Bringing usability to open source. And vice versa.

"Usability" is one of the big raps on the reputation of open source code. Fixing that reputation at the source(s) is the idea behind a series of FLOSS (Free/Libre Open Source Software) Usability Sprints. The first was held in San Francisco this past February. By the time you read this, the second will have happened in July. Others will follow.

The organizers of the first and second events were Blue Oxen Associates (blueoxen.org) and Aspiration (aspirationtech.org), two nonprofits with a mutual interest in improving usability.

At Digital ID World in May, I met with Eugene Eric Kim of Blue Oxen, and interviewed him briefly about what the whole thing was about.

DS: What were your motivations for the first Sprint?

EK: The motivation for the event was multi-layered. We wanted to figure out a way to make open source software more usable. Both Blue Oxen Associates and Aspiration believe strongly in open source and use it ourselves, so we had a strong stake in the outcome. Both of us also care strongly about nonprofit organizations and believe that poor usability is one of the impediments for wider adoption of open source among nonprofits.

DS: Why a "sprint?"

EK: A sprint is a hands-on and collaborative clinic, where everybody works hard for three days to get real stuff done and learn new things in the process.

There's a knowledge gap between the usability and the open source communities; but the problem is actually that the two communities weren't previously motivated to learn about each other and to close that gap. It's a bit ironic, because these two communities share a lot of values: iterative improvement, user-centrism, a desire to improve things. They should have been collaborating with each other, but they weren't. Our goal was to build shared understanding between these two communities and then to see what happened.

DS: How did you select participants?

EK: We started with the largest and most diverse possible group: six open source projects (Activist Mobilization Platform, Chandler, CivicSpace, FotoNotes, Identity Commons, OpenACS, WordPress), twelve usability specialists (OSAF, Sun, Adobe, Adaptive Path, openusability.org, Stanford, Oracle), and a mix of developers and project managers -- for a total of forty people.

Our stated goal was straightforward and concrete -- for each group to improve the usability of some aspect of their projects in three days. On the second day, we brought in about 15 users to help the process.

Our unstated goal was to catalyze ongoing collaboration between these two communities. The event was highly interactive and was expertly facilitated by the good folks at Aspiration (who also do Penguin Day and other excellent events). We emphasized concrete outcomes, reporting out and self-documenting on our Wiki.

DS: What did you get out of it?

EK: We accomplished both of our goals and then some. Each project improved in concrete ways. Everyone learned something. Open source developers learned practical usability techniques, but more importantly, they were motivated to learn more. This wasn't training. It was learning-by-doing. The difference is critical. A follow-up bootcamp will be much more effective if the developers are motivated and if they have a context for which to understand these techniques.

DS: What did the two constituencies learn from each other?

EK: The usability folks learned a LOT about open source development. There were many misconceptions, both ways. It built shared understanding to work closely together in such an intense, high-energy, accelerated -- and, most importantly, FUN environment. Curiously, the usability folks hungered for an opportunity like this. For one thing, most usability practitioners never collaborate with each other. They have good professional groups, such as CHI and UXnet, but they rarely work in twos. A lot of folks approached me afterwards and said how much they enjoyed working with other usability people. These usability folks also hungered to do some good and make an impact. One problem with usability in corporate environments is that you can't talk about your work. Maybe you can point to a final product, but you can't point to the analysis and methodologies that got you there. You also can't show how crappy the project was before you came along.

So, a huge benefit for usability practitioners to work with open source projects is simply the ability to talk about it afterwards.

Both sides learned that usability and open source have an inherent tension. The usability mantra is, "You are not your own user." The open source mantra is, "Scratch your own itch." On the surface, these two philosophies seem contradictory. But, if you dig deeper into both of these, you'll find that it's not that simple. We gained confidence that collaboration wilhelp resolve these tensions.

DS: Did any new projects come out of the event?

EK: Two interesting and unexpected concepts and projects emerged from the event. The first was the notion of "extreme usability". See, usability is like security; it's not something you can just tack on at the end. The best route to usable software is to incorporate it into the process from the very beginning. As all of us explored this notion further, we converged on this analog to pair programming, where you partner a usability person with a developer. This resonated strongly with a lot of folks.

The project I'm most excited about is called OpenWebGUI (http://openwebgui.sourceforge.net/), because it shows the kinds of unique and powerful collaboration that open source enables. We had three CMSes there, and throughout the weekend, they realized that they were faced with similar usability problems. For example, all of them wanted more usable administrative systems. Even though they are all different systems, the functionality and workflow is essentially the same. These projects realized -- independently of any of us -- that because they were faced with the same problem, they were better off working together than in silos. They started a project called OpenWebGUI, which will focus initially on CMS administration GUIs. They will do an analysis, develop some HTML prototypes, and test those. The tested and usable HTML files will be released under an open source license, so that any CMS can incorporate them and customize them as they see fit. I love this outcome because it vividly illustrates the exponential returns that open source enables. We spent a bit of money to improve these projects, and yet the outcome will very likely improve the entire ecosystem of related projects. Now that's ROI!

DS: How else do you expect this to play out in the long run?

EK: Many folks involved with CHI are interested in being more involved with open source, and we are actively working to facilitate this. Jan Muehlig, a usability specialist who flew in from Germany to participate, has an outstanding project called OpenUsability.org, where he's trying to connect usability folks with open source projects. Jan did some excellent work with KDE, and we're actively trying to promote his stuff. And we've got a long list of projects and people who want to participate in the next sprint, at the end of July. We'll incorporate what we learned from this past sprint to improve the next one, and we'll also experiment with some new ideas. And we'll continue to engage with the community as a whole. These events are meant to perturb the ecosystem and catalyze the community. We encourage everybody to participate.

To find out how, visit www.flossusability.org/.