Question 3: Does widespread adoption and commercialization of open source software create new challenges or pressures for open source projects?
CEO
Hyperic
Soltero:Commercialization creates added pressures, especially for projects that are separate from the companies that provide commercial offerings around an otherwise free project. Frankly, the idea that commercial interests become involved in an OSS [open source software] project causes an allergic reaction to a lot of people. The reaction is mostly based on the idea that the commercial interests will overwhelm the decision-making process of the project. Realistically, without some amount of accountability, which comes best in the form of commercial interests, open source projects run the risk of becoming largely academic exercises that don't ship in time and have poor usability. How this accountability is applied into the project is the key factor in whether or not the commercialization will mean more success for the project or not.
Increased adoption of a project fuels the need for some level of governance and direction from the project. There's an interesting scenario in communities where tons of adoption bring way too many potentially conflicting interests, and without a proper governance structure, the project struggles because it cannot reconcile and prioritize the diverse requirements being thrown at it.
Both of these cases really don't apply to companies like ours [Hyperic] where the IP [intellectual property], the project, and the company are managed by the same entity. The community is open and works just like any other OSS project. The company is better able to balance the needs to fund and promote the community with the needs to deliver value to its customers.
Matt Asay
Vice president of business development
Alfresco
Asay: It does, but I believe the very structure of open source mitigates against too much fallout from the success of open source. With all the M&A activity, for example, it's to be expected that some will jump into the market for a quick flip on their investment. But open source isn't something you can force. Community doesn't come easily, and turning adoption into paychecks also doesn't come easily. So I think we are seeing and will continue to see a bit of a gold-rush mentality in open source, but the exigencies of the open source business models will keep us from falling into the same rut that the Web 2.0 world has.
The thing I worry most about, however, is related to my prior point and involves attempts to shortcut open source. Many see it as a mere marketing gimmick. They provide a certain amount of open source code as a teaser to get someone to buy into the "real" version of their software. This diminishes the value of open source for customers and, in my experience, is the product of too little confidence in the open source model. I don't want customers to come to believe that open source is a new vendor-delivered parlor trick and lose interest.
Zack Urlocker
Vice president of products
MySQL
Urlocker: There are projects and there are companies. Commercial growth is not everyone's top priority. Apache is hugely popular even though no one makes money off it. But I think there's greater awareness that you can build a business with open source today. That wasn't clear five or 10 years ago. Companies like Red Hat, Sun, IBM, make hundreds of millions in revenue due to open source software.
But you need to be clear if what you're doing is commercial or just a project. And if it's commercial, you need a business model that delivers value to paying customers. In effect, there are two classes of users in open source, and both are markets to pay attention to. There are your nonpaying community users and paying corporate customers. And you need to serve the needs of both groups at the same time. If you are not commercial enough, you end up like Apache. If you are not community-oriented enough, you'll never get the adoption and scale that works. Adoption must come first before there's an opportunity to commercialize. It's not easy to do this, but if you do it right, it works out well for both audiences.
Mark Spencer
Founder and CTO
Digium
Spencer: First, it is important to recognize the opportunity from the commercialization of open source software. Originally, there were just plain-old open source projects. Then people realized that by pairing open source projects with nonprofit organizations -- such as the Apache Foundation -- certain benefits could be achieved through improved leadership, direction, and an entity that could act to represent the interests of the users of the technology. Digium is a company on the forefront of what I believe to be the next logical step. By pairing an open source project with a for-profit company, there exists an opportunity for the company to provide promotion and marketing, product qualification and formalized testing, documentation, industry certification, and many other benefits that are harder for nonprofits or, certainly, isolated projects on their own. At Digium, for example, the largest single group in engineering is the group that develops for open source Asterisk. Most nonprofits would dream to be able to provide more than a dozen paid developers working full time on open source. Even with this benefit, however, there are people who tend to think of a battle between open source and commercial interests, and there is a challenge in bringing those people around to see how a properly built company can act as a steward for an open source project in a way that benefits the community even more than a nonprofit can.
Further, there are challenges that exist because there are companies who simply use the open source code without actually contributing back to the community in any way. While this is sometimes legally permissible, it is of course detrimental to the project, especially when those kinds of companies compete with companies who are spending resources on improving the core technology. While the community close to the project clearly understands this distinction, it is generally lost on the consumer population at large, and I think educating the consumer on the importance of supporting "real" open source companies will be important to giving this model its greatest effectiveness.
Dave Rosenberg
CEO and co-founder
Mulesource
Rosenberg: Open source “projects” have a much greater possibility of turning into “products” now that enterprises have accepted open source as part of the core infrastructure, and because venture capital firms have been actively funding open source companies.
These do add pressure but also create more opportunity. And since there are now several successful open source business models to look toward, commercialization is becoming much easier.
Bruce Perens
Creator of the Open Source Definition
Co-founder of the Open Source Initiative
Perens: A big problem facing many companies today is that they entirely depend on open source for their operations, and they haven't even begun to deal with that from a corporate policy perspective. I've met CEOs who haven't known they use open source at all, and then they have found out that all of their most critical projects depend on it.
When I wrote the rules for approving open source licenses, I didn't think that we'd get such a gold rush of companies that there would be 70 such licenses today. Dealing with the combinations of those 70 is too complicated. I direct my customers to three licenses that are compatible with each other and that provide for most of the business purposes of releasing open source. That's all you need.
I think our next steps might not be in software. Today, Wikipedia is one of the world's largest content providers, and it's open content. What else can we do like that?
Robert Sutor
Vice president of open source and standards
IBM
Sutor: I don't think there is a lot of common and good guidance out there for how a small open source project can deal with suddenly becoming very popular.
Similarly, it seems to take a long time for some open source projects to yield successful businesses based on them. I think we'll know open source has really arrived when every reputable business school spends a significant amount of time educating its students about the business models around open source. The new challenges and pressures will arise because of business issues, and not technology, in my opinion. We have many, many excellent developers in the open source communities. We need to have many, many more excellent "big picture" leaders emerging from and for those communities.
Sam Ramji
Senior director of platform technology
Microsoft
Ramji: There are at least two dimensions to this. One is simply growth of features and the resulting increase in the code base. If you look at the trajectory of the Linux kernel, as the amount of code increases, so does the complexity. With the added complexity come more defects. This isn’t anything new, nor is it a knock against open source software. All software evolves over time in this way. And it’s not a question of which model has more or fewer bugs. We have to make critical and strategic decisions about how to evolve the best models to maintain quality software.
A second dimension is that there’s a challenge to the community nature of some projects that have been developed in the commons. When these are commercialized, there is typically an uneven return of wealth from the commercializer to the original developers. This is currently causing some strains between the developers and the business people, but I think ultimately the industry will figure out some standards for fairness and generally follow them.
Chris DiBona
Open source programs manager
Google
DiBona: No more so now than in the late '90s. The nice thing about commercial interest in open source is that you get some really good code accepted into important open source projects. The bad thing is that they also need to reject the crappy code that people submit into the same. Not all projects are good at rejecting poor coders and their code.
Andy Astor
CEO
EnterpriseDB
Astor: The greatest challenge for open source projects is to stay smart and passionate. The earliest open source projects were by and large built relatively small groups of people who were excited about a particular idea. And they were built, quite frankly, independent of any monetary award. But as the open source movement grows, practitioners will have to be careful. Somebody once said to me, "As organizations grow bigger, they get dumber." And I think that applies to any new technology or trend in the marketplace. So, what open source projects need to do is to stay smart and retain their edge, and not appeal to the lowest common denominator.
Eric S. Raymond
Programmer, author, and
open source software advocate
Raymond: I don't think it creates any new problems; it just changes the scale a bit on issues we've been coping with (fairly successfully) for at least the last decade. Frankly, all the "will commercialization spoil open source?" worrying that the trade press is so fond of already struck me as old and boring five years ago. Next question?
[ Roundtable home | Topic No. 4: Avenues for acceptance ]