Four of us from Futurice attended FOSDEM 2015 in Brussels. FOSDEM stands for Free and Open Source Software Developers' European Meeting. It is a non-commercial, volunteer organized European event centered on free and open source software development.
During that weekend my goal was to, through discussions, come up with a future vision for the company Open Source program. Based on my notes (and memory) I transcribed some of the discussions (inaccurately) for this post. They may not be in exact chronological order.
There is a picture of my notes after each of the three segments. At the end of this post there are some conclusions and the next steps.
The other three musketeers were:
The FOSDEM 15 shirt sums it up well
Teemu: Hello! You have the Special Beer Promotion, Cuvée des Trolls? How big is that?
Waiter: It’s 1/3 liters.
Teemu: :-/ Aww… okay, well, yeah, I guess… I mean… yeah? Yeah. Let’s do that.
The waiter whirls away, but she soon returns.
Waiter: Pardon? The beer, Cuvée des Trolls, I checked, it’s only just 1/4 liters.
Teemu: Oh dog :(
Waiter: I thought you’d want to know :3 I can do Jupiler, half a liter?
Teemu: Aw Yiss! :D
My colleagues shuffle their feet and look away.
Teemu: Ahem, so, right, during this weekend my primary goal is to come up with a future vision for the company Open Source program...
After a long day in FOSDEM, and some rather disappointing Chinese food at restaurant Beijingya, (if you are the bearded fella who told us it’s the “absolutely best in Brussels”, either you have no appreciation of good Chinese food, or you are no gentleman, sir!), we are back at the hotel lounge bar, wrapping up the day with some tiny troll brews. The bar is about to close.
Kimmo: One potential issue with sponsoring Open Source projects financially could be that any such system quite likely results in project schedules and goals. This can be detrimental to motivation.
Ville: Some Open Source projects have done really well in Kickstarter, though?
Kimmo: That might be a bit different setup, more startup-like. ‘This is a really good idea, all we need is time to implement it – we need to get paid, so we can spend the time’.
Kimmo: An existing low burn project that attracts funding... Suddenly it might feel like you have to accomplish things fast.
Kimmo: Donate buttons and such have existed forever, but crowdfunding is a different beast. For one thing, if you manage to get the funding, you get a lot attention.
Kimmo: If I consider my own projects… Say, there’s a problem that I can’t seem to solve? I give up and might come back to it much later, suddenly it’s trivial. Now, when you feel forced to fix those problems right away, it’s gonna feel like a chore, stress you out.
Teemu: That would be a case of intrinsic versus extrinsic motivation I guess, many studies vouch that intrinsic is the shit, whereas extrinsic is just shit. That worried the hell out of me (and still does) with us already financially sponsoring employee Open Source contributions.
Kimmo: That has a different feel, as the program doesn’t pose any requirements on you… you can still take your sweet time mulling over the issues. You just make some extra when, and if you finally work on it.
Ville: Money might not attract new people to the sport, though?
Teemu: … which would suck, as that’s one of our primary goals ...
Ville: … or perhaps it does, the studies suggest otherwise, but I would be motivated!
Teemu: At least the compensation model has given us the active Open Source chat channel, where people see what projects others work on, and can showcase your own.
Ville: That is crazy important. For people to see and get interested in each others projects, it’s networking at its best.
Kimmo: It is a major motivator to get feedback, comments and requests for features.
Ville: … thumbs up in Flowdock ...
Kimmo: … yeah, very motivating. In my projects, the biggest progress leaps have been when someone contacts me and asks for some feature or a fix. I code day and night for three weeks, publish the changes, idle for half a year, someone contacts me with new requests and there we go again.
Teemu: Still, our plans on getting entirely new people on board Open Source are still vague at best. Our program is largely preaching to the choir.
Ville: Hackathons are good, but perhaps we should provide some clear tasks. Like the statistics visualization of the Open Source contributions done with our compensation. Should be easy to start developing ideas, trying out stuff, marking hours...
Teemu: Yes, and if we do that, we need to pay careful attention to what people suggest – actually accept, or implement, some better ideas, pretty much as they are. That is surprisingly rare, as people tend to prefer their own ideas. It is very common to ask for ideas and feedback, but still pretty rare for actually making it count. Having your ideas implemented (and duly credited) is quite refreshing.
Ville: … the goal being an active and friendly culture of collaboration!
Suddenly a huge man in a uniform sneaks up on us.
Huge man: Excuse me, you need to take your beers outside.
Teemu: Oh dog :(
My notes for the discussion about motivation
We are pushing past the O'Reilly desk, covered with books. It's very crowded here, and there is a steady stream of ridiculously big guys pushing in the opposite direction, slowing our progress considerably.
Teemu: How can they just keep getting bigger?!
Juho: You look small here, and you're not a small guy.
A guy with the physical resemblance of GoT’s The Mountain looms in front of us, blocking our way. He is wearing a dark t-shirt with a Docker logo and he is tenderly cradling a Docker Cookbook in his huge paws.
Teemu: Oh dog :(
We manage to escape with our lives and are walking outside, in the middle of the university complex, swerving this way and that, to dodge people queuing to buy waffles from the carts.
Teemu: I find the benefits for having an active Open Source scene rather obvious for society, companies and the individual. The question is then, how to achieve that. Since I recognize the risk in just throwing money at random projects.
Teemu: Everything I have heard so far suggests that for a working setup, there needs to be a foundation, which owns the IPR for the projects… A decentralized model where the financial support is, by some mechanism, distributed to the projects or individuals, does not seem viable.
Juho: Yes, establishing any kind of control would be difficult, and IPR matters would get complicated in a decentralized model.
Juho: Have you studied how they work? Like Linux Foundation, and the like…
Teemu: Not really yet, no. I suspect there’s plenty of information and examples available, though, as the problems with setting up foundations are unlikely to be that fresh.
Teemu: So the core model – a foundation that owns projects – what would those projects be? Where would they come from? I know companies have donated IPR to Apache Foundation…
Juho: The foundation can own IPR; it doesn’t need to own the projects. Just the contributions to the project, done through the foundation. As the programs and projects are modular. You can develop a new module, easier to get the ownership for that.
We are walking through one of the buildings in the ULB campus. There’s a faint sewer smell in the air. The building looks downright derelict.
Teemu: This campus, it’s really quite unattractive, isn’t it? Like… disgustingly so, maybe?
Juho: Makes you want to just get out of here.
The sewer smell intensifies tenfold.
Juho: Perhaps people graduate faster, so they don’t have to spend time here.
We grow accustomed to the smell and find a free table at the H-building of the campus. There’s a bar as well, but unfortunately with a line of about a hundred very determined people.
Juho: There may be challenges… If you look at the discussion opened in GitHub, about the register for Finnish patient medical records, it wasn’t very encouraging.
Kimmo: It was a good idea, but went off the rails quickly.
Juho: People ended up arguing about technology selections, platforms and such…
Juho: A successful open source project often have two very active roles. You have someone, who makes the challenging technical decisions, and someone, who communicates with the community. Usually one person doesn’t want to do both, nor would there be time for that.
Juho: The person who communicates gets people to participate, so that the project has a friendly feel to it. The other guy can then do the technical “this is shit” calls, and someone must, but if that’s your first impression of the project, it’s not going to work.
I scribble notes frantically, trying to ignore the brouhaha at the neighbouring table. I should rather admire people, who have the stamina to queue for the beer at FOSDEM, but to my shame that is not what I feel.
Juho: How would these people be selected, though?
Teemu: For the pilot, I don’t see a problem. In a long run… it’s an open question.
Juho: As to the project, perhaps it should implement some public service, with integrations to social media. So that people would actually find out about it. Kansalaisaloitepalvelu is one recent example.
Teemu: Yes! Solita built it for the Ministry of Justice and published the code. That is pretty cool.
Juho: Something that makes life easier for people. Identify a repetitive task that people have to do over and over again, then implement a better way to achieve that, based on the existing infrastructure.
Teemu: Looking at what’s been done in Estonia might give some ideas. I understand they have built digital public infrastructure quite efficiently. Not sure how much, if any, of it has been made available as Open Source though... Would be neat to contribute, if there is something that could work here in Finland as well.
Teemu: I suppose it should also be something that can be released to production pretty early....
Juho: That is indeed crucial, if you want to get others to contribute.
My notes for the discussion about methodology
Teemu: So, how can we test this? Setting up a foundation with funding, its own board, members… that’s a lot of work. A less grandiloquent scope is needed, to get some key assumptions validated, all lean and shit. But what are the assumptions? That Open Source projects can succeed? Hurr durr?
Ville: Perhaps we should look into an individual project, then?
Teemu: Yes. We could at least test the selection process – how to select a project. And the actual implementation phase – evaluate it afterwards, against some criteria to find out whether value was produced.
Juho: Whether there are users…
Teemu: Sounds good. Also the model where two persons are appointed responsibility – the technological authority and the community savvy.
Juho: Remember that the latter also needs to contribute to the coding, cannot be just a figurehead.
Our very casual waiter brings four Kwaks to our table. That beer comes in a special glass, with a holder, like so:
The hardest game in the world
Earlier, in another place, we had seen local looking elderly people drinking their Kwaks by removing the glasses from the holder.
Teemu: I have always had my Kwak without removing the glass from the holder… Like in this very establishment, the last year. This wooden holder has a handle. Why would it have a handle like that, if you’re supposed to remove the glass when you drink? Do you think I should remove the glass?
To this question (that one might consider polar), my three highly educated colleagues promptly produce three very varying answers.
Teemu: Excuse me, excuse me, miss! Now. Would I look like an idiot, holding this beer by the wooden holder while drinking, like this?
Waiter: Oui! You would :D
Teemu: Oh dog :(
My notes for the discussion about the pilot goals
When a company owns the IPR for an important Open Source project, there may be challenges in getting the community to stay happy. Core contributors are on constant look-out for signs of foul play and the backlash is severe, quite likely resulting in a new project fork and a mass exodus from the original project.
The way to mitigate this is to have the IPR owned by a foundation with reasonable rules. The foundation can then be funded by companies, for whom the developed technology is relevant. Linux Foundation and Apache Foundation are good examples. Now we also have a Node.js Foundation, founded recently, after the community started to lose faith in the company owning the IPR.
We would all benefit greatly of having a more active Open Source scene here in Finland.
For individuals; motivating work, learning possibilities, and global networking…
For companies; more available talent, employee motivation, and PR opportunities…
For the society; employment, attracting large company R&D, and better public sector systems...
What we need in Finland is a foundation with the sole goal of owning IPR for OSS projects, spawning new ones, and advancing their growth. The funding should be partially from the public funds, partially from the private sector.
The operation model can be validated to some extent with a pilot project, where a general interest service (or an improvement to such) is created as a funded Open Source project, accepting and encouraging community contributions.
Futurice could and should have an active role in making this happen, but conducting the pilot alone does not make sense. We need to discuss this with public administration and other companies.
We’ll discuss this with public administration and other companies.