]

Hi Tim.

Yes, you got roughly right why I cite "code is law" but I'll amplify a bit:

Take some company, let's call it X, and assume that X is a web 2.0 power house: X offers a unique service, based on closely held code. X is hugely popular with some swath of users and, consequently, accumulates a huge amount of highly valuable "back room" data which, perhaps they use to improve their service using Web 2.0 design patterns. So, things snowball: the more useful the data the more popular the service so the more data that arrives. Of course, observant members of the audience are concerned about X's potential for abuse and, regardless, its distorting effects on society. That is pretty much what you are looking at, right?

Well, a good answer there is to observe that company X only gets its data and only manages to leverage its closely held code as long as there is no substitute. What if, instead of using the services of X, a user could choose to use Y, where Y is transparent and benign in its data collection? The possiblity of such a Y comes from software freedom. Indeed, in cases such as Google, if one is paying attention to their technology, free (as in liberty) substitutes seem highly likely to arise.

In that way, Web 2.0 big shots can be both part of a gravely offensive process that is challenging individual liberty and, at the same time, just "thermal noise" along a path in which some rich folks got temporarily ahead of the curve in applying free software to their own advantage but, further along that curve, they lose their magic powers because free software empowers everyone else too.

From a dispassionate remove from a stage show I find that I have some sympathy for the idea that "Web 2.0" is "self promotional hype" but, bear with me, that's a technical criticism not a personal slam. Technologically, it's little more than (a) an incremental evolution of asymmetric corporate surveillance and (b) an incremental evolution in "participatory marketing". Part of the reason that this segment of the industry has grown so is simply that it was, for a time, a space of "easy plays" -- anyone with a spare $100K could have a good go at starting one. It's very winning, in my view, to pimp the "Web 2.0" meme to form a community of interest among people working in that space of "easy plays" but it is a mistake to try to elevate it to something bigger like a historical epoch or a newly formed crisis in the civil order.

Moglen and you are, in part, just talking about different fractal scales of the same thing.

There is a whole separate conversation that, frankly, I'm a little bit itching to have with you. And this is the other aspect of "code is law". The web architecture we have today has some "design artifacts" that really set up the corporate surveillance question. That is, to do something very simple (e.g., host a blog or conduct a search or archive reviews) using today's software means that you *have* to have users contributing to a private database. There's no other way to implement it. There exists a different way of programming the web that rules out that kind of centralization -- that gives users a lot more control and awareness and that, as a technical matter, fails to support data hording as a business model. We're going in that new direction though it is a bit hard to see, still. Some day, you and I will have that conversation, I suspect.


  Tim O'Reilly [08.20.07 10:10 PM]

Me -- did you not notice the link to the Ogg version?

Anonymous, good suggestion about audio only. I'll see if we can get that file up too.

  Tim O'Reilly [08.20.07 10:23 PM]

ricardo --

The fact that Eben claims that I've been neglecting talking about freedom all these years doesn't make it true. I've been talking about the challenges that web applications will bring to the world of free software for the better part of ten years.

I'd suggest you read what I've actually written, and watch some of the talks I've given, before accepting Eben's characterization of my position.

I don't think I misunderstand Eben's position; I do think he either misunderstands or deliberately mischaracterizes mine.

The only reason he sees this as a conflict of rights issue is that he's framing it only as about whether or not it's OK to for someone to make private changes to free software in a software as a service environment. I'm saying that a software as a service environment provides new challenges to those who care about users' freedom. See Tom Lord's comment above. He understands what my point is, even if he thinks it's easier than I do to compete with a Google or other Web 2.0 company.

I do think that the free and open source communities will eventually succeed in a response -- but it will be by challenging the centralized architecture of many web 2.0 databases (and by writing new licenses that have more to do with the users' freedom to own, modify, and share their own data than to modify the software.)

There's a lot of thinking to be done to get this right. But we won't get there if we start by denying that there's any problem to be solved, and insisting that anyone who says there is has no standing because he's been talking about open source rather than freedom.

  Swashbuckler [08.21.07 04:51 AM]

but wasn't the point of defining the term "open source" to avoid having to explain what freedom meant in the context of software?

Nope. It was an attempt to be more pragmatic.

"As the decade progressed, some open source enthusiasts struggled to find a way to reconcile the worlds of free and proprietary software. They believed that the open source and commercial worlds didn't have to be at war, and that free and proprietary software, by coexisting, could benefit both sides. They wanted to figure out how to remove the stumbling blocks of the past and give the corporate world a way to use and contribute to the growth of open source software. Through a combination of savvy public relations and a new licensing paradigm, the "new" open source movement has had notable success."

  steve [08.21.07 06:55 AM]

As is the usual situation, neither of you are completely wrong and neither of you are completely right. I am not completely right either, but here are my comments anyway.

When I posted my complaints against Web 2.0 some time back and alluded to your commercialisation of free documentation, that was some of the same thinking as Eben Moglen.

In that situation, I didn't entirely agree with your direction but I found your understanding was deeper and more nuanced than I had expected. I think that changing and updating your position as you go along is more intellectually honest than sticking to a single idea.

It is a fact of life that commercialisation is necessary for some ideas to grow, and you would not be getting this attention if you had not been successful in your commercial ventures. A purer ideal would be more like myself - typing away in a room somewhere as a person nobody knows.

I agree that it is about the data, not about the software. The software acts as the infrastructure and cannot be neglected, but focussing only on that aspect to the exclusion of all else misses the bigger picture.

Eben is missing some of the picture. I think you're missing less of the picture than he is. Ultimately, further conversation and exploration of these themes will be interesting and timely.

Kudos for taking this on, and respect to everyone involved.

  Jeremy Zawodny [08.21.07 07:11 AM]

Wow. That's kind of sad.

  Michael R. Bernstein [08.21.07 07:18 AM]

Thomas. The factor you're not considering is that X's accumulated mass of data (and the software and hardware infrastructure to deal with it, these things tend to grow in lockstep when you succeed in this space) may give it a natural monopoly that 'Y' can find difficult to compete with, even if 'X' does allow users to move their data.

When a system (as Tim posits in his various 'Web 2.0' essays) gets better the more people use it, whoever has the most users and the most data has not just a quantitative advantage, but a qualitative one. This may hold true in some situations even if the competitor is using the same code.

Tim, something that I found curious was Eben's wholesale dismissal of the 'web app loophole'. His characteriziation of this as 'running software for other people' seems shortsighted.

The FSF has relegated the patch for that loophole to a separate license (the GNU AGPL v3), so we'll have to see whether it gets adopted and enables the sort of ecosystem growth I expect, including dual-licensing business models like MySQL's, in the web-application space. I expect this to be synergistic with 'free data' policies over the next five years.

In my opinion, for the OSI to represent it's position as 'centrist' and 'pragmatic', it absolutely requires that another organization take the ideologically and philosophically pure position. In a way, Stallman's stake in the ground opened up the niche that OSI occupies.

So. OSI is to FSF as CreativeCommons is to _____ ?

Right now there IS no one taking the ideologically pure position WRT Free Data, promoting reciprocal data licensing as a means for ensuring user freedom (the closest we've got so far os 'Open Access'). No one has even formulated what the 'four freedoms' equivalent is for user-contributed data.

  Tim O'Reilly [08.21.07 08:46 AM]

Michael Bernstein -- yes, you've got my position exactly right.

Interesting observation about how the FSF allows the OSI to exist. Humans are funny creatures.

Personally, I don't know why there needs to be so much animus between "free software" and "open source." They are emphasizing different aspects of the same problem. Healthy debate is good, but so often it goes beyond that, into religious wars.

I love the Lao Tzu quote:

People through finding something beautiful
Think something else unbeautiful,
Through finding one man fit
Judge another unfit.
Life and death, though stemming from each other, seem to conflict as stages of change,
Difficult and easy as phases of achievement,
Long and short as measures of contrast,
High and low as degrees of relation;
But, since the varying of tones gives music to a voice
And what is the was of what shall be,
The sanest man
Sets up no deed,
Lays down no law,
Takes everything that happens as it comes,
As something to animate, not to appropriate,
To earn, not to own,
To accept naturally without self-importance:
If you never assume importance
You never lose it.

  Shelley [08.21.07 09:35 AM]

Reminds me of the Gershwin song. Taking liberties:

Things have come to a pretty pass
Our romance is growing flat,
For you like this and the other
While I go for this and that,

Goodness knows what the end will be
Oh I don't know where I'm at
It looks as if we two will never be one
Something must be done:

You say data and I say software,
You say data and I say software
Data, software, data, software
Let's call the whole thing off.

"I regret that I wasn't quick enough on my feet or forceful enough to steer the conversation onto the topics I wanted to talk with Eben about."

A firehose full of ice cold water in the face wouldn't have worked to move Moglen away from his agenda, Tim. He was hostile from the start: to you, the OS conference, perhaps for the fact that this discussion happened in an 'executive briefing', probably all three--but he was antagonistic from the moment he walked into the stage. I imagine, also, the following comment from a previous posting may also have had something to do with it. I heard echoes of it in the emphasis on ten years and freedom:

"On the one hand I agree. When a vendor delivers SaaS, they are delivering a service not the software itself. And that's why the GPL and other licenses conditioned on the act of distribution have no force. But if we step back, and think about the issue of freedom, and ask ourselves what freedoms we'll be wanting, say in a world in which Google has the monopoly power that Microsoft had in the 90s, they are very different freedoms than the original freedoms sought by the GPL, but no less important. The freedom to improve and modify are truly valuable -- the question is how to make those freedoms available in a world of SaaS, and more importantly, SaaS in which the service includes not just software but collective intelligence in the form of vast databases -- that include data that I've contributed, and that is no longer under my control...."

Your mistake wasn't on not being able to redirect Moglen's attention, as much as it was the fact that you invited the wrong person to speak. The FSF, and the organization's legal arm, are irrelevant to any discussion related to web services, data, and user's having control of what data is stored about them.

Open source is also irrelevant, because it does not matter, at all, if the service software is open or not if we still don't have open access to the accumulated raw data, and we still don't have a way to control the data for ourselves (leaving aside how easy such services make for the government to access this data).

You say software, I say data..., which you highlight when you said, position that the fundamental challenge of the Web 2.0 era may not be free software but free data, and the right of users to view, delete, modify, or freely transfer to a competing service the data that is stored about them in centralized databases....

Having said that, though, Moglen has a valid point: we should be talking about rights at a level divorced from software, licenses, and implementations and we should be doing so in a context outside of an executive briefing, and outside of a general framework of, "How can this be open sourced, and how can we monetize it?"

It seemed to me he wasn't disagreeing with your premise, as much as he was disagreeing with the context in which you were having the discussion. This in addition to some defensiveness because, according to the FSF and current licensing, Google is 'good', yet your implicit premise is that Google is or has the potential to be 'bad--and all for the exact same reasons. Could have been an interesting debate, but unfortunately Moglen made his argument in a classic lawyer manner, which in our country is based on an adversarial form of dialog.

BTW, Web 2.0 is dead, long live Web 9.75.

  Michael R. Bernstein [08.21.07 09:41 AM]

Tim, this is a political and ideological debate, and some of the positions taken are certainly justified on moral and ethical grounds (even moral or ethical absolutes), but I think that characterizing any of it as 'religious' goes too far.

To my way of thinking, this is a form of subtle demonization, an ad-hominem (just as much as the ones Eben engaged in), and one I would personally like to abolish from our discourse. None of the prominent advocates on either side is taking anything on faith alone, except perhaps faith in human nature.

I also think we need even *more* healthy debate. in politics, you have Left Libertarians, Right Libertarians, Anarchists, Minarchists, Anarcho-Syndicalists, Constitutionalists (constructivists, originalists, and purposivists), Anarcho-Capitalists, etc.

Not that these schisms and ideologies are necessarily all well-represented, but at least they have some claim to distinct existence. So far, all we've got are the permissive and reciprocal flavors of Free Software (and, have you noticed that the GPL vs. BSD flamewars have largely died down? That debate is about a theoretically pure freedom vs. a slightly (and pragmatically) diminished set of freedoms that cannot be diminished further).

Open Source was devised as a 'Big Tent', and as such it's proponents have always had to struggle to articulate (at least to people outside the tent) why enlarging the tent in some directions was not acceptable (ie. the 'Commercial Open Source' licenses and vendors). Pragmatism is always vulnerable that way.

So asking 'why so much animus' seems naive to me. I'm surprised there isn't a lot more. a LOT more.

  Simone Brunozzi [08.21.07 10:03 AM]

"Free Data" really seems to be the key, but under "free data" I also
put an open and free tool to USE that data.
Google is pretty good and honest in providing search results, but this
can change in the future, and people would not be able to access some
informations just because there is a commercial interest in doing so.
On the other side, there could be an "open" search engine that allows
people to exercise their rights on their data.

Apart from that, I really can't understand WHY today there isn't a company that:
- offers a web 2.0 service
- uses only open source software to run it
- gives users ALL the rights on their data
- wins over competition

Take email, for example, a very basic service.
There is NO way to have the kind of "data" freedom you hope for,
unless you run your own email server.
Same problem with landline, or mobile phones, and expecially with phone numbers.

I think now in 2007 we HAVE to compare email, and phone numbers, to
NAMES. My parents pick my name when I come in the world, and I keep it
my entire life without paying a dime. I think I, as a 2007 citizen,
deserve ANY right on my email; I agree on paying someone to offer me
that service, but I also need the right to move to another service
provider whenever I think it's the case, bringing my data with me.

So, a question: why this dream company isn't on earth yet. What's the
reason? What do you think?



  Tim O'Reilly [08.21.07 10:32 AM]

Very thoughtful analysis, Shelley. And thanks for the support.

I will say, though, that I was struck by Eben's intelligence and the importance of the points he was making. I didn't disagree with his core thinking about the conflict of rights issues at all.

I wasn't prepared for the hostility -- and while your analysis may be correct, I still won't say I understand it. I've certainly said some things that no doubt raise Richard and Eben's hackles, but I've generally thought of myself and the open source movement in general as fairly supportive and complementary.

I would love to have Eben thinking about the issues I was hoping to raise, since he obviously is a deeply passionate and knowledgeable legal thinker. But if he thinks they are irrelevant, I guess that's not going to happen.

  Tim O'Reilly [08.21.07 10:47 AM]

Michael --

I don't think we are using the term "religious" in the same way. It doesn't mean "based on faith" in the context I'm using it. It means that certain free software advocates divide the world into those who agree with them, and those who don't, and demonize those who are on the other side.

I hear you that saying that the other side is "religious" is a kind of behavior that is analogous to what I'm condemning. I'll try to watch that.

Let me say instead that my preferred use of the term "religious" is in the statement I've often made: "Open source is science, not religion." That is, there are no articles of faith. We're trying to find out what's true, and what works. We want to understand the behavior of the free and open source ecosystem, and the benefits of the free and open source processes, and learn from it.

That's what I've tried to do.

  anonymous [08.21.07 11:04 AM]

The poetry and drama is killing me. This is wholly about a struggle for the minds of software developers - and through Moglen's conduct in this episode with the ever "Why can't we all just get along?" O'Reilly, the winner is Microsoft. Moglen and Stallman's captaining of the naive masses of GPL programmers is grounded in Stallman's study of the techniques of Vladimir Lenin and Ignacio López de Loyola. Both Moglen and Stallman apparently know history, politics and propaganda a lot better than their programmer fanboys, who behave like "convenient idiots" at the service of the Bolshevilks at the FSF and SFLC. Maybe begin by asking who are the athiests and what they really want to accomplish, and then you will find the answer to all the other questions. The roots of this run much deeper than a philosophical discussion about software and data.

  nobody [08.21.07 11:12 AM]

Thanks for posting this! Is there a way to improve the quality of the audio? It really bad.

  Michael R. Bernstein [08.21.07 11:33 AM]


Tim, in that case, you're saying that 'open source' is more akin to a school of economic thought, rather than political thought.

Plenty of schisms there too of course, but note that political ideologies all too often ignore economic realities, often by dismissing the concerns of economists as 'theoretical', or finding a few convenient dissenters and then saying 'truth must be somwhere in the middle.

  Anonymous [08.21.07 11:56 AM]


I'm just going to throw in a tangent. I have been following your posts Tim (and your comments to Eben concerning SaS privacy and freedom) about SaS freedoms etc and I wondered if you follow the digital arts in europe very closely? These topics have been high on the list here for some time and the upcoming ARS Electronica festival in Austria is called 'Going Public' and they state on their website the topic is about "our (involuntary) digital transparency and the (voluntary) relinquishment of our privacy".

I think the arts have a lot to offer these debates and seem to often precede the discourse in .com spheres by some years. Here is, for example, an art project from 2001 or so dealing with these issues -

It allows you to create false profiles of yourself for confusing Google. Its a strategy for protecting your data by confusing the trails you create when you use SaS.

Anyways, just thought I would drop that in. Maybe if you want a break from ranting US lawyers you might want to drop into Austria and check out ARS Electronica ;)


  Shelley [08.21.07 12:54 PM]

"I wasn't prepared for the hostility -- and while your analysis may be correct, I still won't say I understand it. I've certainly said some things that no doubt raise Richard and Eben's hackles, but I've generally thought of myself and the open source movement in general as fairly supportive and complementary [...] I would love to have Eben thinking about the issues I was hoping to raise, since he obviously is a deeply passionate and knowledgeable legal thinker. But if he thinks they are irrelevant, I guess that's not going to happen."

I wouldn't assume it isn't going to happen. I don't know the backlot story on this discussion, but Molgen had all the mannerisms of a person who was feeling aggressively defensive even before getting on stage. If you think on it, it's understandable.

The free software movement has always been on the side of the angels, and now all of a sudden, thanks to Google, and Yahoo when it comes to that, and recent issues having to do with data, privacy, and the law, FSF finds itself dancing with the devil. Must be a heck of an uncomfortable position.

If I were you, or the open source movement for that matter, I wouldn't necessarily take it either personally nor consider the conversation permanently closed. Just an opinion, but I don't think timing was that good with the recent GPL discussions, followed on so many stories of Google/Yahoo privacy issues this year--the context was explosive. (2008 is definitely the year where "Web 2.0" is going to have to pay the piper.)

There were so many issues and topics introduced into such a brief time, most at cross-purposes, as you noted that yes, the moment to have a meaningful dialog was lost. It was too bad, but not that much--this was an exclusive meet up.

I hope this discussion and others related do continue at a different time, in a different venue. Perhaps a woman moderating might add a more civil note to the proceedings.

*blink* *blink*

  Clearing My Throat [08.21.07 02:04 PM]


OK let's see it. Let's see all the unpublished code you wrote in perl and python to run your website. Let's see your SQL and shell scripts which create the graphs you publish on the computer book industry. Let's see the modifications you've made to movable type to moderate and de-spam. Let's see it all.

Oh sure, the GPL, the Perl Artistic, the BSD license -- they don't require it. But as agitators for the "SaaS Hole" meme you'll want to be the first to plug it by setting a good example.

Let's see it!

  Tim O'Reilly [08.21.07 02:22 PM]

Clearing My Throat --

From your comment, I don't think you understand my position. I have always been a supporter of the right of developers to keep their code private, and have preferred BSD style licenses to GPL for that reason. It should be the developer's choice when and how to share. (Although I also support the share-alike bargain expressed by the GPL -- I give you my code if you're willing to follow the same rules if you build on top of my work.)

And I've made no suggestion that the SaaS loophole be closed. I agree with the FSF's decision, and think that Eben's reasoning is sound.

The warning I've been trying to sound is simply this: given the trend in SaaS applications to build collective value in a network-effects driven database (see Michael Bernstein's comment above), I've been trying to warn people about the possibility of new forms of lock-in. It seems to me that "forewarned is fore-armed", and that by thinking of this issue now, before any one company has too much power, we can work towards a more open future.

I actually don't think that Web 2.0 companies are wrong to try to accumulate this kind of market power, and as long as they use it wisely, it may even be a good thing. But I also think it's really good to start thinking about what kinds of freedoms will matter to users when applications depend on *their* aggregate data. What kind of control will they want to have over its correctness? its privacy? its re-use? its portability?

I don't have a specific set of prescriptions (although I strongly favor open data policies like the Wesabe open data bill of rights), just a warning that if we don't think about it now, we will likely get into trouble later.

I find it ironic that Eben claimed that I haven't been thinking about freedom, when I've been trying for years to get this particular issue of freedom on the industry radar for the better part of a decade.

  Rasmus [08.21.07 03:44 PM]

You just proved my point.

  Michael R. Bernstein [08.21.07 04:53 PM]

Shelley, you might be interested in watching Eben speak at (and more importantly, to) Google:

Tim, I do think that the ability (not a requirement) to close the SaaS loophole is important for user-facing web-applications, especially ones that do commonly get distributed, but could also potentially be deployed in a SaaS fashion, because it is in that particular intersection that you see less software being created precisely because without an Affero-like copyleft protection it could be suicidal for a SaaS startup to release their crown jewels.

If we want to see Free Software Businesses based on web-applications that DO get distributed, and can still grow to be serious SaaS contenders (especially as active components of a Free Data ecology), then being able to optionally close the loophole is both important and useful, particularly in enabling various sorts of dual-licensing business models.

It is insane that every single photo-sharing site that Flickr inspired was written more-or less from scratch. Where is the 'MySQL of Flickrs'? This isn't quite as extreme for some other Web 2.0 application types ( and Digg for example both have Free Software clones, and there are plenty of webmail apps), but few will ever invest in adding SaaS-scale deployment features to them and releasing those improvement in the absence of meaningful protections for the commons.

In my opinion, once the GNU AGPL v3 is finalized by the FSF and approved by the OSI as OSD compliant, we will see many more SaaS-capable web-applications both released as Free Software *and* deployed as SaaS businesses, participating in various kinds of free-data commons, and further accelerating the demand for standardized utility computing services (because in this environment you *can* meaningfully fork a web 2.0 application).

  Shelley [08.21.07 05:14 PM]

Thanks Michael, I'll add it to my to be watched list for the future. I've already spent too much time related to this thread.

  Michael R. Bernstein [08.21.07 05:24 PM]

Oh, and Shelley, I think part of the back-story here is Tim's claim that software licensing was irrelevant in a Web 2.0 world, more or less.

Tim has since clarified that what he really means is that (given that Web 2.0 applications themselves typically depend on Free and Open Source software for their existence) such licensing is necessary but insufficient, but some damage was already done. Perhaps the clarification bears repeating a few more times.

  Roy Schestowitz [08.21.07 06:45 PM]

Bravo, Eben. There's more to life than personal gain.

  Tim O'Reilly [08.21.07 06:52 PM]

Michael --

I totally agree that the AGPL might be a really interesting solution. It was brilliant of the FSF to leave the SaaS loophole in the GPL but to have a separate license that closes it. To do otherwise would of course have made GPLv3 a non-starter. But the AGPL will allow Web 2.0 and SaaS players who believe in the free software ideal, or just want to get other developers working on their software, to harness that outside community under a "share-alike" kind of license.

Matt Asay makes this point well in a blog post entitled GPL is the new BSD in Web 2.0, and why this matters.

I disagree however with your assertion that I said that software licensing was irrelevant in a Web 2.0 world. What I said was that current licenses, in which various conditions were triggered by the act of distribution, were obsolete -- i.e. in need of rethinking to come to grips with the new architecture of how software applications are being delivered. That's a very different statement. "Obsolete" is a provocative word to be sure, but it has a very different meaning than "irrelevant."

It is certainly true that remaining obsolete can doom licenses to irrelevance, but they aren't there yet, nor does this mean that licenses in general are irrelevant, just that the ones that were based on the software industry of the 1980s need to be updated.

(And I'll note that I started making this statement in 1999. It got widely publicized when I made it a focus of my remarks at Oscon last year -- but I did that because I didn't see evidence that the FSF was really coming to grips with the issue.)

Even now, I sense a kind of reluctance to talk about the issue. Frankly, at Oscon this year, I was expecting Eben to say why the AGPL strategy was developed, and how it addresses at least some of the issues (even if it doesn't touch new issues such as data portability and freedom.)

Someone who has more insight into the politics of this whole issue at the FSF might be able to shed more light on the backstory behind the GPLv3/AGPL strategy. How coordinated was this effort? And how much was it just two approaches that couldn't be reconciled, and so went forward separately?

In any event, it will be really interesting to see the uptake of the AGPL. I can't see it being retroactively applied to any GPLd software, but I can definitely see it being used by someone who doesn't want to reinvent the wheel in web applications, and would like to get an external developer community cooperating on software for a web application.

(That being said, I still believe that without a decentralized architecture for data sharing between instances of such applications, they will still turn into winner-takes-all (or at least takes-most) applications because of the benefits that accrue to the company that has the most data and the most ecosystem activity. If that's not true, if there aren't real network effects associated with the application, such that it gets better the more people use it, then it's not really a Web 2.0 application, at least by my definition.)

  Tim O'Reilly [08.21.07 07:00 PM]


I actually agree with that sentiment as well ("There's more to life than personal gain."), and have spent most of my career working to live up to one of my company maxims, "Create more value than you capture." That's why I have some reason to be put out by Eben positioning me as some kind of crass capitalist who doesn't care about the bigger issues, but only about making money.

I believe that making money is like putting gas in the car: something you do to help get where you're going, rather than where you're going.

Go read some of the essays on my business philosophy over at

  Peter Rock [08.21.07 07:19 PM]

Tim says:

"I have always been a supporter of the right of developers to keep their code private, and have preferred BSD style licenses to GPL for that reason."

I don't understand. What does the license have to do with the right to keep their code private? Whether GPL or BSD, the choice of keeping code private or sharing is independent of the license (assuming a FOSS license that gives you the choice).

If what you are saying is, "I have always been a supporter of the right of developers to release free software under a proprietary license." then please say that. Don't use the phrase "keep their code private". Privacy is a right that I support. It is an essential freedom. To make "privacy" synonymous with "proprietary" confuses the issue.

Can you please clarify what you mean?

  Steve [08.21.07 07:25 PM]

What an incredible waste. Would Eben have been any less gracious on a podium with Bill Gates? As someone with considerable sympathy for the free software movement, I would have really liked to hear a thoughtful discussion of the topics addressed. As it was, I thought the issues you and he were raising were important and difficult ones.

  Patrick Mansted [08.21.07 08:03 PM]

Thanks for the video, particularly for non-flash video (although the volume seemed very low and the video quality could have been better. From looking at it I would say it was transcoded from the flash version, using the original source ma y yield better results).

  Roy Schestowitz [08.21.07 09:49 PM]


> I actually agree with that sentiment as
> well ("There's more to life than
> personal gain.")

I haven't a doubt about. I admire your work and I just think that the conversation enables many people to just keep on eye on balance (striving for reciprocity and escaping egocentric consumption).

> Eben positioning me as some kind of
> crass capitalist who doesn't care about the
> bigger issues, but only about making money.

Yes, he was a bit over the edge, but he admitted being grumpy (maybe caught in the heat of the moment). Without so-called Web 2.0 and companies like Google, the value of Free software would be harder to reason about. So I, for one, welcome all of this.

> I believe that making money is like
> putting gas in the car: something you
> do to help get where you're going,
> rather than where you're going.

Excellent analogy. I'll try to remember that one.

  steve [08.22.07 04:50 AM]

To the other Steve... could you check which names are being used before you start using the same name as an existing person?

  Sachin [08.22.07 09:57 AM]

To me it seems that Mr. Eben talks about 'Freedom' in a philosophical way but has a blind spot because of the way he holds this grudge against you. So even if you did not talk about it in the way Mr. Eben wished you should, what then? You are free to talk what you legitimately felt you should be talking about, you are free to pursue your priorities.

In his own words, he should have used 'subtle diplomacy', 'negotiation' to bring you around his point of view. How does such collar grabbing help?

But the man does have great insights shorn of his rhetorical rant.

There is clearly no readily available solution to the issue he raised. Ruing the lost 10 years, and putting it as if Tim, you alone were responsible for no solution seems bizzare!

  Steve 2 [08.22.07 12:13 PM]

Steve - sorry.

  Doug Cutting [08.22.07 01:10 PM]

It turns out that AGPL doesn't really plug the SaaS hole, at least not according to Eben. I asked him about this a few years ago when trying to figure out what license to use for Nutch. I wondered whether any license could require all sites running modified versions of Nutch to share their ranking algorithm, thus ranking transparently. Eben explained that folks can easily plug in proprietary extensions, and that no license can really prevent that. They ran into this with gcc, where folks can always pipe intermediate code through proprietary optimizers. Most open source software is extensible, and thus prone to this.

  Joshua Gay [08.22.07 02:51 PM]

Hi Tim,

I'm not sure I agree with you completely on this statement:

Meanwhile, I continue to feel that the focus of the free software movement on "software" rather than on "freedom" is the real lost opportunity.

Because it makes it sound like they are somehow binary switches -- one or the other, instead of both. I think the movement has done a great job of focusing on user freedom in general.

First off, the focus of all of our licenses, the GPL, FDL, and the AGPL are strongly focused on user freedoms that goes beyond just "software."

Secondly, we have a good track record of taking strong stances against many issues through mailings, public outreach and education, and protests. These issues range across standards, documentation, data, and more --- with each issue being focused on user freedom, and not just software freedom

And lastly, you would certainly attest that campaigns like Defective by Design are good examples where the free software community has taken on an issue that goes well beyond "software."

But, most importantly, please don't distance yourself from the free software community -- you, too, can be a part of the free software movement and join us in actions, events, and discussions so that we can further discuss your thoughts on user freedom and what good next steps, today, should be.

I would suggest becoming an associate member of the Free Software Foundation as well as signing up to some our mailing lists so you can see all of the great freedom related discussions and actions that are taking place that go beyond software.


Joshua Gay
Campaigns Manager
Free Software Foundation

  Michael R. Bernstein [08.22.07 05:19 PM]


You're right, I was thinking of your 'obsolete' comment, and the associated foofooraw. My apologies for potentially adding to the confusion with my comment here.

I will note though that you did use the word 'irrelevant' in a similar context at least once (and this is what caused the conflation in my mind); this is some text from your 'What is Web 2.0' talk (yes, I know that this is somewhat taken out of context):

"Software licensing and control over APIs--the lever of power in the previous era--is irrelevant because the software never need be distributed but only performed, and also because without the ability to collect and manage the data, the software is of little use."

Let me say again that for MY part you have long since clarified what you mean by 'obsolete' 'irrelevant' and similar statements (something more akin to 'necessary but insufficient'), but I also think (or rather, guess) that those and similar soundbites probably had something to do with Eben's animosity. I have no inside knowledge to support that intuition, though.

I also was expecting Eben to talk about the AGPL, and was disapointed that he didn't do so. If you do find someone who can give that backstory, I'll read/listen/watch with great interest.

Eben's talk at Google (linked to in an above comment of mine) did go into some length about why an AGPL-like provision was not put into the GPLv3 (although I was surprised at how emphatic he was that such a provision had no chance of being incorporated, despite public perceptions to the contrary). Basically he holds that the right to private modification is too important to abridge without some serious external danger.

Consequently, even though he hasn't spoken directly to this point, I get the feeling that he is somewhat hostile to the idea of AGPL-like provisions even in a separate license. This, in fact, is why I was hoping the AGPL would be the primary topic of your conversation. Oh well.

BTW, I certainly think that relicensing existing projects from the GPL to AGPL will not be a *common* choice (relicensing is a relatively rare event in any case), I do think that some projects will do so (and that some of those projects might well fork as a result). Any user-facing web-app project that is commonly deployed by hosting companies as a 'value add' with proprietary enhancements and/or the credit stripped off would be a candidate. Don't underestimate the power of continuing annoyance in this context, particularly for small projects. Some developers are definitely going to scratch that itch.

Joshua, the FSF's Deffective by Design campaign is (as I understand it) mainly motivated by the fact that DRM cannot be implemented meaningfully by Free Software (ie. you have to abridge the user's freedoms to make DRM work), and is mostly a targetted extension of the longer campaign against proprietary data formats in general. Similarly, the GFDL is motivated primarily by the need for Free Texts as manuals for Free Software. Both of those however tie directly back to the need to ensure users' freedoms WRT software.

Upon some further reflection about this whole episode, I think that the general need for 'Free Data' may be largely orthogonal to the core competencies and interests of the FSF and the SFLC. There are certainly areas of overlapping interest (just as there are with, say, Net Neutrality), but RMS has stated on various past occasions (I am paraphrasing from memory) that while there are certainly other important issues in the world, some arguably much more important than software, he didn't know how to solve those other problems, and that the problem of freedom for users of software was the one he DID know how to fix, so that is what he concentrates on.

Free Data may simply be too far removed as an issue from that core competency, at least for now.

  Joshua Gay [08.22.07 07:50 PM]

Michael R. Bernstein

"Both of those however tie directly back to the need to ensure users' freedoms WRT software."

Yes, Michael, the free software movement does tend to be rather heavily focused on software freedoms.

There are free data movements and free culture movements that are very active and blooming, and many of us are friends, and many of us are part of both movements. But, you understand that already.

That aside, Tim's sentence can be read a number of ways. Personally, it was an irksome last sentence to a rather inconclusive and confusing post.

What irked me is that a schismatic open source movement leader (parse it as you'd like) was calling the free software movement out on not concentrating on freedom, and this 'calling out' was itself a response to the, at the time grumpy, but always diligent, free software movement leader calling him out for not focusing on freedom over the past ten years. But, why?

It certainly wasn't about free data. Eben addressed that in several ways, directly in the discussion, and I'm sure they both understand that.

But, anyhow, here's what I think. If Tim really wants us to be focusing on freedom, and he sees some way that we can do that better than we already are and have been, and he means that, then there is plenty he can do. But, if he can't put his money or his time where his mouth is, then I'm not convinced that statement was really about freedom. Instead, I think that statement is just directed at Eben, and for reasons that are none too impressive.

Joshua Gay

  Tim O'Reilly [08.22.07 09:36 PM]

Joshua --

Let's try to end the "who's more focused on freedom" subthread here. I'm sorry I've said anything that's contributed to that.

I've tried to keep this discussion to issues, and I'm not going to spend a lot of time responding to comments like "can't put his money or his time where his mouth is," and "reasons that are none too impressive," which are both directed at me personally rather than at the issues, and also devoid of specifics. What money or time would I have to spend to get your stamp of approval? Be a member of the FSF, as you suggest in your first comment? That's a pretty narrow definition of spending money and time. There are quite a few people who can attest to how I've spent money and time on the cause of freedom, just not the exact same freedoms as the FSF concerns itself with.

Obviously, the FSF has been focused on many kinds of freedom for many years, and it's a central focus of the organization. My comments about what it seems to be missing are based on direct interactions. In 1999, Richard Stallman told me "it doesn't matter" what Google does on their computer. It only matters what's on my own computer. That seemed short-sighted. Similarly, when Eben Moglen refers to a fundamental paradigm shift in how software is delivered as "thermal noise," that doesn't suggest that he's taking seriously the issues I'm trying to bring forward.

Now it's perfectly OK with me if the FSF doesn't want to focus on those issues. But it was a bit surprising to have Eben agree to come discuss them, and then instead spend time telling me how something (I'm not quite sure what) is all my fault. I've spent time trying to put a particular set of issues on the radar of the free software movement -- just like I've spent time trying to put them on the radar of businesses making architectural decisions about how to deploy their software.

It would be nice if there were some indication that the FSF even understood the issues I'm trying to highlight, even if to disagree with them, rather than continuing to say that I don't have sufficient standing as a freedom fighter to be taken seriously.

It's certainly true that I'm not trying to politicize the issues that I'm highlighting, and that may be confusing to you. I'm just trying to say "here's how the world is changing; think about the implications," to anyone who will listen. Unlike the FSF, I don't see this as a battle between good and evil, but as an evolutionary landscape, where we can be more or less fit, and where the choices we make can make the world we live in better or worse.

It's sad that somehow that seems to have put me on the FSF enemies list. It shouldn't have to be that way.

And by the way, I'd love to know where in the discussion you think Eben addressed free data. In some followup email conversations we've been having, he's indicated a willingness to talk about our respective ideas here, but I don't think they were addressed at all in the session.

  Robert H [08.23.07 07:16 AM]

What do you mean by "free data"?

  Joshua Gay [08.23.07 12:41 PM]

Also, you may want to check out all the discussions that are happening at the Open Knowledge Foundation and perhaps jumping in on their discussion list discussion list.


  Clearing My Throat [08.23.07 05:44 PM]

You can't meaningfully call the "Saas Loophole" a loophole when freedom zero of one of the licenses with the loophole is:

The freedom to run the program, for any purpose (freedom 0)

It's not a loophole, it's the premeditated, considered intent of the license. It's been thoroughly reconsidered for GPLv3, and it stands.

Apparently you agree with the sentiment:

"And I've made no suggestion that the SaaS loophole be closed. I agree with the FSF's decision, and think that Eben's reasoning is sound."

So you have no problem with freedom zero, yet you label it a loophole, and hence: a problem.

You're not communicating clearly.

  Tim O'Reilly [08.23.07 07:35 PM]

Clearing my throat --

You're right. I'm using "the SaaS loophole" as shorthand to refer to the issue that SaaS developers are not bound by the distribution provisions of the GPL, because they don't have to distribute software. Whether or not you consider this a true "loophole," it is definitely a change in the fundamental conditions under which free software is used. The question is whether that architectural change raises new issues that need to be addressed, whether about the software itself or (as I principally argue) about other kinds of freedoms that are potentially constrained by SaaS applications.

In particular, I have tried to highlight the issues that are raised by the fact that Web 2.0 SaaS applications typically collect a great deal of user data -- in fact, my definition of a Web 2.0 app is an app that gets better the more people use it (largely because of the data it collects.) What rights do users need to have to:

* remove that data?
* modify that data?
* transfer that data to another provider?
* build value added services against their own data?


If it's freedom that matters, and not just free software, these are questions that need to be considered.

  Michael R. Bernstein [08.24.07 12:12 PM]

Tim, it is worth noting (and Eben made this point) that in one sense this loophole isn't new at all (he gave the example of a proprietary compiler service based on GCC). It's been called the web-app loophole, the ASP loophole, and now the SaaS loophole. Note that it has always been commonly called a loophole over the past several years, so you don't need to apologize for that.

In any case, what is truly new here is that for new user facing applications, this has now become the *default* deployment style, rather than the exception that proves the rule.

In other words (with apologies to William Gibson), the future was already here, only now it's evenly distributed. ;-)

Calling this a loophole is completely valid, because although Freedom 0 suggests you should be able to run a program on behalf of other people, it is a loophole in the requirement to provide source to all users of a program, and in so doing, you abridge their Freedoms 1-3.

Just because closing the loophole for all GPL software would impose costs that are too high to make it worthwhile, does NOT mean that it isn't a loophole. It is.

  nim [08.25.07 11:29 AM]

I just watched the video. Even if everything that Eben said is accurate (I personally agree with him on mostly everything), he was somewhat rude, while Tim and the audience were being very polite. The recap text of this blog entry is also polite (but still quite informative) in describing what happened... :)

  anonymous [08.28.07 07:12 AM]

This video proves that it is not only geeks that are too-frequently lacking social skills. Some lawyers for geeks apparently are socially inept as well.

  Sarah Claire [08.28.07 10:36 PM]

Tim, you are the biggest fraud to perpetrate a false business model on the internet. Promoting the confusion you monetize with books and conferences that explain it.

  Chris Travers [09.04.07 04:37 PM]


I think that we have to be extremely careful about adopting a conflict mentality. If we worry too much about "open source vs proprietary software" as a conflict or struggle, we risk signing away our freedom for the sake of security. In short this is whay we are told: Use the GPL or the Big Bad Companies will steal your code. This is no different than being old that warrantless wiretapping is OK because otherwise terrorists will destroy our freedom.

The GNU GPL v2 provided a reasonably fair and generally useful framework for protecting a commercial product in the first stages of release into a Free/Open Source Software environment. It also provides a tool for businesses to continue to use open source code without giving up control over the direction of development. As a result we have seen the rise of products like JBoss, MySQL, and Asterisk, perhaps at the expense of real community open source projects like PostgreSQL and Bayonne.

The issue is not one of licensing or legalities. The BSD License is *as good* at encouraging contributions to the community as the GPL is, if the community wants to use it in that way. The issue is one of committing to a community-oriented development model which exists between rather than within businesses.

(I know some people will have issues with my comments about the BSD-style licenses, but my experience around projects like Apache and PostgreSQL suggest that economic rather than legal means are generally used to achieve the same ends.)

  güzel sözler [04.14.08 07:49 PM]

Thank you Tim O'Reilly, great post..

  Thomas Lord [04.14.08 11:08 PM]

Hey, Tim.

You should revisit this post. You should do a retrospective on it.

Last chance, is my guess.

Some points to consider:

You speak of mutual "regret" but you and Moglen seem to use that word in different senses.

As a rule of thumb, never play speed chess with Moglen unless your objective is to be humbled.

Return the dialog to freedom. In the words of the Roches: well, read for yourself.

In the words of the Simpson's (Moe speaking, from the bomb shelter): "He's right, you know. About everything."

Chickens are coming home. Chips are falling. or as Watson put it: "Think."

I am, perhaps, overly kind. So much wasted time, indeed. Literally, lives destroyed thanks to self-promoters evading the topic of freedom. Rapes committed, literally. Secret police created, literally. In the name of "Open Source".

Stand up.

"All you need is..." -- Lennon-t

  Thomas Lord [04.14.08 11:36 PM]

Oh, and, I half take back my earlier statement. It's not yet a slam -- but it is personal.


  Tim O'Reilly [04.15.08 07:10 AM]

Tom --

You miss the point. I HAVE been talking about freedom, in every talk I've given about open source and the inevitable consequence of lock-in by data in Web 2.0. Go read my debate with Stallman in 1999 at the Wizards of OS. I've been pushing on the FSF to think about how the necessary freedoms will change in an era of cloud computing. They've been in denial, and in this case, decided to try to impeach me with the idea that since I was saying "open source" and not "free software," I wasn't talking about freedom.

In this "debate," the FSF once again failed to address the real issues, of how access to source code is no longer sufficient to ensure real software freedom.

  Thomas Lord [04.15.08 08:05 AM]

Tim: over beer will work better. I'm hearing both sides, I'm pretty sure. I do appreciate your sincerity and impulse towards concern. I do appreciate some of the process and discipline by which the abstract O'Reilly (your shop) and you yourself work issues and think through possible futures.

You say "denial" about the FSF and, really, the thing here is that you still seem to not be hearing what, for example, Eben was saying. I say "seem" here because, at some point, one has to lay aside the presumption that it's just a communication problem -- that "not hearing" or "misunderstanding" might not be the issue, after all.

Eben says something in the talk, interrupting you bruskly when you start to say along the lines of "That relates to my larger point...." He goes after you saying "I knew that whatever I said here that, somehow, it would 'demonstrate your larger point'" or words to that effect. That's why he came with a tactic. I see where he was coming from.

There's an urgency here that you seem to deny and that you are in a position to actually take up. "What we have here is a failure to communicate." 10 wasted years is about right. Enough money has been made at the margins of the disaster that is "Revolution OS" onwards -- back to business.

Now, as for web services and licensing: Eben answered that very clearly by talking about the fundamentals of freedom. Those tricks and those debates leveraged to create "Open Source" in its contemporary form are not the issue.

The freedom issues with "Web 2.0" have much more to do with the business models. From the core concepts of how "value" is supposedly created and extracted. Licensing isn't going to fix that cultural problem.

Freedom is a subtle practice but it's not that hard an idea. I'm not sure, at this late date, how to make it any easier for you.


  seks hikayeleri [05.13.08 06:52 PM]

This video proves that it is not only geeks that are too-frequently lacking social skills. Some lawyers for geeks apparently are socially inept as well.

  Anonymous [09.02.08 08:54 PM]

I have to agree with Tom and playing speedchess with Eben. On the other hand, I have to chuckle at the fact that I am not the only one getting a tongue-lashing by Eben :-)

There is no point arguing. They are out there, fighting for Freedom. And they don't get what you are saying, until a truck hits them that is.

Monetizing F/OSS was never a problem to them, then a company with 20+ billion in the bank started "buying" up Linux vendors outside of the { Red Hat, Ubuntu } set, and some of us noted, making money is important - to defend your organizations, or you will have to constantly re-spawn them (well, the ones which are not foundations, anyway). But they did not, it is free enough if the code is floating in the foam, on the ocean out there somewhere.

$ matters, because certain proprietary vendors have a lot of it - and "Freedom" is moving on their turf. But sorry, I forgot, the problem is the chardonnay-Socialist issue of Free vs Open software.

Post A Comment:

 (please be patient, comments may take awhile to post)

Type the characters you see in the picture above.