An Open Letter to Tim Bray about OSCON

Hi Tim,

I'm sorry I didn't get a chance to meet you at OSCON; I've been following your blog for a while and it's been a treat.

I'm writing to compare your experience at OSCON to my experience.

Interestingly, both you and I were added to the schedule at the last minute, but for very different reasons.

The submissions board did not think my talk was interesting or appropriate, largely it seems because it was about operations. Tim O'Reilly blogged about how cool Microsoft's operations are, though, paying particular attention to how important the topic is these days, and my (relatively bitchy) comment on his post led to Nat offering to let me fill in a slot opened up through a cancellation. I expect that my proposal was worded horribly, which likely had something to do with the rejection (I had no idea how to phrase the abstract or talk so it would be appropriate, because I had little knowledge of the attendees) but it was one of very few talks on operations and the only one on automation other than Capistrano, which is Rails-specific.

Your talk, on the other hand, seems to have been accepted at the last minute because you were able to email Nat directly and he clearly knows you, and because the topic is very Web 2.0. I expect that you could have gotten a slot on about any topic, but your proposal discussing Atom seems like it was a shoo-in given the people choosing slots. I might be getting the wrong impression about how you got your speaking slot, but your post implies it took only a few words from you to Nat.

This is relevant because I think it leads directly to our divergent perspectives on OSCON. I thought it was the most insular conference I've ever attended; every party was invitation-only, and invitations were not exactly trivial to come by. Well connected people got even better connected, and poorly connected people (such as myself; I'm well-connected in the systems world, but apparently not in the O'Reilly world) stayed poorly connected.

If part of OSCON were converted to a foo-style camp, then it would in fact be foo-style; that is, only friends of O'Reilly would be invited. This would just increase the insularity of the conference. There is a very good reason why people have created bar camps -- we live in an open culture, and it's not right that there be gatekeepers to the culture. Tim O'Reilly clearly has the right to do whatever he wants in his own camps and conferences, and he's done a great job of meeting and connecting a lot of great people, but it should be obvious that the open source world prefers open processes and open conferences.

And don't think that I was the only person who noticed or commented on the conference's insularity -- it came up quite often at the tables in the hall, which is where I spent a significant portion of the conference.

I, too, was disappointed with the content in the conference, but I do not think the solution is to let the luminaries speak and everyone else listen. I think it should take a bit more of a page from science -- peer reviewed talks would be a great way to judge both interest and quality. I can only assume that OSCON's acceptance board is as insular as the rest of the conference, and it likely had a large impact on the conference.

Really, though, I think the conference just lacks a purpose -- is it to enable better OSS software? Bring developers and users together for networking? Inform developers and users? Help build business around OSS? Or just to talk about anything and everything related to OSS?

I think not having a clear purpose resulted in a conference that didn't do its job very well. The greats got to talk and they did great, because that's why they're famous, after all. There were no good mechanisms for choosing the not-greats, so it was pretty random as to whether something good got chosen.

Why is it that a conference organized entirely around openness is so closed in its production and experience? I cannot find any detail about the organization of the conference beyond Nat and Tim. Compare that to USENIX's LISA organizer's page -- it doesn't show all of the paper reviewers, but this is everyone involved in decision-making, anyway.

I feel a bit like an ingrate, because I was provided an opportunity to present, after all, but the whole thing was a strange enough experience that I don't think I can skip the commentary.

Is it time to create an Open Source conference that's created in the same way that Open Source is created?

(I'm sorry this is so long, but I didn't have time to make it shorter.)