OSCON, Pt. 2.2: Participate 08 (Sponsored By ... Microsoft?)

Let's rewind a bit. My Monday afternoon at OSCON 2008 was taken up by "Participate 08," a Microsoft-sponsored discussion panel chaired by a whole panoply of folks -- including, yes, an open source liaison from Microsoft. The whole thing was neither a "corporate apologia" (as one wag put it from the audience) nor a pile-on where Microsoft got the worst of it. Their approach was only one of a diversity of perspectives, and sometimes not even the most eyebrow-raising.

Let's rewind a bit. My Monday afternoon at OSCON 2008 was taken up by "Participate 08," a Microsoft-sponsored discussion panel chaired by a whole panoply of folks -- including, yes, an open source liaison from Microsoft. The whole thing was neither a "corporate apologia" (as one wag put it from the audience) nor a pile-on where Microsoft got the worst of it. Their approach was only one of a diversity of perspectives, and sometimes not even the most eyebrow-raising.

The participants:

  • Karim Lakhani (moderator), of Harvard Business School
  • Allison Randal, O'Reilly Radar
  • John Wilbanks, Science Commons
  • Siobhan O'Mahony, UC Davis
  • Bryan Kirschner, Microsoft
  • Zack Urlocker, MySQL

Because of the sheer amount of material covered and the pace of the discussion, I settled for chomping out some of the most illuminating tidbits:

What surprises you most about open source? Siobhan, whose background is in sociology, noted that in her field of study there's a long history of failure for collectives. "It's remarkable to see how large projects [e.g., Eclipse] are being well-managed" -- especially when there's initially not much order to be discerned. (The order does exist, Allison maintained; it's just something you have to look a little harder for.)

Zack was most surprised by how late Microsoft came to the open source table; in his purview, they're missing out on a whole generation of coders who take open source for granted. Bryan's stance was that by coming late, they could "stand on the shoulder of giants" and not repeat other people's mistakes. (There was also some note here to the effect that this new generation of open coders is not using Microsoft tools or languages, but I found that a debatable assertion at best.)

What do you think of hybrid business models? For Zack, MySQL was a fine implementation of a hybrid model: They have both paying customers and a highly competitive base of users who don't pay them but nevertheless have serious influence on the rest of their audience. For him, there had to be some awareness of how openness could be measured as a measure of success as opposed to how much money you made.

Allison's take was that hybrid models work best in the corporate world, but there are plenty of business models where the code is not the moneymaker. (Classic example: Some Apache founders decamped and founded Joost.) For Siobhan, there were two different kinds of "open": 1) transparency, 2) accessibility. Corporate-sponsored open source projects tend to go No. 1 (what she called the "fishbowl" model) for the sake of control.

John felt that hybrid open content models are not nearly as developed as models for software, and are probably even more complex. Creative Commons-type licenses that require more than just simple attribution have a chilling effect in integration and federation, which is bad for science that has commercial applications (which, when you get down to it, could be anything). Putting data in the public domain is the only strategy that gives you certainty, but in a lot of developing countries that's simply not possible. He noted that there was a certain irony in the warning that "You don't know what's going to happen if you use public domain" when delivered by people in the open community -- it's an ironic echo of Microsoft's own finger-waggling about open source.

One commenter called from the floor that companies "are misusing open source the way they're abusing the term 'green': How much in the way of open source tendencies do you need to be a real hybrid? How much commercialization is too much?"

Allison: "You have to be open source first, somewhere."

Siobhan: "That's an issue of the license you choose, as opposed to the development model or the business mode."

Bryan: "We have programs we're sharing source codes for a long time under academic/government licenses -- our Shared Source Program -- and we don't call it open source out of respect for the full definition of open source."

Someone else commented from the floor that Microsoft might well be an "assassin" if admitted into open source circles.

Allison: "I was on a panel was Sam Ramji, and we talked about how when you get a new contributor into the community, they often do dumb things at first. You do have to kick them a lot, but at a certain point you have to stop kicking them and start training them. The more you talk about something with others [in this case, the more MS engages others about open source], the more your own perspective shifts. There is dialogue that wasn't there before."

Karim: "Sure, you can say there's no trust. Or you can simply be prudent and see what happens. If the words don't match the action, now we have someone [Bryan] to shout at."

Bryan: "Judge us on our actions going forward, as Sam said." (An aside: the Windows installer toolkit was the first OSS project at Microsoft, and it was a huge exception at the time [in 2004]. Today there are 400-plus such projects, along with whole product groups building it into their strategy.

On the subject of inclusion:

Karim: "You have to have people come and join your community, to be sure, but communities are inclusive and exclusive, both. Comments?"

Zack: "You don't have to just write code to be part of the community - you could be a user, a debugger, just someone in the loop."

Siobhan: "Newer communities seem to be easier to get into, since there's not as much hierarchy to traverse. That said, people who do more boundary-spanning work advance higher in any given hierarchy."

Allison: "It's not about putting up a signboard and waiting for people to gather; you need to give them a bridge from where they are to where you are."

Zack: "Sometimes exclusion is a healthy thing. Sometimes you need to keep out people who would otherwise be a detriment."

Karim: "Connectivity is now the default." (This was an observation I clued in on pretty strongly: you have to assume that people are watching and listening, and that some of them also want to get in touch and say something.)

Allison: "The perception is that there is no order [in many open source communities], but there is a system of governance that has to be found."

Siobhan: "Sometimes you have to design the community to fit the audience in question."

On the subject of motivation in open source:

Zack: "Large corporations will contribute to open source to solve a problem they have. They make it public to allow other people to make sure it's the right fix. And it shows they're someone to participate with for future efforts" (what I've called "being a good open source diplomat").

Siobhan: "Commitment to a given firm that's doing open source is much shorter than commitment to the community as a whole. If a programmer is with a company and they feel the company isn't a good open source player, then they'll go somewhere they feel is more in line with those feelings."

Allison: "Many OSS developers have a genuine humanitarian interest in helping others. They get tremendous intellectual satisfaction from this work, and since they would do it anyway ... "

On the subject of whether intellectual property was irrelevant to software:

Zack: "Er, no?" [laughs] "Open source does not automatically imply public domain, after all." (Not by default, of course -- there may be open source projects licensed as public domain works, but that amounts to only a fraction of the total.)

Bryan: "Intellectual property is a set of tools in a toolbox. Just as there's diversity of motivations in open source, there should be diversity of tools."

Allison: "IP is both our bane and our lifeblood. At the moment it serves us well, even if there could be changes." (The GPL itself couldn't exist in its current form without copyright.)

John: "Whatever you think about copyright, it's here; but IP needs to be 'unbundled' and thought of more granularly. Is the right to give credit the same as the right to file a lawsuit for failing to give credit? I'd hope more that we use trademark and branding (following the norms of a community and earning their branding) instead of lawsuits, especially when you are entering an age when you have software that writes other software."

How do you approach criticisms about one's approach to IP? I asked.

Zack: "It helps to have a thick skin and to operate transparently. We have both our paying customers and just plain users which we try to satisfy, but sometimes you have to make a decision and be transparent about the reason behind it."

John: "Creative Commons has retired a couple of licenses, such as the Developing Countries license, because they devalued the rest of the core licensing. So we do have to listen to the community and value their feedback."

Someone from the floor asked: "If you have attributions of attributions, at what point do you give up attributing the original source?"

John: "I'd say speak with an attorney. We call that 'attribution stacking,' and it's going to get worse as machines write software. You don't always know when you've triggered such a thing."

Allison: "I'm not convinced attribution is the biggest thing, since many licenses don't even require that specifically. The biggest thing is not attention to you, but innovation delivered to you thanks to open source in the first place."

What gave the panel an extra little cachet of niftyness was the live "mind-mapping" of the conversation that took place on whiteboards behind the speakers:

Editor's Choice
Brian T. Horowitz, Contributing Reporter
Samuel Greengard, Contributing Reporter
Nathan Eddy, Freelance Writer
Brandon Taylor, Digital Editorial Program Manager
Jessica Davis, Senior Editor
Cynthia Harvey, Freelance Journalist, InformationWeek
Sara Peters, Editor-in-Chief, InformationWeek / Network Computing