Hacker News Reader

Comment by vasco | Hacker News Reader

I have a difficult time thinking through how to justify donating to some of these niche libraries. It's not wrong to make a living out of open source of course, but it is slightly strange sometimes to find projects that are super core to internet infra running on shoe strings but then some random dude in SF makes 100k+ / year on some minor library that is syntactic sugar on other stuff. This is not about this person or this project but more of a recent realization. In my company when developers brought up supporting open source, the thing they first remembered is whatever niche npm library they remember seeing a begging readme, when we use so much more important things that could use the help.

For example git-scm belongs to a foundation, which you can donate to and ensures transparency and some structure around spreading contributions. "Donate to some dude" is quite strange to me in comparison.

View on HN

It's disheartening to have the top comment on this discussion be a discouragement from individual contributions because the solution is imperfect.

Let me suggest a thought pattern which might ease your difficulty.

It's hard to imagine a community better positioned than the devs themselves to evolve better solutions to these problems than they are. So the "ways to fund FOSS" they advocate, by for example asking represent some flavor of best solution we've come up with, despite all their flaws.

So do what they ask, to the extent you find their packages useful.

Magit isn't exactly a library (although I think it is also that). It's more of an application that happens to be implemented within emacs. There's a number of commercial git clients, so paying for Magit doesn't seem unreasonable to me.

Magit is by far the best interface for using git and managing git data structures that I have ever used - and I am a vim user! Almost every emacs user I know uses or recommends Magit. It may seem like a "niche library" but in some areas of software development, it's absolutely well-known and considered the gold standard.

You use a similar lib on vim ?

No, I use emacs with Magit exclusively to do any interactive or complex git operations and then return to vim. I think Magit is that good.

should be a quote on magit.vc

The lesson there should be that projects get funding when they have advocates. The importance of a project to the world means nothing: it’s the importance of the project to people willing to fund it. Your co-workers should feel empowered by this: if some inconsequential software can make money, your co-worker’s advocacy can surely help a piece of business-important software make money :)

Magit's call to donate addresses itself to a completely different demographic than the companies that should donate to "super core internet infra" projects.

Individual developers might not care about their company that uses libcurl in 95% of their applications, but might get enough benefit from magit to make it worth their while to donate to it.

I looked at the project's sponsorship on GitHub, Patreon and OpenCollective. Guessing the average GitHub sponsor donates $3/mo, he's at ~$1,360/mo before fees and taxes. If the GitHub average is $6/mo, that's ~$2,140. The GitHub average would have to be $30/mo to hit $100k/yr.

(Wasn't sure if you were suggesting he's earning 100+k/year)

This argument is not new: it happens also with regular charities. You could donate to the charity with the highest impact (measured by QALYs per dollar or whatever) but a lot of people donate less efficiently. Sometimes they even get criticised for it. But maybe an inefficient donation is still better than no donation, and should be lauded to encourage more people to donate, however efficiently?

At first I was going to suggest, that someone set up an effective-altruism-for-open-source type of project, that ranks open source projects by how efficiently they can use donated money.

But then I realised also that would fall for the same trap. Why donate to give libcurl higher revenue when you can donate to save people from malnutrition?

(On the other hand, you could argue that the productivity gains afforded by libcurl would allow us (humans) to allocate more resources to victims of malnutrition. So what do I know.)

Reply on news.ycombinator.com