Read time 6 min
From the very beginning, Reaktor has built most of our software using open-source technology. Experience has shown that working with open-source software (OSS) comes with many benefits: no vendor-lock, no license fees, the support of a community, and the option to contribute fixes when problems arise are just some of the many advantages.
We have also attempted to use our projects to contribute to open source whenever possible. Examples include our work at the Finnish Board of Education (e.g. Koski), open-source libraries at the Finnish Broadcasting Company (YLE), and training on open-source strategy at several clients. We often advise our clients to leverage the potential of open source, not only as a consumer but also as a producer.
Making an impact
It’s worth noticing that in addition to our work as a company, Reaktorians have also made an impact in their own right and on their own time. Some well-known examples, with hundreds or thousands of Github stars and dozens of contributors, include the Bluebird promise library by Petka Antonov, Bacon.js, the FRP library by Juha Paananen, the Calmm.js collection of state management utilities and libraries by Vesa Karvonen, and dozens more. There are many of us, with thousands of projects. You might notice that Reaktor is pretty close to the top when it comes to open-source contributions by Finnish organizations on commits.top (by Lauri Piispanen from Reaktor), which is naturally open source itself.
All this hard work – and the open-source community in general – has brought many benefits for Reaktor, so for a few years we’ve run the Reaktor Grant program, sponsoring some high-visibility open-source projects including Bluebird and Bacon.js. However, we were conscious that it wasn’t benefitting the many other lesser-known – yet still valuable – projects.
Supporting open-source work
Last year, a group of Reaktorians started to think about how we could, as a company, offer support for open-source work more systematically. In order to encourage even more Reaktorians to participate and to acknowledge our respect for Reaktorians’ open-source work, we launched an internal open-source support initiative called the Juice program.
When we were creating the program, there were two things we wanted to avoid. First, we didn’t want to reinvent the wheel – if there was a solution already available, we were ready to try it out. As it turns out, several software companies have searched for ways to support open source, so there were certainly experiences to learn from. Second, we didn’t just want to put something together and keep our fingers crossed that it would work. Instead, we wanted to do what we do best: experiment, iterate, and find the solution that works best for us!
Providing a balance
We ended up starting a three-month experiment of a new, and partially borrowed (thank you Futurice and Wunderdog), model of supporting open source. The premise was simple: by logging your open-source hours to a Slack bot (with the help of our friends at Wunderdog), you get monetary compensation of €15 an hour for a maximum of 20 hours per month. The bot makes your work visible to colleagues, helping to share kudos, spark discussion – and, even better, inspire collaboration.
The hourly rate and monthly limit are deliberately set reasonably low, so people aren’t encouraged to do a lot of hard work in addition to their regular job. It’s more about showing respect for and bringing visibility to the good work Reaktorians do while inspiring more of us to participate in open source and learn new skills at the same time. Ideally this benefits the company, employees, and the open-source ecosystem, too.
We received a lot of feedback from enthusiastic open-source hackers and problem-solvers in our community as we started to explore whether the format works for us. Some of the initial concerns were whether the program encourages burnout in people that are working on OSS in addition to their regular work, if the program is too developer-centric, and if it only benefits those (usually younger) people with a lot of free time on their hands. For some of the questions, we had answers or hypotheses available: the program likely could not cater to everyone’s needs, but that doesn’t mean it’s not a good thing to implement. Many of the questions we had no answers to – hence the experiment.
The experiment was well received and had a great number of happy hackers participating. Here are some of the observations we made from the experiment data:
- Participant ages varied from 20 to >50 – it wasn’t just youngsters
- All contributions so far are code
- Four out of five participants observed an increase in their OSS activity
- Almost all participants would like to continue the program and were generally happy with the experiment!
What we achieved
Perhaps the most interesting results (for both you and us) were the contributions made. Here are a few cool and interesting picks:
- Typera Node.js routing library by Petri
- Signal K Stash by Jouni and Teppo
- Type definitions for Ramda by Miika
- Contributions to AWS SDK by Henry
- Robotini Racing Simulator by Mikko and two other contributors
- Bacon.js finally converted to Typescript as a team effort
- AWS CDK fargate starter kit by Markus
- Optimizing BrainFuck -> x86_64 compiler by Hannes
- Consultant Herding Game by Joona + a bunch of other Game Jam productions
- Harmaja frontend framework by Juha
Based on the results, we decided to continue with the original program since our feedback survey didn’t detect any major flaws. None of our perceived risks seemed to materialize – instead, we observed a lot of happiness and collaboration. Overall, we felt that the experiment had a positive impact on both Reaktorians and the wider open-source community.
Over to you
Fast forward to the present: where are we now? The Juice program is going strong and open source at Reaktor is (in our humble opinion) in a good state. The program has attracted new participants, even ones who joined the company after the experiment, which we consider a good sign.
There is still work to be done – the program has so far attracted mostly developers although the aim was for it to cover all openly licensed intellectual work. We hope to find a design perspective on the program in the future.
In the spirit of open source, we are sharing our findings and experiences with the Juice program in the hope that it will help people who want to try a similar initiative. Just as Juice was inspired by similar programs at Futurice and Wunderdog, your company can learn from our experiment to help enhance your support for open-source and personal development. Feel free to copy our recipe if you feel that it resonates with you – after all, that’s what open source is about!