ACT - Protecting Small Business Innovation

LA City Council Reads ACT's "Paying for Free" Paper

OK, maybe nobody on the city council actually read Paying for Free, the paper where we describe the security, privacy and sustainability costs of "free" software. Still, on Monday the Los Angeles city council delayed its vote on a contract to replace the city's email system. with Google Apps. The reason: concerns over costs.

Cloud-based software is often touted as an easy way to save money instead of using software installed locally on desktops and servers. Often, cloud-based software and services are even "free" to use (in the sense that they have $0 license fees) but will be supported by advertisements. Users often provide personal information or consent to targeted ads based on their interests to help online companies pay for these free services.

In Los Angeles, the city is rightly considering the costs as well as the benefits of the software that it chooses.  There are also cloud computing security concerns over storing sensitive city/law enforcement information on remote servers. These too are a potential cost that should be weighed into the mix of total costs of ownership. Free of licensing fees does not mean free to use.

A discussion about Apple's choice of licenses used in open-sourcing Grand Central Dispatch

Caution: because I am talking about code libraries, this discussion will be technical - a combination of the worst of geek and lawyer-speak.  Sorry.

I had a discussion with a colleague yesterday about Apple’s choice of licenses used for open-sourcing of its Grand Central Dispatch (GCD) concurrency framework, which is a piece of software Apple credits with driving great Mac OS X performance improvements.  GCD comprises both a library component, called “libdispatch”, licensed under Apache 2.0, and a C language extension, called “blocks”, a runtime distributed under the LLVM license, a BSD derivative (someone has already begun porting blocks to iPhone and older versions of OS X, distributed under the MIT license). 

Where I had speculated that Apple had chosen the license mix that it had to facilitate incorporation into GCC (and avoid the problems Apple experienced with developing Objective C as a series of libraries and a front-end that was a derivative of gpl-licensed GCC code), it looks like in reality Apple is trying to wean itself from GCC entirely, by its support for, and use of LLVM (a compiler runtime that is an alternative to GCC’s runtime) and a compiler front end called clang.  The LLVM runtime still supports the goal of portability because of its availability on linux (some distributions use a GCC-llvm compilation method for the kernel).  David Chisnall, in a thoughtful article, claims that Apple's transition to LLVM (away from GCC) presents a perfect example of the principle that "you don't necessarily get good results from forcing people to do things."

Apple’s support for LLVM may not be entirely due to the licensing issue, because LLVM is considered to have significant performance improvements, and a reason for Snow Leopard’s improved performance.  The performance improvement, however, can be due to Apple’s development support for LLVM, which may be due to the licensing issue.  So we may have a chicken and egg issue.  I do think that Apple started supporting LLVM before it got really good (and most likely in order to be free from GCC), and then the performance improvements followed (with help from a large open-source community that arose to support the project).   John Siracusa at ars technica says:  “By committing to a Clang/LLVM-powered future, Apple has finally taken complete control of its development platform.” 

So what do the respective licenses  and distribution methods get Apple?

GCD and libdispatch – Apache 2.0.  As we have discussed, one distinguishing feature of Apache 2.0 is that it provides for automatic license termination if the licensee instigates patent litigation regarding the work (in clause 3).  This provides Apple with control over its own patents regarding the work, and is similar to how Apple released Mini DisplayPort, a standard that has been incorporated into VESA's DisplayPort standard.   Apache 2.0 also allows for “viral” licensing, although it is not “copyleft” per se; in clause 4, redistribution allows either source and/or code redistribution and derivatives as long as the distribution is also licensed under Apache 2.0.  For that reason, the FSF considers Apache 2.0 licensed code to be “one-way” compatible with GPLv3, meaning that a GPLv3 project can include Apache 2.0 code, but not the other way around.  

The Apache 2.0  license is not compatible with GPL2, because of the automatic patent license termination, but that does not automatically mean that the linux kernel project cannot use the new libdispatch library.  Ryan Paul at ars technica believes that the license incompatibility does prevent adoption by linux.  Remember that the linux kernel today will allow linkage to fully proprietary driver libraries, let alone alternatively licensed open-sourced libraries.  See http://blogs.techrepublic.com.com/opensource/?p=276&tag=rbxccnbtr1 for a discussion of the debate about allowing linkage to proprietary drivers.  So Apple can get adoption or use of libdispatch by linux, if that is its goal. 

Apple is promoting use of GCD as a standard by also releasing OpenCL, a standard it has been developing with nVidia, incorporating a set of libraries to take OpenGL a step further and provide parallel processing on graphics cards and core processors.  OpenCL has been set up under the “stewardship” of the Khronos group that controls OpenGL, but John Siracusa says to "make no mistake, OpenCL is Apple's baby."  OpenCL builds on GCD, (and LLVM, but does not incorporate blocks) and provides a higher level programming API, just as OpenGL did, as well as language extensions.   Just as OpenGL has been implemented on linux, there is no reason to believe that OpenCL will not also be implemented on linux, by nVidia, if no one else.

Blocks – released under LLVM license (a BSD derivative) and as part of LLVM project.  It may have just made sense to release blocks as part of the existing LLVM project and under the license already adopted for that project.   This license is compatible with the GPL Because it is a BSD derivative, it can be used in GPL projects; the llvm-gcc compiler front end is released under GPLv2, (no update since GPLv3), but there is a GCC link time optimizer released under GPLv3. .

Because blocks are a c language extension, and not just a library feature, there is extra benefit to Apple by encouraging adoption (language extensions that developers do not adopt are useless).  Of course, by inclusion in Apple development environments, there will be some level of adoption, but there are network effects in having more developers use this construct.  Ars technica explains that “It's Apple intention to submit blocks as an official extension to one or more of the C-based languages, though it's not yet clear which standards bodies are receptive to the proposal,” (further noting that blocks are supported by all of Apple’s compilers (including Apple’s GCC distribution).

Apple has many very smart IP lawyers guiding its choice in open source licenses; the fact that neither Grand Central Dispatch nor Blocks were released under the GPL is not an accident.  Ultimately, I don't agree that Apple's choice of a non-GPL license will prevent its adoption by anyone other than GPL evangelists, and I believe that Apple would not have open-sourced this work if its only alternative was to use the GPL.  I think Apple's example shows how careful companies need to be in open sourcing code, especially if such code is part of a larger whole.

When You're Getting Software or Other Stuff on the 'Net for Free, what are the Costs?

Come one, come all. ACT will be hosting a lunch event next Tuesday (June 23) at noon on privacy, free software, and government procurement.  

We’ll discuss “free” software (ie. no license fees, free as in beer). It’s a nuanced take on some of what Chris Anderson will surely be talking about in his upcoming book on Free—where does the $ come from in software that we all use for free on the web, or that we download to our computer?

To answer this question, we’ll attempt to update traditional Total Cost of Ownership analysis for ad-based software and services. There’s a lot of discussion about privacy, security and sustainability considerations of cloud based solutions. In addition, the event will deal with skeptics who think that “free” means no business model at all. We’ll describe how free software and services are usually just one aspect of a larger enterprise geared toward expanding market penetration and increasing revenues. Mike Masnick described this in a recent Techdirt post.  

I'm going to moderate, and our speakers will be Rob Atkinson at ITIF, Tom Schatz at CAGW, and Peter Corbett of iStrategyLabs.  

We’ll be releasing a paper on all this, so come join us  for lunch and a lively discussion--and best of all, it’s FREE!!   

Further details are here.

TADC Conference - 'Other people's stuff should be Free, not mine'

Today at the Carnagie Institution the Trans Atlantic Consumer Dialogue held a conference called “Patents, Copyrights, and Knowledge Governance: The Next Four Years”. Here’s the top blurb about the event: 

As a new Administration will take office in Washington, and the European Union renews its institutions, what should the political agenda be for intellectual property? 

The globalisation of the challenges faced by consumers and rights holders have made intellectual property policy one of the main features of global trade policy, and stimulated both international and domestic debates about how best to promote innovation and access to knowledge, including "knowledge embedded" goods such as medicine, software, agriculture, inventions that address climate change, scholarly research, databases, films or recorded music. 

Both the United States and the European Union are facing demands to modify policies on patents, copyrights and other forms of intellectual property protection, coming from different perspectives. There are high profile right-owner lobbying efforts directed at higher standards and tougher enforcement of intellectual property rights, and growing interest among consumer groups, academics and many innovative businesses to protect the public domain and retain or even expand user rights. There is also much interest in exploring newer approaches to the support of creative and inventive communities, that do not rely on notions of exclusive rights. 

With the organisation of this event, the TransAtlantic Consumer Dialogue calls for two days of discussion on the assessment and on the prospective of the American and European political and policy Agenda on intellectual property practices and policies. 

The whole event is set up with a very strong anti-IP bent, and I am used to hearing that at NGO sponsored events.  But I was dismayed by the comments of Guilherme de Aguiar Patriota, the UN representative for the Government of Brazil.  Guilhereme gave a presentation very similar to one he gave back in 2002 where he said: 

“The intellectual property rights (IPRs) system, enhanced by the WTO’s Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), is actually hampering development of research and development (R&D) in developing countries; and industrialized countries, while heavily subsidizing science and technology (S&T) and R&D in their own countries, are placing all sorts of restrictions on attempts of developing countries to cooperate in market-oriented and commercially charged sectors of applied sciences“.

 But at today’s event he went so far as to suggest that copyright holders that expect Brazil’s schools to pay for books or software are actually hurting the children of Brazil, because it is taking money away from the kids. 

wow. 

I guess I just don’t know how to think about that in a rational way. Part of me ”gets“ what Guilherme is saying, but part of me looks at Brazil’s economic standing in a very different way than I might look at Kenya. Moreover, the fact that Brazil has the second most advanced industrial sector in the Americas means that Brazil might have to shoulder their part of the burden for education, including paying full freight for the tools developed outside their borders (if the tools are useful to them). 

But really, this issue of who pays and how much is kind of window dressing for the exchange he and I had on the fundamental concept of IP-Value. One little known reality about Brazil is that they are a leader in deep sea oil extraction and green energy. In fact, the State owned company ”Petrobras” is a major player in the field of new energy tech. But don’t trust me, here’s what Petrobras says about their own research and development laboratories: 

The technologies that are developed at the Cenpes turn Petrobras into the company that generates the most patents in Brazil and abroad. The number of patent request submissions in 2007 show the Cenpes is one of the world’s biggest applied research centers. In 2007 alone, in Brazil, 22 patents were granted and requests submitted for 59 more. Abroad, 129 requests were submitted and 29 granted. 

Given Petrobras’ success, and the vital import of green technology to the entire globe, I asked Gilhereme if, given the spirit of Access to Knowledge, Brazil should freely give up all patent rights and royalties from their energy tech. 

He did not seem to think it was a good idea. 

However, there was an interesting thread in his answer. He pointed out, with justified pride, that Petrobras was a company built in spite of all of the naysayers in the developed world who said it couldn’t be done. Gilhereme went on to point out ‘that Petrobras had poured millions into R&D to develop the technology and build it all on their own’. (I think he missed the parallelism there). 

Brazil is not alone in their call for others to give up what they themselves wouldn’t, but they are in a lead position to recognize and rethink a position that ultimately harms domestic innovation within developing countries. I’d hope that events like the one today at Carnegie would provide a more diverse and nuanced worldview that respects the fact that there are no simple answers when it comes to judging the value of the products created by others.

Awoogah! Anti-Open Source FUD Alert!

The Evil Empire is back at it again.  It's telling lies and half-truths about Open Source and Linux in an effort to keep the Open Source movement down!  From a series of recent blog posts, it appears that the dreaded "Get The Facts" campaign is back to tearing down the concept that Open Source is inherently cheaper than proprietary software.  Just look at these recent blog quotes:

"If you hear a CIO say he is going to reduce his costs by not having to pay for software licenses any more, then he has the wrong idea." - Filthy Microsoft Shill #1

Another blogger responded to a new Gartner report that found "through 2013, 50 percent of mainstream IT projects using open-source software (OSS) will not achieve cost savings over closed-source alternatives," and had the gall to say:

"I'm convinced that [the author] will be proved correct in spirit (maybe not 50 percent of mainstream projects!). I say this because I've seen his prediction play out already in the past year and a half...The current economic environment will result in companies considering the use of open source for a variety of projects in 2009. They'll do so because of the lower cost value proposition that most open source vendors seem to be including in their current marketing. But what happens when some open source projects don't meet the TCO test?" - Filthy Microsoft Shill #2

Wait a tick.

I'm now being told those two bloggers are apparently NOT Microsoft Shill's and it is NOT part of the Get The Facts Campaign.  Sorry, my bad.  I had no idea anyone else would dare question Open Source's cost advantage over proprietary software.

Instead, these quotes are apparently from noted Open Source advocates Simon Phipps of Sun and Savio Rodrigues, Infoworld Blogger and IBM employee. 

Wow, 2009 really IS the year of CHANGE!  Maybe Obama can bring civility and intellectual honesty to the debates on the web too. 

Be Careful What You Wish For: How the Jacobsen v. Katzer Decision Could Hurt the Free Software Movement

Last week, the U.S. Court of Appeals for the Federal Circuit, in Jacobsen v. Katzer, issued a very important decision in a case of first impression relating to the enforcement of software licenses.   In particular, it was the first federal appellate court decision to clarify whether failure to follow obligations imposed by an open-source license results in copyright infringement or breach of contract.  The distinction is important, as it controls the remedies available for such a failure, and can impact the question of whether lawsuits can be filed in state or federal court.  As Larry Lessig says, “trust me, this is huge.”  This may be true in ways that the free software movement did not foresee – and ways that the movement may very well regret. 

In this decision, called a “major victory for open source,” the court held that failure to abide by provisions requiring attribution in the Artistic License  (which wins the award for best software license name) will result in copyright infringement, because the license provisions were labeled and drafted as “conditions.” This decision is considered a victory  for both open source software licenses and the Creative Commons  licenses for literary and musical works, because it clears up questions about the enforceability of their licenses under copyright law. However, in strengthening copyright holders’ ability to make license provisions a condition of the copyright license, the free software movement may have won the battle, but lost the war.

As Braden Cox of ACT explained, because the “rationale of the decision is not limited to only open source,” this decision is a boon for all copyright owners who impose conditions on the license to their products. This could deprive the free software movement of one of its most important sources of new products: reverse engineering popular proprietary products.  Under Jacobsen, conditions, including a prohibition against reverse engineering, MAY now be enforceable under copyright.

This practice has already been under pressure in light of decisions such as Davidson & Assocs. v. Jung, 422 F.3d 630, 639 (8th Cir. 2005), which held that license provisions prohibiting reverse engineering (in that case, in relation to the Blizzard gaming system) are not preempted by copyright law.

While there are examples of important free software that did not originate by reverse engineering proprietary software (e.g., the Apache web server,  the Firefox browser ), many of the most important free software projects began by reverse engineering proprietary software systems.  For example, Richard Stallman and the GNU Project reverse engineered  Unix tools to create the “unix-like” GNU system which is a major component of what Stallman calls "the GNU/Linux operating system."  Stallman considered installation and use of Unix, a proprietary operating system, a “justifiable evil,” “applying the same reasoning that leads to the conclusion that violence in self defense is justified,” but he did not apparently apply the same rigorous analysis to his decision to copy Unix system tools rather than create his own system.

Another example of an important free software program that could not have been created without reverse engineering is the Samba project, which is a set of services that allow a non-Windows server “host to interact with a Microsoft Windows client or server as if it is a Windows file and print server.”   Samba was created by using the “french café” method of reverse engineering: having Microsoft clients and servers communicate with each other and using a “network sniffer” to discover the details of the network communications between them, akin to “sitting in a French Café and just listening to the conversations” in order to learn French.   

Yet another example of an important open source program that relied on reverse engineering is the Open Office suite of productivity applications, including a word processor and spreadsheet, which both reverse engineered the Microsoft Office data formats and “emulate[d] its look and feel where suitable.”   

The ability to reverse engineer software has been considered subject to the “fair use” defense under U.S. copyright law, 17 U.S.C. § 107,  including by the Ninth Circuit in Sega Enterprises, LTD v. Accolade, Inc., 977 F.2d 1510 (9th Cir. 1993).   Many software providers consequently have made it a condition of their license agreements that the licensee not engage in reverse engineering, as “private parties are free to contractually forego the limited ability to reverse engineer a software product under the exemptions of the Copyright Act.”  Bowers v. Baystate Techs., Inc., 320 F.3d 1317, 1325-26 (Fed. Cir. 2003), quoted in  Davidson, 422 F.3d at 639.  This is where the Jacobsen decision could impact reverse engineering practices. 

If a software license is “conditioned” on not reverse engineering the software, then after Jacobsen, breach of such a license “condition” could result in a finding of copyright infringement, rather than simply a breach of the license contract.  It’s not the reverse engineering per se that would constitute the copyright infringement, just as it’s not the lack of attribution that constituted copyright infringement in the Jacobsen case, but rather the unlicensed exercise of one of the exclusive rights granted under copyright law (i.e., the right to make copies or derivative works, which takes place during the process of reverse engineering).  And just like with the attribution provision in the Artistic License, as the Federal Circuit explained, “these types of license restrictions might well be rendered meaningless absent the ability to enforce through injunctive relief.” 

This is the result that Eben Moglen has long argued for, stating “[l]icenses are not contracts: the work's user is obliged to remain within the bounds of the license not because she voluntarily promised, but because she doesn't have any right to act at all except as the license permits.”   But if this result prevents development of new free software projects that rely on reverse engineering popular commercial programs, then it could deprive the free software movement of one of its most significant sources of new projects. 

Make no mistake – this decision strengthens the hand of copyright holders, a large number of which are the types of commercial organizations that insist on being paid for their hard work when their products are successful in the marketplace.  By pressing for a short-term victory, the free software movement may have sown the seeds of its own defeat.

This is not legal advice

FSF FUD Wrong Again: First FLOSS Program Already Available for iPhone

Thanks to Mike Sax for pointing out that the Chicken Little's over at the Free Software Foundation tried to grab some attention around the launch of the new iPhone by publishing 5 Reasons to Avoid the iPhone. Given that Apple sold more than a million of them in the first 3 days of availability, perhaps the FSF needed a whole lot more than 5 reasons to get people not to give in to their technolust.

More importantly, the FSF might have been more persuasive if its reasons were ACTUALLY TRUE.

The FSF's number one reason for people to avoid the iPhone is:

iPhone completely blocks free software. Developers must pay a tax to Apple, who becomes the sole authority over what can and can't be on everyone's phones.

But wait a tick...doesn't WordPress have an iPhone application that is open source?. Why yes! Yes it does!

If it isn't true, why would the FSF say it? Despite all their protestations, the FSF seems to have fallen in love the tactic of FUD.

Bust Open the Cristal, ACT Wins Award from Some Angry Brit with a Blog!

Finally catching up on my blog clippings after a week of travel, and I found out that ACT has apparently won the inaugural "Scum of the Day" award from some angry British guy with a blog!  I've ordered a case of Cristal for our DC and our Brussels offices in celebration. So try not to call us on Tuesday...the ringing will hurt my head.

Anyway, it seems that the prize is in honor of all our work to identify some serious problems with the GPLv3 during the drafting stages.  What is truly amazing about the accomplishment is the competition we beat out!  There were so many who did so much to highlight the dangers of the new license.  Other "libelous" critics of the GPLv3 read like the a Who's Who list of the World's biggest Linux haters!

• The Notorious DLC (Debian Linux Community) did some really ground breaking work FUDDING the GPLv3 on behalf of their corporate masters of Microsoft.  Wish that we too had noted that the GPLv3 fails the "Dissident Test" and could be used by repressive regimes as “a weapon to be used against people holding unpopular or illegal views.”
• The Linux-Hating Members of the Linux Kernel Development Team which made some crazy claims about the GPLv3's "potential to inflict massive collateral damage upon our entire ecosystem and jeopardise the very utility and survival of Open Source."  They also noted the same concerns WE had about the patent provision.  Obviously, these blatant lies were a favor to Bill Gates after he loaned them his Lake Washington home for a massive kegger and promised to open source Microsoft Bob.
• That Guy Who Hates Linux So Much He Put His Name On It.  I figured Linus Torvalds' refusal to move the Linux kernel to GPLv3 would have made him the frontrunner. But, after he called the Free Software Foundation and its acolytes a bunch of "controlling," "condescending," "hypocrites," that are full of "hot air," it seemed like a foregone conclusion...Linus did more than anyone else to raise concerns about the GPLv3 and the FSF'ers that were designing it. 

Yet, somehow little ol' us won the award... funny that.

If I were to hazard a guess why, perhaps it's because reminding the world that most of the open source community's leaders have criticized and/or rejected the GPLv3, would undermine Mr. Angry's entire version of reality:

"the oceans of FUD Microsoft has bought itself, including low-key payments to academics (never forget academic kickbacks), were at best able to slow down — but never to defeat — adoption of this evolutionary licence."

The reality is that Microsoft didn't need to do anything to FUD the GPLv3.  There were so many legitimate problems with the license that the Open Source community led the charge on the so-called "FUD" campaign. 

ACT engaged on GPLv3, because we believe that the development of the license was a turning point in the software industry.  After years of moves toward collaboration and mutual respect between open source and proprietary software companies leading to hybrid solutions, the GPLv3 was akin to the Bonfire of the Vanities.  It was an effort to purify the Open Source community back to its Free Software roots, a move that was destined to split the community... and, as you can see above, we were right.  In the end, the Free Software Foundation even made some of the changes that we pushed for including rewriting the DRM section and parts of the patent provisions.

In the end, though, we stand ready to accept our award with open arms.  Mr. Angry, just tell us where the awards ceremony will be and what the dress code is and we'll bring our own red carpet.  Just one question...does the award resemble a tin foil hat or is it more like a Faraday cage?

SFLC steals a page from the Business Software Alliance - good thing there is no patent on strike lawsuits!

Those of us who thought maybe Free Software was all about love and peace and breaking with the "tyranny" of copyright and other intellectual property might have been surprised by four recent heavily publicized lawsuits filed by the Software Freedom Law Center (SFLC) on behalf of BusyBox, a lightweight set of unix utilities licensed under version 2 of the Gnu Public License (GPL). 

These suits, filed in rapid-fire succession in September, November, and December of 2007 against Monsoon Multimedia, Xterasys, High-Gain Antennas, and Verizon, all involved the same claim that these companies (perhaps inadvertently) distributed the BusyBox program with their products without the source code as required by GPLv2.  The lawsuits all seem to have settled the same way as well: by promising to release the code, appointing an "open-source compliance officer", and paying an undisclosed sum of money to the plaintiffs.  (See links for the settlement details for Monsoon Multimedia, Xterasys, High-Gain Antennas, and Verizon - the Verizon case is unique in that the settlement mostly penalized the third-party company that supplied the product containing the offending BusyBox code).

These lawsuits in the U.S. are mirrored by actions in Europe, most recently the lawsuit brought against Skype, which distributed a flyer with its product listing the internet address where source code could be downloaded by interested consumers, but failed to distribute the text of the gpl with the product.

These companies were all punished for adopting open-source software in their products, but doing so imperfectly.  If these actions seem familiar, you might be thinking of the Business Software Alliance (BSA), which has been criticized for what some call intimidation tactics to "punish businesses that may be trying to play by the rules".  And GPL lawsuits are only going to increase: the people behind the SFLC just formed a for-profit law firm so that open-source businesses can sue for violations as well.  The "free software" folks must be thinking: "why should proprietary software have all the fun?"  It is high irony that the RedHat blog highlighted actions by the SFLC and the BSA within a few days of each other.

What should you do if you receive the dreaded call or letter from the SFLC?  Well, it might pay to take the advice given to those who have received similar [edited] audit letters from the BSA.  Some good advice from Baseline magazine:

1. Hire a lawyer.
2. Cooperate -- carefully - "Remember the sole propose of the BSA [and the SFLC] is to get as much money from your business as possible."
3. Don't rush out and buy or remove any software - it won't make any difference at this point.
4. Preserve confidentiality of evidence by having a lawyer involved in the internal collection of information, which imbues that effort with privilege.
5. Find your allies.  Baseline suggests involving your software vendors.  Verizon successfully demonstrated that it pays to make sure that the people who helped you get into this mess help get you out of it by involving the vendor who supplied the products containing BusyBox.
6. Create a compliance plan.  You will have to have one as part of any settlement with either the BSA or the SFLC.
7. Negotiate non-monetary aspects.  While the BSA might agree not to publicize any settlement, that doesn't seem to part of the SFLC's new game plan, which has involved heavily publicizing its settlements with press releases, etc. 

Litigation is expensive.  Both the SFLC and the BSA are well aware of the incredible costs that targeted companies would incur just in discovery, let alone fighting the full case in court.  Be careful out there.

While I am a lawyer, I am not your lawyer.  This is not intended to be legal advice.  I sincerely hope that you do not come within the cross-hairs of either the SFLC or the BSA [edited].

Innovation Takes a Bazaar and a Cathedral

Over at the Open Source Blog, IBM's Savio Rodriguez discusses a presentation by Spring Source's Rod Johnson at EclipseCON. 

Rod and Savio come to the same conclusion that we at ACT has been discussing for years: Innovation takes a Bazaar and a Cathedral. 

As Savio writes:

Rod challenged the notion that the Bazaar model is the best or only way forward. Rod claimed (near quotes):

"The bazaar model encourages competition in implementation, but may not produce innovation.
....

The cathedral model is more likely to produce innovation. Remember that Eclipse began as a cathedral project from IBM.

....

Now, the combination of the bazaar model and the cathedral model drive innovation to a much higher degree than either would alone."

It is a simple matter of economic incentives and the incentives in the proprietary model enable companies to invest more in revolutionary innovation than the incentives in the Bazaar model.  Of course that doesn't mean that revolutionary software can only come from proprietary models or that all proprietary software is revolutionary.  From a macroeconomic perspective, however, the proprietary model is more likely to get companies like IBM to invest in big innovations. As Rod concludes:  

the combined use of the OSS model and the proprietary software model is the future. I've seen (and lived through) proof that proprietary vendors are learning from OSS vendors. Are OSS vendors doing the same with lessons from proprietary vendors? Or have OSS proponents led OSS vendors to believe that there is nothing to be learned from "the past". I hope not.