May 1, 2006

FSF supports average users with high-priority list

Author: Bruce Byfield

The Free Software Foundation (FSF) is frequently considered an organization for developers rather than end users, but Peter Brown, executive director of the FSF, would disagree. "We don't just want freedom for software developers," Brown said in a telephone call interview last month. "We want freedom for all." One of the ways that the FSF promotes this goal is with its
high-priority project list.

The list is vaguely reminiscent of the recent Novell survey of applications users would like to see ported to GNU/Linux, but its purpose is more specific -- to encourage the development of the tools ordinary users need to run a free system that can do everything a proprietary one can.

By calling attentions to gaps in functionality on free systems and by a combination of advocacy and practical assistance, the high-priority list, Brown believes, has emerged as a modest success in the last year. In fact, he says that several items on the list can be deleted as done.

Currently, the high-priority list consists of six items. At the top of the list is a free BIOS (a.k.a. LinuxBIOS), which the FSF considers an essential safeguard against so-called trusted computing, and free drivers for ATI graphic cards. The next four items are about to be retired: a release of 2.0 with a free implementation of Java; Gnash, a free Flash player; and GNU Compiler for Java (GCJ) and GNU Classpath, a collection of free Java libraries. The, GCJ, and Classpath items are closely connected, because GCJ and Classpath were needed to build with a free Java implementation. Brown expects these items to be replaced with others as a result of feedback from volunteers, including some from the FSF general meeting held on April 1.

Putting together the high-priority list

Brown explains that items are added selectively to the list. Rather than "overwhelm people," Brown says, the goal is focus on the greatest needs. For this reason, the list does not usually contain feature requests for mature applications. Similarly, with so many potential candidates for potential inclusion, the FSF is not strongly concerned about improving an item once basic free functionality is available. For example, Brown considers it a matter of secondary importance that, in versions of built with GCJ and Classpath, Java-based features are painfully slow. He'd rather encourage the project not to include further Java-based features. "There's always room for improvement," he says. "The important thing is that the people at now are aware of the issue, and they are committed to not allowing more proprietary dependencies creep in. So, we've made the major step."

The Foundation may also choose items because they represent a larger problem. For instance, although only drivers for ATI video cards are on the list, Brown agrees that they are there largely as a representative of their entire industry. Drivers for Nvidia, he agrees, could just as easily be on the list instead. "At the moment," he says, "the major manufacturers are not supportive of free software and of making their drivers available Sunder a free license]."

In order to conserve the FSF's resources, other criteria also seem to be unofficially applied. Priority seems to be given to items that are already supported by a project that can achieve success with only a little more support. Also, all items on the high priority list are centered on GNU/Linux, not because the FSF does not support other free operating systems, but because it is by far the most popular.

Suggestions for new additions come from FSF supporters and volunteers. For example, Brown attributes the drive to support Gnash directly to a Lawrence Lessig speech about the needs of graphic artists at the FSF annual meeting in 2005. Suggestions are not voted on, but adapted after internal discussion at the FSF. A particular concern is software that has no free or open source equivalent. "We're always very concerned about proprietary applications that are seductive and pulling people away from free software," Brown says. "When we see such applications, it's very important that we react to them and try to hold the line."

He expects a new forum organized by region that the FSF implemented after this year's general meeting to be a source of new suggestions for the list. Those who wish to advocate including a specific project can send the project's site and the reason for including it to

Advocating list items

Whenever possible, the FSF tries to works with existing free software projects that are already working on list items. "We don't want to come in and take over," Brown stresses. "Our aim is help developers decide for themselves how the FSF can support their efforts."

At times, as with GNU Classpath, the FSF's support extends to server space for hosting the project. More often, Brown says, FSF support amounts to "moral support and a call to attention." Besides general publicity in the media, this support can include calls for volunteers, as happened with both Classpath and Gnash within the past year.

In some cases, advocacy takes stronger form. At least for a few days last spring, the FSF seemed to be conducting a campaign to discourage the project from adding Java-based features in the future. More recently, FSF support has taken the form of consumer campaigns to advise supporters about which companies they should deal with.

For example, as part of the push to develop a free BIOS, the FSF is urging supporters to buy Tyan servers and motherboards with AMD chipsets. AMD, Brown explains, "is very helpful toward the LinuxBIOS project." It provides an engineer who tests that LinuxBIOS works with AMD chipsets and other hardware. Tyan, too, Brown says, has "provided a great deal of support." When visitors come to the FSF office, Brown says, employees like to mention that the office uses Tyan servers. By contrast, the FSF Web page for the project discourages supporters from buying IBM or Intel hardware because of these companies' incomplete support for the project and endorsement of trusted computing.

All these attitudes are tentative and somewhat qualified, Brown says, but, "In all these things, we have to draw a line and say, 'Well, no one's perfect, but how can we leverage one corporation's good work against another corporation's less supportive work?'"

Evaluating the list's effectiveness

Is the high-priority list effective? At times, it is hard to tell. Although ATI and Adobe, the present owner of Flash, both develop software that items on the list hope to replace, neither would officially comment on the the effectiveness of the list. However, to what extent this lack of response indicates ignorance, caution, or cunning cannot be determined. It seems likely that at least some employees of these companies are aware of some of the projects on the list. Rob Savoye, the maintainer of Gnash, for example, notes that several Flash developers subscribe to the Gnash email list. Because of their employment, these subscribers are restrained from participating in development, but they are aware that a free alternative is being developed.

A few developers involved in high-priority projects are also uncertain about the results. Caolan McNamara, who did much of the work of building a version of with free Java, feels that his project was added to the list too late in its development to have much of an effect, while Tom Tromey, a long-time developer on GCJ, was unaware that the list existed until he was contacted for this article. Similarly, Ronald G. Minnich of the free BIOS project writes, "If it has affected us, we can't tell."

However, others say that FSF advocacy through the high-priority list had a strong effect on their project. Mark Wielaard, one of the main Classpath developers, notes that having Classpath on the list encouraged journalists to contact project members, giving them a chance to explain the importance of the project. In addition, he says, "People take our project and goals seriously since we have a good relationship with the FSF." He says the FSF has supported Classpath with server space as well as moral support and assistance with recruiting.

Savoye agrees. The FSF's support, he says, "has helped attract other developers" and to increase interest in Gnash throughout the free and open source software communities. At times, Savoye feels that working on such a high-profile project is like being in a fishbowl, since he is now correcting his own mistakes very publicly. He considers the attention "a mixed blessing," explaining that "I get more bug reports, but I also get more patches." Yet, on the whole, he seems to welcome the attention. "I guess it'll encourage me to make Gnash a more stable, portable, and maintainable project," he concludes. From such remarks, it seems that, in many cases, the FSF priority list is the success that Brown claims.

The future for the list

Brown is cautious about suggesting what items might appear on the high priority list in the future. However, he does suggest that laptop video drivers are a large ongoing issue. "We're obviously very concerned with hardware," he acknowledges.

He also hints that future items might center on increasing open access to government resources on the Internet. This is an issue that is of particular interest to the FSF because the number of charities and non-government organizations that run on free software is higher than in the general computer-using population as a whole. The issue has attracted some attention because of the promotion of the OpenDocument format, especially in the FSF's home state of Massachusetts, but Brown suggests that only some of the concerns are being discussed. Companies support an open format "because they want a level playing field," yet "when it comes to computer users' freedoms, these same corporations shy away and forget the linkage between these issues."

No matter what the upcoming content of the list, Brown sees it as continuing to be one of the tools that the FSF uses to achieve its goal. "We want to keep this list current," he says, "and we want to keep it pushing forward on these issues." Contradicting the perception the FSF is oriented towards developers, he adds, "It's really only the FSF that draws attention to these computer user issues."

Bruce Byfield is a course designer and instructor, and a computer journalist who writes regularly for NewsForge, and IT Manager's Journal.


  • Free Software
Click Here!