Mako's Essay on Funding Free Software Projects (extract)
Regardless of the steps that a project chooses, transparency will act in a central role in maintaining volunteerism. — Mako Hill I thought we could use this text, taken from Mako's essay, "Problems and Strategies in Financing Voluntary Free Software Projects", as a starting point for a discussion on what sorts of activities Sugar Labs should be funding. In the essay, Mako takes a strong stance against the introduce of paid labor into a primarily voluntary free software project, so this is far from a neutral starting point. Nonetheless, let the discussion begin! --Walter 13:26, 28 May 2008 (UTC)
"The safest way for voluntary free software organizations to spend money without compromising voluntary labor is to not pay for labor -- and the production of code in particular -- or to fund labor indirectly.
"One great way to do this, employed frequently by the Debian project, is to use donations to purchase hardware to help facilitate development. The most frequent and largest types of expenditure by the Debian project and its supporters is hardware for servers, other forms of infrastructure, and bandwidth. Each of these are viewed as shared resources within the project which avoids the appearance of paying an individual and valuing some contributions more than others — even if an individual makes disproportionately large use of the machine or infrastructure in question. When a developer steps down, the project infrastructure they use is passed on to another volunteer.
"Another good way to fund projects is to build "capacity." Capacity describes work or resources of any organization devoted to something other than the groups' core goal but that makes the project more effective or efficient. Developing or building an accounting system is a good example of capacity. Because it is outside of the core goals and work of the project, funding capacity is less likely to crowd voluntary workers out. Since infrastructure and increased or improved capacity can help an entire project run more smoothly, funding capacity in this way can help a voluntary organization become more streamlined and effective with only very minimal risks of crowding out volunteers, affecting transparency, or resulting in a lower quality product.
"Another good low-risk solution is to use funds to organize or sponsor conferences, seminars, meetings or workshops. For groups with less funding, money can be spent on securing a venue, bandwidth, food for attendees, or other conference expenses. For projects with more funding, need-based or merit-based travel aid can be offered to help attendees reach the conference. Conferences provide a venue for sharing information between members of a volunteer team, allow for in-person brainstorming and design sessions, help increase the quality of relationships between members of the project, and can act as a reward to "sponsored" developers.
"DebConf, the annual Debian conference, is a good example of strategic conference funding and is how the Debian project spends the vast majority of funding each year. Lodging, bandwidth, the conference venue, and food are normally paid for all attendees based on a first-come-first-served basis until funds are depleted. Additionally, travel aid is offered on a combination of need and merit based systems going first to speakers with accepted proposals, then to official Debian developers, and finally to qualified contributors to the project over the previous year. Funded conference attendance acts both as a reward for active volunteers and a step toward creating a positive, fun, and highly educational event for all attendees. Each year, there are participants who are less involved in the Debian project before the conference but who increase their involvement after. Additionally, DebConf acts as a venue for airing ideas and proposals between developers and for gathering feedback on controversial issues.
"A related, and sometimes overlapping model of funding involves funding code sprints. Popular in the Python and Zope communities and with increasingly popularity elsewhere, sprints are intense sessions of development — usually around one week long. They are used as catalysts for development and have seen major leaps forward in the development of features and code within very small amount of time. Sprints are like conferences in most aspects except that the emphasis is more on the production of code and sustained hack-sessions than on presentations and discussions. They also usually involve less people than a conference and attendees are usually limited to the most actively involved volunteers on a project. The Plone Foundation and SLX Debian Labs in Norway has effectively used the funding of sprints to accomplish time-based development goals.
"While conferences and sprints can be a clever way to spend money without sacrificing the voluntary nature of development projects, it is worth keeping one important caveat in mind. When a project selects people for funded attendance at the expense of others, it demonstrates favoritism that can be divisive. In the process of organizing these events, it is important to maintain a high degree of transparency and fairness. Organizers should use published and fair criteria to determine conference funded attendance and leave attendance open to all. Good criteria for fair selection includes "first come, first served," constructive activity on mailing lists, the number of important commits to a source code or documentation repository, or a good reputation among fellow developers (e.g., as determined by a fair and representative committee)."
Need for a Full-time Educator
Sugarlabs should acquire funding to hire a full-time field-educator to manage communication b/w developers and teachers. This is really critical to make sure that Sugar meets "felt needs" of teachers in the developing world rather than perceived needs. This person would also manager feedback on activities and requests from the deployments.
It is really critical that this person be an experienced teacher who has worked full-time at the primary or secondary level, ideally in a developing country.
An education Ph D from Harvard or MIT would be the wrong person. Those folks tend to focus on policy and theory, and tend not to have teaching experience in public schools. We need someone who is thinking about the problems of teaching long division in a constructionist way and the basic English grammar.
BryanWB 03:01, 29 May 2008 (UTC)