December 29, 2022

The year 2022 was a busy and productive one for the FreeBSD Foundation. Continuous Integration (CI) and Quality Assurance (QA) is but one of many successful efforts the Foundation continues to support as a dedicated resource. We spoke with Li-Wen Hsu, one of the Foundation’s Software Engineers, to discuss his work on this project over the past year, and the plans he has for it this coming year.

Q:  Share with us please, some of what you have accomplished this year. Do help us understand both your accomplishments and the Foundation’s role in supporting this effort please.

Hsu: In 2022, we’re mainly doing two things, maintaining and extending the current continuous integration (CI) system, and working with the workflow group to build a more productive and streamline workflow for the developers and contributors by leveraging or reusing what we have in the current CI system.

Under the realm of maintaining and extending the current CI system, we continue to add new jobs, such as gcc12 and powerpc related jobs, to fulfill the development needs. We also retire some jobs whose targets are not supported anymore, such as mips related jobs.

We also spent a lot of time working with the contributors and developers to fix the failing test cases, and added some more test cases.

If readers would like more details on any of this, they can find them in the “Continuous Integration” entry of each online quarterly report at https://www.freebsd.org/status/.

We brainstormed with many people about what the workflow of theProject needs, and continue working on setting up the proof-of-concept system for evaluation.

Also, it’s important to mention that the FreeBSD Foundation also supports the hardware within the CI system. During 2022, the artifact storage space has been expanded two times more, enabling both the storage of a greater variety of artifacts and a longer retention period. These benefit debugging and testing efforts by a lot.

Q: What are your ideas and plans for 2023 and beyond?

Hsu: There are few top priority tasks, which include:

  • System for pre-merge testing for developers 
  • A new workflow that can at least be in the stage of public testing 
  • Snapshot and release builds can be generated by the CI system to reduce the release engineers workload

There are already too many items in the TODO list in WIP and open tasks in the status report. But if possible, I’d like to help more software adopt FreeBSD in their development CI and see major players of the hosted CI solution receive better support of FreeBSD.

For the hardware part, we hope to acquire more hardware to fill the space of the broken and discharged machines, and better leverage cloud resources.

Q: Is there anything else you’d like to share with us about your work, plans, or the Foundation’s role?
Hsu: I would say improving FreeBSD’s CI is just the beginning of my contribution as a technical staff from the Foundation. The essential focus of the Foundation’s role should be to support any technical needs of the Project development. While currently CI is the most important part, just like in the past we supported Git migration, we will of course work on anything that can improve the experience of contributing to the Project.

Besides the CI work, I also spent some of my time recruiting and mentoring students to let them know more about FreeBSD and help them contribute to it. I spend much of the rest of my time trying my best to initiate collaboration between the Project and developers from both corporate and academia. It’s always good to know when someone is interested and willing to help.

– Contributed by Pam Baker