Sugar moves from the shadow of OLPC

Fourteen months ago, when One Laptop per Child (OLPC) announced that it was preparing to work with Windows, the free software community treated the news as a betrayal. However, nowhere was the reaction stronger than within OLPC itself. Within days, Walter Bender, who oversaw the development of Sugar, OLPC's graphical interface, had resigned and announced the creation of Sugar Labs, a non-profit organization for ensuring Sugar's continuation. Now, looking back over the year since then, Bender considers that Sugar Labs has progressed steadily towards its main goals: getting organized, taking advantage of the nature of free software to enhance education, and developing a community of teachers and students to help direct Sugar's development.

"It turns out that One Laptop per Child didn't really go down the Windows path," Bender admitted. "They're still shipping Linux and Sugar with every laptop, and they've just announced that their 1.5 machine is also going to be Sugar-based." However, he had no way of knowing that last year.

At any rate, the proposed change of operating systems was not the only reason for the creation of Sugar Labs. "At the same time, I was thinking that Sugar should be broader than just one particular hardware platform," Bender added. "We spent a lot of the first year undoing any specific dependencies on the One Laptop per Child hardware. What we've done is made Sugar run pretty much everywhere."

Sugar is now available in most major distributions, and Sugar Labs is currently in the middle of developing Sugar on a Stick, a USB installation. In addition, Sugar Labs has been discussing Sugar deployment with several netbook manufacturers.

"I think we've got a lot of ways to go in terms of raising awareness about Sugar," Bender said. "I don't think the world knows that Sugar's out there, and that it's a separate thing that can run outside of One Laptop per Child. That will take time. Our marketing budget is the same as all the rest of our budgets — zero — but we'll get there."

At the same time, Sugar Labs has established a structure more in keeping with the free software and interactive learning ideals of its members. Modeling itself after free software organizations like the GNOME Foundation, Sugar Labs is run by an elected Oversight Board, which is deliberately designed "to be pretty toothless. About the only power that the Oversight Board has is to appoint a committee to solve problems."

Instead, it is the project's various teams that make most of the decisions, with all discussions occurring on the same IRC channel. As with most free software projects, this organization is born out of necessity, since Sugar Labs does not actually have any office space, but Bender suggested that it is also an advantage in building the type of community that project members desire.

By contrast, he suggested that OLPC, which does have a physical location, "was struggling a bit to maintain communication with the global community. Conversations were not deliberately obfuscated, but, because they were happening in a room as opposed to online, a lot of people weren't hearing the conversation, and a lot of people felt they weren't part of the project because of that. We don't have one physical center of activity, and that plays out to our advantage, I think."

All in all, Bender stated, "Things have gone remarkably smoothly. We've been a pretty disciplined bunch. For instance, if you look at our release map, we're not letting features get ahead of our ability to deliver something that is robust and on time. For the most part, it's been a great year."

Free software ideals and education

However, for Bender, Sugar Lab's greatest accomplishments have not been in its organization, but in the advancement of its educational goals — goals that Bender views as meshing remarkably well with the ideals of free software

From the start, Sugar has been intended as more than an interface. Instead, Sugar Labs prefers to describe the software it develops as a learning platform. "We're not interested in anything except learning," Bender said. "as we make decisions about what we release and what we do, the first question we always ask is, 'How does this impact learning?' But 'platform' is important, too, because Sugar is not just a finished product that we give you to use. The implication is that Sugar is the platform on which you are going to build as well. We give you some scaffolding, but that scaffolding is there for you to build upon, as opposed to something you just use."

In this respect, Bender views Sugar as radically different from proprietary learning systems. "Often times, the tendency is to give children toy versions of a program, so that they don't hurt themselves. Also, the professional versions are expensive. Sugar takes a different approach. We want to give them tools that are real, and don't limit them in any way. But, at the same time, we're very cognizant that we have a path that doesn't require them to be an expert to get started. Really, that's an approach that is possible to achieve in free software, but quite difficult to achieve in any other way."

For example, Sugar's music activities begin with "tools that are literally accessible by a two year old — a sort of pound-on-the-keyboard activity. They can take that tool and use it over the network to make a band, and use it for sequencing and to compose music. then they can go into a synthesizer and start to play with wave forms, random number generators and envelope curves. Then they can go into a scripting language called cSounds and start to understand how music is scripted in the machine — and that's the same language that the pros use in Hollywood for scripting special effects. Then they can go into our View Source editor and hack the Python code that's underlying all these tools. They have the ability to go deeper into anything — and anything is determined by the child's interests, [or] by the teacher. It's not determined by the people writing the code."

In much the same way, Bender considers the collaboration framework that is part of the basic Sugar interface as a direct invitation for open-ended critical discussion. "This has a very direct connection to free software," he said. "One of the things about free software is that not only do you share, but you also engage in a critical dialogue about what you're sharing. The idea that ideas are there to be critiqued is one that a child has got to learn, and Sugar collaboration is as much about being engaged in criticism as it is about sharing." In other words, Sugar Labs considers the collaborative development found throughout free software projects as a model of exactly the sort of interaction that is ideal in education.

Community Building

Another area of success in the past year has been the increased participation of teachers and students in Sugar development.

"We've been really persistent about constantly going back to the teachers and saying, 'You've got to participate. You've got to give us feedback,'" Bender said. "'If you don't give us feedback, we're not going to learn, and you're not going to get the kinds of tools you need.'"

Bender acknowledged that finding the right channel for this feedback is difficult. He would personally prefer IRC, "because on IRC you've got this discussion among experts [who are] solving problems," as well as access at any time of the day. However, experience has taught him that IRC is not "a tool that teachers are going to be comfortable with, at least initially." Instead, teachers seem to prefer email, despite the fact it is not instantaneous and that communication is asynchronous.

But, regardless of the medium, teachers and students are starting to let their views be known. For instance, in Uruguay, teachers using the Turtle Art activity (Sugar's name for an application) requested a square root function they could use to teach the Pythagorean Theorem. At first, Bender told them to write the function themselves. But, eventually he realized that such a contribution was beyond most teacher's ability. He added an extension that gave them several different ways to add to the code, including some that did not require programming expertise, and the square root function got done.

What was more important than the specific function, though, was the lesson Bender learned about coding styles in general. "I wasn't thinking how I could make things extendable by teachers when I [wrote] it," Bender said. "Now, that's at the forefront of my mind."

The last year does not seem to have produced any example of student involvement comparable to one Bender remembered from 2007, when Igbo-speaking students in Nigeria produced their own spell-check dictionary for Sugar's Abiword-based Write activity. However, Bender did mention that he was scheduled to be interviewed soon by students in Boston about Sugar, and an upcoming project to have students write Sugar documentation, so Sugar Labs is trying to engage students in discussion as well.

Still another mechanism for receiving feedback is the relatively new practice of establishing Sugar Labs wherever the interest exists. The goal, Bender said, is to "have those local centers be responsible for local support and localization. Those centers can be pretty much structured in any way that's appropriate to that region. We aren't going to impose a structure on them. So long as they share our core values, they can be Sugar Labs. That's beginning to happen now."

A Sugar Labs now exists in Colombia, and others are being organized in Washington, D.C. and Lima, Peru. "Each of them has its own set of issues they're trying to focus on, but all of them at the same time are participating in the global dialog," Bender said.

Building for a pedagogical future

Fourteen months after Sugar Labs was established, OLPC machines remain the major deployment for Sugar. However, Bender is encouraged by signs that the educational ideas behind Sugar are starting to be adopted elsewhere in free software.

Sugar has a long relationship with Fedora, the basis for the original OLPC operating system and an active participant in everything from Sugar on a Stick to the Oversight Board. But, in the last year, Sugar Labs has started to work more closely with distributions that are specifically geared towards education. For example, Bender suggests that his discussions with the developers of Skolelinux, an educational distribution based on Debian, may have helped them move their planning beyond the mechanics of installation and system maintenance or of software selection.

"I think they were only just beginning to think deeply on how Linux can impact on learning," Bender said. "I think that's why there's a lot more interest in Sugar recently, because we've been thinking about that question right from the beginning. I'm convinced that there's not just technical merit in Linux for learning — there's cultural and pedagogical merit."

Increasingly, Bender hoped, other parts of the free software community will take over some of the technical aspects of Sugar, such as solving their own problems with adapting Sugar to specific distributions. If that happens, Bender said, "that means our attention can be elsewhere, focusing on the dialog with teachers and with students, and making sure that deployment is specific to the needs of schools is being done."

Concluding, Bender said, "This is important stuff. We've got to do something about giving every child an opportunity to learn and be a critical thinker. There's all these problems that the wold is facing right now, and the only way that these problems are going to be solved is by putting more minds around them. Whether it's Sugar or something else, I think that the community really needs to be working hard at engaging more people in problem-solving and engaging them in learning. And I think that's the strength of the Linux community."

At a time when discussions about free and open source software tend to be centered on business, such comments may sound outdated in their idealism. But, after listening to Bender, the only conclusion one can come to is that is why Sugar Labs was created in the first place: To provide an alternative perspective that is harder to maintain in a commercial venture like OLPC. By that criteria, Sugar Lab's first year can be counted as a solid success.