Sam Hartman <hartm...@debian.org> writes: > I'd like to ask you to look at the elephant in the room. This > conversation came up specifically because we were talking about an > organization loosely associated with Debian paying some Debian > developers.
> Yet, you didn't consider any of the middle options in your > analysis--only the option of staying effectively all volunteer or going > all industrial consortium. Yes, this is a very good point and I didn't address that. > do you have any thoughts on the middle options like having industrial > consortiums that we work with so that Debian developers who are > comfortable in that model can pursue that? Personally, I'm pessimistic that this would be stable in practice, but it's possible that I'm borrowing trouble where none need exist. The specific outcome that I would expect from that sort of hybrid model is that the folks with the stable funding stream and full-time engagement in Debian would come to dominate the project, not out of any malice or nefarious intent, but simply because they have more resources and engagement and Debian is effectively their job. In other words, were the consortium to be successful at the level required to close the resourcing gap, I would expect the rest of Debian to effectively collapse into it, leaving an industry consortium core surrounded by a halo of volunteers. The incentives that I can see would seem to push in that direction. For example, it's difficult to do project management when some folks on the project can put steady and predictable effort into it and other folks on the project are volunteers and by the nature of volunteer work are often not going to be able to promise timelines or level of effort for large projects [1]. The natural tendency is to shift the most critical work onto the most reliable people. This is a perfectly fine model for an effective non-profit, but it doesn't look very much like Debian does today. Note that I wouldn't expect this to happen with more modest engagements with increased funding, such as what Raphael is proposing. It would require very significant increased funding and corresponding staff. But neither would I expect more modest efforts to have much long-term effect on the overall resourcing gap, unless we coupled that with decreasing Debian's aspirational scope. That said, maybe I'm completely wrong. I am certainly not an expert in non-profits, and if someone has experience or research on volunteer-run non-profits with paid staff that have avoided this outcome, I'd be curious to read it. [1] I want to acknowledge that some volunteers do promise timelines and level of effort and fulfill those promises. Indeed we ask this of the DPL candidates every year. But I think this is unusual for work that requires more advance scheduling than the sort of shift sign-ups that you'd see at, say, volunteer food banks. Most volunteers will not be able to do this for the length of a major Debian project, hence the regular questions to DPL candidates about the viability of the time committment. I also have personal qualms about whether asking for reliable commitment without compensation from people who live in an otherwise capitalist society is entirely ethical. -- Russ Allbery (r...@debian.org) <https://www.eyrie.org/~eagle/>