Monday, December 23, 2013

Guest post: In 2014, let's make open source software more usable

This is a guest post from my friend Asheesh from OpenHatch.org. Asheesh and I met at an OpenHatch event sponsored at the university where I work, the University of Minnesota Morris. It was a lot of fun to participate in this open source software event; I recommend your university also look at hosting a similar event for your students and community.

I want everyone to be able to choose free software and feel empowered. But right now, so much free software—even some of the most widely-used software—is barely usable.

If this is a problem that matters to you, too, and you're willing to put in 1-3 hours per week of volunteer time over 2014, I think we can make a huge difference in fixing this. You don't have to know anything about usability today to do this; you can learn as you go.

So, we need a volunteer to step up. But first, a story:

I had the opportunity to meet the founder of the FreeDOS project, Jim Hall, yesterday in Morris, MN. Having created a popular (within its niche) operating system project, he's come to realize that the next frontier for open source is not more code: it's building software that people can use to get things done.

Free software hackers often think of usability as a vague notion, wondering what to do when confused users issues in their bug trackers. By contrast, usability researchers like Jim do something concrete: they name a specific task a user should be able to do with that app, and then tell some users: "Do that task as well as you can. Narrate what you are thinking, and we'll record the screen." If the goal is to build software that people can use, the test is finding out if they can actually use it.

The results show that people often get stuck at the same points when trying to achieve a task. It's excruciating for developers, who want to help by saying, "The save button is right there." In Drupal, recordings of these experiments caused the community to change links and text within the admin module. When they did that, user confusion decreased.

We can say that "user confusion decreased" because usability is in fact measurable. In Drupal, they performed the same experiment a second time, with that UI widget changed, and asked: "How many users were able to achieve the task in the given time?" If that number increases, then we know we are getting somewhere.

Jim is going to run a handful of tests as part of his masters thesis. He's not the first in open source to care about usability—the now-defunct OpenUsability project began in 2006. The question is, when his masters thesis is over, will the community continue to understand how to do this important work?

Here's my plan:

We—the free software community—need someone to take the burgeoning interest in usability and turn it into something that becomes embedded in existing projects and grows from there. That's something you can do.

To do that, you'll need to set a goal that in 2014, you'll find two projects that are interested in improving their usability, run one user test for each project, and work with Jim to document how anyone can run these tests. The measure of success is if one new person in the world runs a user test on a project they care about and shares that result with the project's community.

I think that won't take more than 1-3 hours per week over 2014, most of which will be spent reading about usability, chatting with Jim, and writing documentation on how to run user tests.

I, and probably other people, can support you, help you find those two projects that want usability help, and help get the word out. We can work together to help you present your usability data in the clearest way so those projects understand the findings.

You'll know that you'll have begun to change the culture of open source software in one of the most high-impact ways.

Do you want to?

1 comment:

  1. Yes. Where do I start? You can reach me at irina@valdyas.org .

    ReplyDelete

Note: Only a member of this blog may post a comment.