Bob, First of all, let me thank you for your years of dedication to the FOSSology project and in particular ...
Robert Gobeille wrote at 13:10 (PDT) on Wednesday: > Since HP is “deinvesting” in FOSSology, meaning that HP is to stop > funding us to work on it, my job (HP pays me to be the project leader) > is going to go away. ... I want to express my sadness that HP has eliminated your position. Of course, the wonderful thing about Free Software projects, particularly those under the GPL, is that they can have a new life after their primary sponsor moves on. As an employee of an organization that uses FOSSology regularly, I hope very much that FOSSology survives and finds a home that seeks to help FOSSology to thrive in the long term. First of all, I strongly encourage public discussion about the benefits and disadvantages to any non-profit home. In my nearly 20 years of experience in helping Free Software projects find non-profit homes, I find that such public discussion not only serves as an excellent record of why the project chose one non-profit home over another, but also can help other projects who later face the same decision. I realize Bob specifically asked about infrastructural resources, and I give "Conservancy's answer" on that at the bottom of this email. However, I would suggest the FOSSology community consider these other issues I list as points below, and I hope they're helpful. If not -- just skip down to the bottom where I answer Bob's specific questions. ;) * Not all non-profits have the same type of mission. If you're considering exclusively USA Open Source and Free Software organizations, there are roughly two types of possible homes: a trade association or a charity. A trade association serves a common business interest of its member companies (who are for-profit businesses, typically), while a charity serves the needs of the general public. This is a huge cultural difference and will set the tone for your project as either primarily for-profit-business-oriented or community-oriented. * The size of the non-profit you join matters a great deal. While a smaller non-profit might have fewer resources to put forward in the short term to the project, a smaller non-profit can typically "weather storms" better with the project, since a small non-profit is quite adept at doing more with fewer resources. The question to ask is whether FOSSology expects to grow quickly and have tons of finanical sponsorship, conferences (etc.) quickly. A large non-profit might be a better choice if such is expected to happen right away, as a large non-profit can ramp up faster. * Carefully review the agreements that the project would engage in to join the non-profit, and be sure you understand what the non-profit will and won't do for the project, how much of the funds' raised go specifically to the project, and how decision-making is handled between the non-profit and the project leadership. Conservancy's agreement templates, as well as FAQs about these issues, are publicly available here: https://sfconservancy.org/members/apply/ * Consider the fundraising goals of the project. Some non-profits might not be willing to assist with grass-roots and "small time" fundraising, while others typically focus on that. Depending on what annual budget you'd like to set for the project (and realistically believe you can raise,) you might chose one non-profit over another. For example, most Conservancy projects have about $5k-$30k/year in financial throughput, which is primarily spent for developer travel (Page 11 of https://sfconservancy.org/docs/conservancy_independent-audit_fy-2013.pdf gives a good overview of this). Conservancy has found that most projects -- even very large and important ones like Git and Samba -- simply don't need more revenue than that, since their developers are either volunteers, or are funded by their employers to contribute to the project. If the project does want to raise enough to fund developers directly, almost any Open Source or Free Software non-profit (Conservancy included) has experience doing that. Then, the question becomes how such arrangements are structured, and how to design reporting structures for the funded parties such that both the goals of the project and the non-profit are met. In my experience, such consideration generates lots of questions and discussion. If hiring developers is on your roadmap, explore these questions in depth with the non-profit before joining. Now, to Bob's question: Robert Gobeille wrote at 11:55 (PDT): > What the project needs is: > 1) A public FOSSology instance (if Matt at UNO wants to move this off > their server). > 2) A package build/test infrastructure. We build packages for > several distros and distro versions, and currently use VM’s at HP to > test them. > 3) A test instance that is fairly large to identify performance > problems at scale. Our current test machine has a 154 GB database > and a repository that is almost 2TB. You've listed a number of infrastructural needs of the project. In my experience (and unlike non-software development initiatives) software developers tend to prefer a good degree of control over their infrastructure. Conservancy has therefore focused on providing as much VPS "power" to our projects at no charge as we can, by negotiating agreements with many different VPS hosters, including Rackspace, Gandi, and OSUOSL (which you mentioned), to give gratis services to our projects. Conservancy focuses on getting raw computer power donations for our projects, as they usually know best how to use it. While nothing is long-term guaranteed, I suspect that given that nearly half of Conservancy's projects rely on donated VPS hosting, Conservancy will continue to find more VPS hosters and ensure such donations are consistently available to its projects. Thus, ultimately, the difficult issue in my experience is not CPU cycles or virtual instances, but maintaining the development infrastructure that runs on those boxes. This takes sysadmin work, and often projects simply find volunteers, and/or raise money to pay someone to do that work. Thus, if FOSSology *can* extract a contractual assurance of guaranteed sysadmin staff time devoted to the project from its non-profit home, then that non-profit may be the best choice (if these infrastructure needs listed above are indeed the top priority). However, I'm not aware of any non-profit that offers this, other than Apache Software Foundation (from which FOSSology is disqualified due to its license). Ironically, one of the reasons that Conservancy *didn't* design hosting infrastructure (such as the items you mention above) as part of its extensive service plan (which is listed at https://sfconservancy.org/members/services/ ) was due to the fact that more commonly, developers on ASF projects were complaining that the ASF infrastructure wasn't in the form the project wanted. (Cf: the "why can't I host my Apache project on Github?" arguments of a few years ago.) ASF's more recent relaxing of their infrastructure requirements on their projects indicates to me that Conservancy made the right choice. I don't think I could identify two Conservancy projects that have the exact same infrastructure setup, so imposing a specific one would have attempted to put many various non-round pegs all in round holes. Meanwhile, I suspect the issue of infrastructure concerns may simply be culture shock of the existing changes for the project. FOSSology has been cared for well by a for-profit company with substantial resources for a long time, and that era is now ending. No matter what happens to FOSSology next, the project will have to figure out how to survive with substantially fewer resources. I'd encourage you to pick a home that can help you do that, as I suspect it's impossible at this point to find a home that's prepared to dedicate long-term financial resources. For Conservancy's part, we can offer FOSSology on this front what we offer all our new member projects: access to the most experienced non-profit staff in the community to help you figure out how to raise as much money as you can for the project, and their advice on how to make best use of whatever resources can be collected through charitable contributions. I'd suspect that's just what FOSSology could use right now, but ultimately the project should settle into the home that's the best fit for your needs. I hope my lengthy email above has fully shared my expertise on the subject to assist the FOSSology community in making the right decision. Finally, I should note: Robert Gobeille wrote at 08:52 (PDT): > the Software Freedom Conservancy [has] expressed a willingness to > accept FOSSology. Conservancy's staff is very interested in FOSSology (since we're users and fans of the project), but FOSSology would still have to go through Conservancy's standard application process. However, Conservancy's Evaluation Committee meets monthly, and thus once the FOSSology community decides to apply to Conservancy officially, a final decision from Conservancy would likely be forthcoming in a month or less. -- Bradley M. Kuhn President & Distinguished Technologist of Software Freedom Conservancy _______________________________________________ fossology mailing list [email protected] http://lists.fossology.org/mailman/listinfo/fossology
