- selling plugins/extension onto an open platform (basically what you call open core, not a fan of that term though). Works great for https://vcvrack.com/
- offering support for $$/hr. You'd think that would work better than it does, but even large companies try to find ways around getting your help. They'd prefer spending 100 of their developer hours than 10 of mine it seems.
- GPL/commercial barely works, but it's better than $0. Honestly I'd rather just make it fully commercial than this, because what's the point. I don't actually consider GPL to be fully open-source (an unpopular opinion.)
My stance is that you should plan ahead of time whether you want to monetize so you can determine if the time is worth it at all. I'm a firm believer in separating fun and work, so if you want to develop software for fun, keep it that way, and if for work, have a viable business plan. Otherwise bite the bullet and say that the project isn't meant to be exist.
I held the same opinion for a long time, I'm more on the fence now. It's an unpopular claim, but I don't think it's an unreasonable claim.
Implication being here that pure software product houses are not a healthy model. I understand this might be somewhat radical view, especially on HN where so much revolves around software products.
Giving away your work for one and all to come and take is, ultimately, unsustainable. All of your suggestions ultimately boil down to altruism by commercial entities, and that's also not sustainable. Companies care about legal compliance but don't want to contract with a million little entities to comply. That's why FOSS is so popular for companies and so terrible for maintainers. Companies get to build their business on this great software and completely ignore the externalities that it takes to get there.
If there were one central organization (or heck, a handful) where companies could register and pay for one license that entitles them to use any software licensed by that organization I think we would be far better off as a community.
There are already tons of source available models that tons of companies use. Nothing is stopping you from putting your code on Github right now with a proprietary license. Of course, even in that scenario most users prefer single-time purchases to subscription models.
If you want to go off and develop a new proprietary subscription license, then fine. But don't market it as a solution to OSS problems, that's deceptive.
> All of your suggestions ultimately boil down to altruism by commercial entities
Quite literally none of my suggestions boil down to that, I deliberately left donations off the list. In every single model a commercial entity pays for something they would not be able to get for free: advertising, permission to bypass GPL infection of their own code, support and access to developers, new features, or an enhanced proprietary version.
I'd suggest not to use the term when promoting your idea, it's kind of misleading and will provoke such discussion every time.
I don’t think this statement is accurate as there are decades old projects (linux/gnu, apache) to dispute your claim. I think this is possible because software actually is possible where a few individuals can have great reach with part time or employer funded time. And also because companies can benefit by funding and releasing activities that aren’t core to their business as either advertising or inclination by developers to be altruistic or other reasons (MSCode, Google stuff, etc).
I suspect this is because OSS software on Patreon usually doesn't come with many actual benefits. One of the models I'm interested in playing with in the future is having Patreon-specific issue trackers and access to separate development repositories.
In any case, you should not be trying to fund yourself purely through Patreon unless you have a good reason to believe you're going to be different.
Kickstarter and crowdfunding on the other hand has been real promising.[2][3][4][5]
I think there's strong evidence that OSS developers should consider charging upfront for feature development. It probably doesn't work for every type of project, but it seems to work for many types of projects.
[0]: https://liberapay.com/explore/teams
[1]: https://marijnhaverbeke.nl/fund/
[2]: https://www.kickstarter.com/projects/1681258897/its-magit-th...
[3]: https://www.kickstarter.com/projects/aiforeveryone/mycroft-m...
[4]: https://www.kickstarter.com/projects/johnonolan/ghost-just-a...
[5]: https://www.kickstarter.com/projects/mbs348/diaspora-the-per...
This is the other problem I ran into trying to figure out how to get Patreon set up - Github and Gitlab have pretty broad permissions.
It would be good to see people experiment with a wide variety of permissions surrounding issues, from access, to commenting, to creation, to upvoting. A lot of Open Source projects aren't even really looking for funding, but still want something in front of people who are just showing up to complain.
I don't know of a public issue tracker that supports that kind of granularity though (if you do, let me know), so the next best thing I could find is to make a private group on Gitlab for development branches and issue tracking, and then to mirror everything to a public repo for stable releases with issues disabled.
If anyone who works on Gitlab/Github is lurking around, it could potentially be a really cool differentiator to have those kinds of granular permissions.
Then it's not Open Source.
I agree that funding Open Source software is hard, and I want a solution as much as anybody: It's not even a theoretical problem for me, I am currently self employed and would really like to fund myself with just Open Source software.
What you're describing is a really bad idea though, it isn't actually FOSS. Heck, it's actually considerably worse than many of the other licensing models out there, because it basically boils down to being a subscription service. Many of us are involved in Open Source because we specifically hate the 'software rental' model.
Better ideas that I've seen people kicking around that I wouldn't immediately dismiss:
- putting a paywall in front of issue trackers
- dual licensing under the GPL and a proprietary paid license.
- requiring upfront payment for new features (ie Kickstarter/Patreon)
- open core models (ie Gitlab's strategy)
- advertising and sponsorships via READMEs and official sites.
Many of these ideas also have problems, but at least they don't require us to completely abandon FOSS licensing.