The 2018 Letter to Stakeholders (Year-end)
2018 was another successful year for our group, continuing the growth of prior years. We published in top journals and conferences and further ramped up revenues through an array of DFG, BMBF, and BMWi public grants that we won as well as industry funds that we acquired. The personnel situation remains dire and we were not able to hire enough people to fill all open positions. Perhaps most noteworthy, this year, we also kick-started the transition of the Sweble research project into a commercial startup.
The timeframe for the research section of this letter is the 2018 calendar year.
The research group’s work is structured by research domain; a domain may have several projects. Our primary research areas are open source and inner source as well as the three projects JValue ODS, QDAcity and Sweble.
In more detail:
- Open source. In 2018, we continued our research into open source governance and compliance within companies. We extended our handbook for open source governance with industry best practices on getting started with corporate governance, on open source component approval and reuse, and on supply chain management. We worked hand in hand with three industry partners to implement and evaluate parts of our research results in a real-life setting. We also published and presented a paper on Understanding Industry Requirements for FLOSS Governance Tools at the 14th International Conference on Open Source Systems in Athens. In addition, old work bore more publication fruits with a paper on A Qualitative Survey of Episodic Volunteering in Free/Libre and Open Source Software Communities in the IEEE Transactions on Software Engineering journal.
- Inner source. We continued and expanded earlier work. The three main focal areas within inner source were management accounting of inner source code contributions, transfer pricing of contributions that cross tax boundaries, and quality assurance through code review practices of said contributions. As part of our management accounting work, we published a paper on The Patch-Flow Method for Measuring Inner Source Collaboration at the 5th International Conference on Mining Software Repositories. We expect this work to continue in 2019, with an added emphasis on transfer pricing.
- The Sweble project, our longest-running project, is folding tents as a research project and setting up shop as a commercial startup, see the startup section of this letter.
- The JValue ODS and QDAcity project remain in research stealth mode with no publications this year.
We won research grants from all three major funders in Germany, the DFG, BMBF, and BMWi.
We continued our collaboration with our colleagues Minghui Zhou (Peking University), Klaas-Jan Stol and Brian Fitzgerald (University of Limerick), Ademar Aguiar (University of Porto), Rosni Abdullah (University of Sciences, Penang), and Ian Chai (Multimedia University, Kuala Lumpur). We are collaborating on various Ph.D. level research projects associated with our own projects.
In addition to this, we provided the usual academic services like reviewing for journals and conferences.
The timeframe for the teaching section of this letter is the 2017/18 academic year, that is, the winter semester 2017/18 and the summer semester 2018.
Student interest in our courses kept growing, as the following chart shows, presumably in line with overall increased student numbers:
This academic year was the last one where we tried to scale and meet student demand. In future semesters, we intend to reduce the number of students in our courses, because we do not have the capacity to meet the overall demand.
The number of final theses students roughly stayed on the same high level as in prior years. The following figure shows the total number of final thesis students we served.
It is worth noting that we continue to have more Master students than Bachelor students. This is probably still due to the poor positioning that our courses, all of which are electives, have in the overall computer science colloquium. Students only find their way to us later in their degree program, when at least a bit of appreciation has developed for software engineering research and development topics.
For many years, we have integrated industry into our teaching, most notably through our course on agile methods, called The AMOS Project. This was done for mutual benefit: Students experienced more realistic projects and industry got in touch with promising students. In AMOS, students learn agile methods and develop a useful piece of open source software. Also, in 2018 we had an interesting and successful ARCH (software architecture) project, in which students analysed the code architecture of an automotive OEM’s infotainment software.
Competition for students is fierce, and in particular our flagship course AMOS suffered in 2018. For this reason, we will pause AMOS at FAU for a year and therefore will not offer it in the coming summer semester.
The timeframe for the industry section of this letter is the 2018 calendar year.
Most of our industry engagements are through our research projects, see in particular the research project domains of open source and inner source. In addition, we engage with industry through our teaching projects like AMOS or ARCH, see teaching.
In 2018, industry outreach was strong like in prior years. In more than 30 invited talks we communicated our research work and results to current and future industry partners.
Our 2018 industry partners and sponsors (who were willing to be named) were:
If you are interested in sponsoring research, there are two main models for industry to engage with us. In one model, you directly sponsor a Ph.D. student, as described here: https://goo.gl/B1v6Xb. In the other model, we define classic projects with deliverables, as described here: https://goo.gl/XLrE6M. We are open to both models, but the specific contracts possible depend on our current research projects.
The Sweble research project pioneered a new way of collaboration in knowledge projects, dubbed “Github for wikis”. Now called editive (commercial name), this web service lets users create wikis and fill them with content. Then, unlike traditional centralized wikis where everyone edits the same central instance, editive lets users copy a wiki at the push of a button and work on that copy. Work from different copies of the same wiki is integrated into the original wiki by way of merge requests, and the updated and integrated new content is pulled from the original wiki to the copies. Hence editive’s vision:
A world in which everyone can collaboratively develop and individually own the content they care about.
Editive’s approach to collaboration cleanly separates the task of original knowledge creation from the task of knowledge integration. We view it as a major contribution in making knowledge collaboration in companies more efficient and effective.
editive acquired initial public funding of 710.000 Euro and intends to incorporate in early 2019, ready for its first customers.
In 2017, we acquired the Softwarecampus 2.0 project for FAU, bringing a financial project volume of 1.854.636,09 Euro (+20% overhead) to the university. These funds are available to all professors who successfully file a so-called microproject. In 2018, we had one such project, for 2019 we already know that we will have four more projects. Our group manages the program for FAU so that our colleagues can easily apply for and receive the funding.
In addition to this, we provided the usual administrative services in various university committees and affairs.
The timeframe for the fundraising section of this letter is the 2018 calendar year.
The following figure shows our revenues over the time of the group’s existence.
Our revenue growth in 2018 was about 25% and thereby significantly higher than in prior years, when it hovered around 20%. As always, revenues are closely trailed by expenses, as we spend most of the money on salaries and related personnel costs. We expect 2019 to be even better, going well past the 1 million Euro mark.
We supported our university by paying 10% of our revenues or 90.000 Euro as an internal tax (“overhead”). This is slightly less than in 2017 and is due to the lower percentage of industry funds raised in 2018 as well as internal accounting adjustments. We now use a university-internal service called wtt to provide to our industry partners a smoother experience in billing and contract matters. Of total revenues, industry funds were only 38% in 2018, down from 50% in 2017 and 2016. We expect the relative growth of public funding to continue.
My team and I would like to say thank you for your interest in our group and for reading this letter. I hope you found it interesting and we will either keep collaborating or we will get together for new groundbreaking research.
For the open source research group,
Prof. Dr. Dirk Riehle