Why I’m not an open source person any more

I’ve been having this conversation a bit lately so I just wanted to put it out there.

From 1998-2007 I worked full time in open source software. I considered myself a member of the open source community. Open source was kind of my “thing”.

This is no longer true.

I still use open source software extensively (I’m writing this in WordPress, using Mozilla on Gnome on Ubuntu), but then, so does everyone, whether they know it or not. Sometime around the early 2000s, Linux and other open source software stopped being a fringe, weirdo thing and started just being a sensible choice for most Internet projects. And since almost everything’s on the Internet these days, well, open source is just something that is.

To put it another way: if the open source movement were a software project, I’d say that software project is in maintenance mode. It’s out there, it has widespread adoption, and while there’s still work to be done, it’s more the ongoing work of keeping things going than the initial big push to get it launched. And I’m not much good at maintenance projects.

So what am I doing these days? When people ask me I usually say, “Open… stuff.” And then I wave my hands a bit. In my day job with Freebase I mostly work with open data. But I’m also interested in those sort of open principles as they’re applied to other aspects of our lives.

A short list of things I consider to fall under the umbrella of “open stuff”:

  • Intellectual property reform and alternatives to the current copyright system (eg. Creative Commons, anti-DMCA efforts, etc.)
  • Increased access to knowledge, information, and art (Wikipedia, open access journals, Scarleteen)
  • Decentralised social networking platforms (StatusNet, Diaspora)
  • Radical online collaboration and novel ways for groups to work together online (Wikipedia, of course, but also Anonymous, which I think is fascinating and important even if I mostly disagree with them)
  • Using technology to connect and empower members of marginalised groups (Genderplayful Marketplace, disability hacking)
  • Using the Internet for social change and grassroots political activism (too many to list, but #jan25 seems timely)
  • Non-traditional, non-hierarchical ways of working on projects (Agile, consensus-driven, anarchic)
  • Grass-roots, community-run, egalitarian events (unconferences and the like)
  • Unofficial/unlicensed fan activities, especially creative/critical/transformative fanworks and the communities around them (Organization for Transformative Works, vidding, scanlation)
  • Small-business and micro-entrepreneurial activities on the Internet, especially as they enable independent artists/writers/musicians/creators (Etsy, Kickstarter, Bandcamp)

There’s more, of course, but all those are things that excite me. It feels like there’s something broader there — not just software, but a whole cluster of Internet-related things that are about giving people more options, more ways to express themselves, more ways to make a difference, more ways to (at the risk of sounding a bit woo-woo) realise their potential. Ideally while not being beholden to, or at the risk of being shut down by, any one corporation or government or institution.

Of course open source software is a part of this, but I don’t think it’s the only part, and it’s definitely not the leading or most important part for me any more. So, if you invite me to speak or write or come to an open source event or whatever, and I say “I don’t really work in open source any more,” this is what I’m talking about. Hope that makes sense.

(That said, if you read this and you’d still like me to speak/write/attend your open source thing and talk about “open stuff” in a more general sense, let me know.)

Dolores Park mural

Random pic is random: Dolores Park mural, at the corner of 18th and Guerrero, San Francisco. In the style of Georges Seurat's A Sunday on La Grande Jatte.


13 thoughts on “Why I’m not an open source person any more

  • Skud Post author

    Yeah, I like a lot of what Tim says. Though actually I don’t think he goes far enough. He has an idea of what he’s thinking about when he says “alpha geek” and I think it predisposes him to certain stuff and makes him miss other things that are going on. I am *completely fascinated* by some of the stuff that’s coming out of fandom recently, but it seems to be flying entirely under the O’Reilly radar (pun absolutely intended). For instance, there was an amazing discussion last week around ebook piracy, mostly taking part on Livejournal and among women in the intersection between the fandom and social justice communities, that resulted in some really crunchy, fascinating discussion and some minds being changed… and I saw very little awareness of it outside those circles. You’d think with O’Reilly running the Tools of Change/Future of Publishing conference, they’d be all over that, but apparently not.

    Dammit, I should probably make a post about that. I’ve been saying so for a week now. In the meantime, posts to read:

    http://kanata.dreamwidth.org/1393254.html (general link roundup)
    http://troisroyaumes.dreamwidth.org/38222.html (pull quotes from the best posts)
    http://karenhealey.livejournal.com/918268.html (published author changes her mind and apologises)
    http://jimhines.livejournal.com/551156.html (another published author changes his mind)

  • Daniel O'Connor

    Hey Skud, where do you see open data heading?

    I agree a bit with the feeling that open source has arrived; and in a way freebase has come along and done that for much of the semantic web ideas – it’s now a tool to go out there and build something…

    … but from where I am, not much is being built on (freebase specifically). IE: there’s a lot of data, but it’s slightly incomplete or hard to combine in a useful way. Alternatively someone else has come along an done the practical step using another method (ie: XBMC eating TV information with scrapers from tvdb or whatnot; rather than just pointing at freebase or similar).

    Do you see the open data you work with having an impact on the other problems you listed? If so, how; and can I play too?

  • Skud Post author

    Oh boy, that’s a very work-y question for a Friday night ;) Consider the following to have extra “not speaking for my employer” sprinkled all over it.

    When I started working on Freebase in 2007, there was almost nothing out there in the “open data” space. You said “open data” and people looked at you a little blankly. Now there are governments worldwide releasing open data, there are regular conferences about the subject, and a lot of the stuff that was completely arcane just 3 years ago (like, say, talking about strong identifiers and reconciliation for identities) is becoming commonplace… at least in the circles I move in.

    Last year at OSCON I gave a talk called “Open source, open data” where I talked about the similarities between the two fields, and how I see open data facing the same challenges now as open source was 10 years ago. People are starting to know what it is and want to know how to use it for their particular applications and stuff, but we are missing stuff like toolchains, quality control mechanisms, or even a strong understanding of what an “open data license” is. We’re working on them, we have a lot of ideas, there are a lot of experiments underway, but there’s not yet any such thing as (say) a github for linked open data, or even a good equivalent of diff and patch.

    So yeah, it’ll take a while. Were you using Linux in 2000? How often did you have to recompile your kernel to get USB working? Or sound? What toolkits were you using to build websites (I was using CGI.pm)? What did you use for version control, and how did it handle branching (CVS, and ugh)? That stuff was a complete PITA, and now it’s a no-brainer. So I think with open data, we’re where we were with open source back then. The adoption looks to me like we’re going over the crest of the hill right about now, and it’s all going to accelerate and become easier from here on.

    As for the impact on the other problems — I think we’re already seeing some effect, and it will continue to grow. Open source, open data, open standards, interoperability, platform choice and competition, freedom of information, all these tie in together in my mind and support each other and make something big and important. How you can play too… I’d say find a particular problem you want to solve, and dive in deep in that vertical. Pick something local and important to you and to people around you. One of the things we need most right now in open data land (IMHO) is real use cases. There’s a lot of people making broad, theoretical generalisations about “what we need to do”, but it’s hard to know what’s true til we see more people really getting their hands dirty.

  • Daniel O'Connor

    Heh, sorry about the work focus!

    I’ll take your “not speaking for my employer” and raise you a “this is just my personal opinion”.

    I definitely think you are right about the use cases being required. My workplace was acquired by RP Data; who have built their entire business around identifying use cases and selling (http://www.rpdata.com/about_us/our_data.html).

    Unfortunately the idea of taking any of the data they have and making it open sounds like it threatens the business model – at least on the face of it.

    I just came from reading an article that made me want to help solve problems with data –
    http://www.newyorker.com/reporting/2011/01/24/110124fa_fact_gawande?currentPage=all

    Unfortunately, I’m struggling to describe exactly what would actually be solved by providing open information on houses.
    Most often any idea that is actually useful would be exactly what you keep non open.
    The alternative then becomes identifying data which appears “useless”, but useful to other parties.

    I guess what I’m looking for is a cross between lazyweb and the concept of an open problem list – the link on http://data.australia.gov.au/ which says “request some data” with a box for “How would you use it?”.

  • Skud Post author

    So when I think about existing closed data sets, and whether we could get institutions to open them up, and whether we’d want to, I think about how that worked with open source. There have been dozens of cases of closed projects being opened up and released under an open source license, and when that happened, it turned out that the code was horrible, there was no support from the organisation (who often just wanted it off their plate), no support infrastructure for a developer community, etc. The rare cases where a closed project went open and succeeded — Mozilla springs to mind — too a REALLY long time to get there. Remember jwz’s nomo zilla post?

    The places where open source has really won, were areas that we hardly knew *existed* back when we were starting out. Web browsers. Web servers. Instant messaging. Bittorrent. Blogging.

    So I don’t think we should necessarily be looking at areas where there’s heaps of closed data and thinking “how can we open this up?” That stuff’s too often encumbered by legal complications or legacy business models that depend on it remaining closed. Instead I think we should be looking at new projects and new fields as they emerge, and realising that open data is one of the things in our toolkit that might help us succeed. Then we make the data open from the start.

    (You work at RPData; I used to work at realestate.com.au. I’ll tell you now that the above is bloody difficult to do in Australia, because Australia tends to lag on tech innovation. But if it were possible, I would love to see a real estate website in .au that crowdsourced data from homebuyers/sellers and made it available in aggregate/anonymised form. The trick of course is reaching the crowds, rather than going via the agents, because existing businesses kind of have a stranglehold there. I’m thinking more like Everyblock meets Mint.com but for homebuyers/sellers. Not that I want to do it myself (ugh), I’m just throwing ideas out here.)

  • Cesy

    Thanks for the links about the ebook discussion – I’d heard about that, but hadn’t had the energy to wade through the whole thing.

  • David Gerard

    As I have noted, the next big battle for Wikimedia is to get everyone else using free content licenses.

    e.g. Al Jazeera have released a pile of their Egypt coverage as CC by-nc-nd, but we have asked nicely and they have released some under proper free content licenses. Other news agencies have done the same. I am currently trying to collect a list and get it made into a WMF blog post. The point being to make releasing stuff under a proper free content licence normal.

    But we’ve about half won already.

  • Ari Saks

    Thanks for the blog post.

    I’m interested in open source software and data but from the perspective of non-profit work. I’d love to integrate that kind of technology with data on what drives people to do good works, not necessarily data connected to ways to find the closest taxi. Do you know of anyone or any site that is interested in the nexus between open source and non-profit work? Thanks.

    ari

  • David Gerard

    @Ari – that would be Wikimedia. If you’re not interested in working with them as such, there’s plenty of research by Wikimedia and people around it trying to understand how Wikipedia can actually work, which would be useful for thinking on the topic.

Comments are closed.