Michael Dexter at Linux Fund

We recently asked the OSAlert community for input to create questions to send over to the folks at Linux Fund. The questions were compiled and sent over to Michael Dexter, program director at Linux Fund, and he graciously complied in addressing them and clarifying just what it was we all wanted to learn.

Is your organization an investor, expecting a return
on its investment, or a disburser of no-strings-attached donations?

We interpret “Fund” as a verb rather than noun: we fund things and have
done so for a decade. We are a full 501(c)(3) public-benefit nonprofit
and are not after the returns that, say, a venture capital fund would
seek. In fact, in contrast to the pursuit of such returns, we have
negotiated with a few proprietary software vendors about products that
are commercial failures but could be open source successes if
re-licensed with an OSI-compliant license. As for “Fund” the noun, we
are considering the establishment of an endowment that would support
nonproproprietary software and hardware.

Is it necessary for open source projects to receive funds like
this? Are there some key open source projects that just can’t boostrap
themselves with volunteer labor, consulting, etc? How does your typical
open source project sustain itself? And what are the shortcomings to
this “traditional” path that you’re hoping to address?

Yes, with regards to the necessity and your touch on three related issues
here: one-time project assistance, project bootstrapping and long-term
project sustainability. With regards to one-time assistance, we often
find that funding is the last tool that is appropriate to help a project
in need. Developers are generally well-equipped to make technical
decisions on behalf of a project but are unprepared to serve as strong
financial and administrative decision makers. We have encountered
situations where projects more urgently needed help with nontechnical
matters such as attending an event, shipping equipment or finding a home
for a domain name. We always try to help projects exhaust any and all
free resources that are already available to them and are preparing to
formalize this help with select conservancy services that would provide
such administrative support in an unintimidating way.

Project bootstrapping takes a number of forms including the authoring of
new code from scratch, the adoption and completion of promising code
that has never quite reached production quality and the relicensing of
proprietary code. Our direct donation projects have addressed all of
these approaches to varying degrees and with varying success. In all
cases, we believe that the public awareness of a project is just as
important as any new lines of code because both bring the project closer
to a critical mass of community that should result in priceless amounts
of future technical and nontechnical contributions. While new and
relicensed proprietary code projects promise to best meet any given
technical needs, such projects usually face the challenge of beginning
without a community. “Build it and they will come” is genuinely possible
thanks to the Internet but mindshare is still an elusive and precious
commodity. I am most excited about promising pre-1.0 projects as they
often have established communities and could genuinely be “best kept
secrets.” The hangup is that often such projects have strong but
introverted communities of developers who have satisfied their own
technical needs with the project and either do not recognize the value
of a broader user and developer community or are actually hostile to the
idea. Two promising but under-the-radar projects that fall into this
category are the GNUstep and Lazarus development environments.

As for long-term project sustainability, we are witnessing a gradual
transformation of the software industry away from licensing-based
funding and business models to countless alternative models, the best of
which haven’t been invented yet. While project bootstrapping is
important, projects can face financial challenges at any stage of
maturity. For example, the foundation behind the Gnash project recently
lost its funding. How an organization responds to such challenges is the
key: the Free Software Foundation turned to a membership model to
address its funding challenges. Our organization was once faced with the
loss of our main credit card but survived. We’re pleased to see that the
FSF is trying the credit card model of funding that Linux Fund
pioneered. This is simply the nature of public good in the USA: it comes
with no guarantees and, as Andy Grove put it, only the paranoid survive.

There is, however, the occasional unexpected success: the Mozilla project
received so much money from its Firefox Google search toolbar that it
jeopardized its nonprofit status. You highlight consulting and, like the
Firefox search toolbar, there are ways to incorporate such “commercial”
activities into nonprofit software foundations. Some nonprofits, in fact,
wholly own for-profit consulting divisions which in turn advance their
mission. As long as software costs nothing to duplicate, I am convinced
that nonprofit software foundations will eventually dominate the
software industry, but again therein lies the challenge: what sources
of revenue will sustain them?

To address these issues, Linux Fund is continuing to develop new funding
models and introduce unintimidating administrative services that will
gradually grow unincorporated, volunteer-driven software projects into
authoritative foundations.

What do your donors and sponsors hope to achieve with their
support? Do you find that it’s mostly interest in a particular project,
or is it OSS in general?

Donor motivations vary greatly and require great care to leverage.
Individual donors who are project developers often see the value of the
work to the project as whole, and may personally see their lives made
easier by it. Non-developers are more inclined to want to benefit either
a given project or category of projects. I love when people support two
loosely-related projects like the OGD1 and gEDA/PCB as I know they’ve
recognized the relationship. Corporate sponsors are primarily after the
PR value, and we highlight a number of possible approaches they can take
in directing their donation. The majority of our direct donations to
date have been from grass-roots individuals, and we see great value in
this growing community.

Do you provide any non-monetary assistance to the people behind
open source projects (such as public relations, organizational, or
technical help)?

Yes, as outlined before; I also invite you to watch for formal
announcements in the coming months. Our Ubuntu Massachusetts LoCo
project is, in fact, a beta test of several elements of these services
including event planning and printing sourcing. Their enthusiasm has been
amazing, and we see no reason for them not to benefit from the collective
wisdom of the community.

How do you determine which projects to fund and the cash goal?

I lead a Technical Advisory Board that identifies projects with unique
potential. Project selection begins with a lot of listening, and it was
several months into our direct donation projects that I discovered the
FSF High Priority project list only to find that we were already
addressing several of its concerns. Our ideal partner project is one
that is promising but not yet production ready but also aims to shatter
a glass ceiling in the software industry: enabling free and open source
software to enter a field where it currently is absent. Mechanical and
electronic CAD and EDA design software are a perfect example as
virtually everything made by humans has gone through design software. An
incalculable amount of data exists in .dwg and related AutoCad formats,
and this is the exact same situation the world faced with .doc and .xls
formats. Literally our collective knowledge is saved in those formats,
and it is critical that we have open access to the data locked in them.
Our gEDA/PCB and VectorSection projects address these issues, and if the
majority of the top 500 supercomputers can run Linux, it is absolutely
possible for nonproprietary software to penetrate traditional industry.

Every project begins with a budget to achieve a certain goal, and from
this we determine cash goals ideally in incremental steps and
milestones. Incremental projects allow individuals to adopt certain
portions of a project and know they are making a difference. This
effectively allows a nontechnical individual to make a solid code
contribution through a financial contribution, and that’s exciting. I am
glad to see that the Google Summer of Code project has encouraged many
projects to think about development budgets for the first time.

Does a project’s choice of open source license have any bearing on its
eligibility for funding?

Our top concern is that project’s license is OSI-compliant and we
encourage the use of the more popular copyleft and copycenter licenses
such as GPL and MIT/BSD/ISC. My personal opinion on nonproprietary
softare licensing is that we’ll need to wait about twenty more years to
see who has the best approach. While there is clearly a valuable role
for lawyers to play in this community, we would prefer see funds first
and foremost go to developers.

How will the introduction of the BSD Fund affect the allotment of
time and resources to Linux projects?

Fortunately, BSD Fund coattails on virtually all Linux Fund legal and
mechanical infrastructure and does not introduce significant time or
financial overhead. With the recent introduction of the BSD Fund Visa,
BSD Fund is a bit like Linux Fund circa 1999.

What proportion of donations go towards administration and running costs?

If you want an exact figure, under 8.25% to date, and only upon the
successful completion of a project. We continually refine our model for
partner project direct donations and hope to eliminate any overhead
withholding through direct donations to our General fund and other
revenue sources such as our credit cards. Donors should be aware that
with the relatively high cost of electronic processing fees, you can
help any given project most effectively by writing an ordinary check.

OSAlert would like to thank the OSAlert readers for supplying the questions, and of course, Michael Dexter for taking the time to answer them. The next interview with community questions: the Arch Linux team. That interview will be up later this week.