This makes me think. Is there a service that handles crowd-sourcing and processing payments a bit like Kickstarter but on a smaller scale? Say, if someone proposes a feature and the developer says "I'll do it for $300", users could donate until the goal is reached. With the promise of payment, the developer can go ahead and spend the time/effort, and everybody wins. But setting up a Kickstarter, IndieGogo, etc is too big of a deal for a $300 feature/bug fix.
GitHub Sponsors (https://github.com/sponsors) already exists, but I feel that they missed the point. As a user, I don't want to throw money at a project like it's a charity, I want to throw money so issues will be solved. As a developer, I don't want to accept unconditional donations (and all the unwritten, assumed responsibilities it comes with), I want to be funded in a concrete way so I can work on the issues users want so much they'd pay for it.
I think bounties create an unhealthy incentive structure. The kind of thinking you expressed in your post that payments shouldn't be seen as "charity" leads to people trying to not "waste" their bounties.
They'll add it to issues the maintainers won't do otherwise, which in turn leads maintainers to hold the code hostage, not doing something they'd have done had there been no bounty structure, hoping that someone posts a bounty for it. Of course a maintainer is one of the world experts in that codebase so knows really well which things can or have to be done. So that behaviour will lead to maintainers not addressing issues they know will annoy users. It leads to annoyance driven development.
Regular payments are far better IMO for projects long term because they allow maintainers to actually maintain the software instead of hunting for bug bounties. They give more security to maintainers (rent is pretty much a constant for example) and less overhad for users (unless they want to save money and use it most efficiently, but IMO that's not what donations should be about). That being said, bounties aren't bad overall, they are definitely great ideas if used here and there, but funding of open source projects shouldn't be bounty-driven.
I can't speak for others, but many of my company's open-source projects are not open-contribution (meaning I don't accept PRs unless I hire you). Features are added because funding comes from somewhere else, but it would be nice to have a second option for users to fund a feature directly.
If you're asking what we benefit from doing this, not much. We could have released proprietary software, but as users of software ourselves, we enjoy having these freedoms, so we don't want to restrict users of the software we write.
Mate, I haven't felt as happy reading a comment on this website as when I read this comment of yours. Keep up the excellent work and I hope you find a lot of success!
There's room for both models. Some people operate better under one incentive structure and others under a different. There's probably one structure that is better for more people than others, but maybe we should normalize both and encourage exploration.
Agreed, if bounties conventionally worked like that I would feel incentivized to not contribute to OS projects unless there was a bounty. I dunno, maybe it'll still be a net benefit with an influx of bounty hunters. Hard to say really.
Depends. If the bounties are created by the maintainers themselves, funded from something like Open Collective, I think it can be a healthy way to get the community involved and help maintainers get things done that they wouldn't be able to work on themselves. If bounties are only exclusively created and funded by the community, I can see how this could be a problem.
> As a user, I don't want to throw money at a project
To be fair, GitHub Sponsors doesn't even allow you to sponsor a specific project, but only users or organizations.
I think BountySource (https://www.bountysource.com/) was the first player on the "Bounties for GitHub issues" side of things. Long time ago I actually came across a issue with a bounty though, so not sure it's actually working at the moment.
https://gitcoin.co/ is another attempt at aligning the incentives, but it's heavily involved in the cryptocurrency ecosystem and also been spamming projects in order to be setup on their platform, so I've been kind of avoiding it.
I think Open Collective seems to be the best platform for supporting projects with funding, but they don't have any "per issue" funding as far as I know. But it's a open project, someone should propose that feature!
It's a tool where users post bounties for features. Usually the owner of an open source project would set up a bounty and say "$xxx to fix this" and open source developers can apply to work on the problem, once its approved they get the funds.
But, if community members really value the project they can chip in more funding to the issue.
Currently the system is set up to use Ethereum and DAI (an ethereum USD stablecoin) to handle transactions, and Gitcoin's smart contract acts as the escrow account for the work.
$300 just buys you a half day of developer time, I suspect the sponsor wants a proper professional job done that works with and is tested with a range of hardware.
On my own open-source projects, there are hundreds or thousands of issues I'd solve in half a day if I got $300/ea. That'd keep me fully employed for months or years! And I actually have the users that would donate 6 figures / year cumulatively to make that a reality. But orchestrating the money would be a nightmare without a service that is simple to use from the users' perspective. BountySource is alright, but I'm not impressed. It's still a bit difficult to use and navigate compared to the average checkout system, or Kickstarter itself.
Kickstarter is more suitable for larger projects.Github could work with a little bit creativity on top, however that's the place for technical people. A platform,where a feature could be requested and payouts are as simple as one click would be fantastic. I think the platform part would be the easiest to implement.To ensure the whole thing works well for devs and clients well+ doesn't attract bad actors,etc. would be more difficult to get right.
With 300 for a half-day, it would be 600 for a full day and 12000 for 20 days of work.
In most parts of the world, that's an extremely high level salary and could easily fund most peoples month, granted they live outside of Silicon Valley.
I'm sure many people would be more than happy to receive 300 or much less for half a day of work.
The going rate for a developer is about £300-600 a day in North Europe for relatively common skillsets such as C#. It isn't just silicon valley.
The $10,000 bounty is about right for this sort of work. In fact it might be a little low, but I have no idea how OBS works internally so I could be completely wrong.
I don't think it's a problem of OBS, but the problem of writing a dummy v4l-device it seems. which IS more complicated! (though I wonder what it's used for)
Looks like a direct clone of BountySource (in the same way that GitLab cloned GitHub), but this looks a lot better and solves some of the UI problems of BountySource.
To test this, I signed up and pretended to contribute to a random project. "Where's the place to enter my credit card info to fund a grant??" 99% of users would leave at that point.
Yes, I'm aware this is based on a cryptocurrency system. But I would never want to force my users to navigate that mess in order to simply give me a few bucks. I'm just joking around above, but I don't think Gitcoin gets close to solving the problem.
GitHub Sponsors (https://github.com/sponsors) already exists, but I feel that they missed the point. As a user, I don't want to throw money at a project like it's a charity, I want to throw money so issues will be solved. As a developer, I don't want to accept unconditional donations (and all the unwritten, assumed responsibilities it comes with), I want to be funded in a concrete way so I can work on the issues users want so much they'd pay for it.