Okay for real. I’m all for debates about open source. But we have to start acknowledging that part of why it’s impactful *is because it’s free*.
Most OSS has no (economic) worth until many ppl use it to create *something else* valuable. If you built that same thing and tried to charge ppl initially, it wouldn’t work.
It seems obvious, but it is such a big part of why people don’t want to pay for it even after it accrues value.
Think about why you wouldn’t pay for twitter or facebook (no you actually wouldn’t), even though they have huge value to you (yes, they actually do).
I should’ve kept these connected to this thread.
If OSS wants to charge money, they really need to rethink what it is they’re supposedly selling. The thing you think is worth money isn’t.
It’s very hard to take something you gave away for free and convince people that same thing is now worth money.
It’s possible to upsell people by offering them *additional* value. Many people in OSS have been trying to crack that model.
But this thing where we think our free labor is now worth money because it’s got lots of users and it’s not fun anymore? Nah homie, just stop.
Here are some current thoughts on the problem. Keep in mind I would call myself a layperson on these issues.
With a few exceptions, open source projects don’t become worth money even after they become hugely popular.
The original example was about RxJS. Pretty popular. No one wants to pay. I tried to get my last company to pay for popular python projects. No deal.
Some exceptions in the OSS space that can command real dollars when popular: operating systems, programming languages, databases.
One theory is that instability and bugs at these lower levels are much more costly. People will pay to mitigate risk once they’ve invested heavily.
Another related theory is that once you’ve invested heavily at that lower level, switching costs are prohibitive. People will pay if they’re locked in and their investment is threatened.
These theories seem to hold up when you think about higher level open source libraries.
If there are bugs in app libraries, they suck, but most aren’t that much worse than the bugs internal teams produce.
App library bugs are also more accessible. Internal teams can usually fix them if they took the time (though they often don’t).
This is important because you perceive less value in a thing if you feel like you could do it yourself.
You can see this in the fact that many internal teams do fix bugs and just don’t contibute back upstream.
A lot of times internal teams *do* want to contribute bug fixes, but the project won’t give commit access and isn’t responsive to pull requests.
I suspect that popular OSS projects get offered labor contributions more than money. But they either don’t want them or can’t coordinate enough to accept them.
On a different note, there’s also a customer acquisition problem in OSS. It’s hard to charge *everyone* who uses a project.
Projects try to get *anyone* to pay, and they end up trying to charge a few people way more than is reasonable. Or get a few companies to chip in thousands or 10s of thousands.
If you could actually charge everyone who uses a popular project, you could charge less to each. Assuming they would actually pay and it was easy to do so.
There was a small wave of startups creating platforms to easily pay to support people’s previously free labor. Patreon emerged as the most popular.
Finally I want to come back to a more constructive take on my original rant. What are you selling? Remember it can’t be the same thing that used to be free.
OSS is trying lots of things here. Bounties for specific bugs, “influence” over the direction of the project. There are scattered success stories, but nothing sustainable.
Let’s set aside the ethical issue of how we should be paying more for the already accrued free value. It’s important but not realistic right now.
I think about the open source I use and what the pain points are. What would I pay to fix in the current dynamic?
I would pay for professional support. If your shit is broken, I want it fixed. I want an SLA, I want test coverage.
I would pay for access to the project to fix it myself. My engineers have autonomy to fix upstream. But you need to respond promptly to issues and PRs.
That doesn’t have to mean direct commit access. But my fixes need to land upstream in reasonable time so I can avoid maintaining internal forks.
This opens up a really big can of worms around architecture and roadmap though. These project maintainers are building their perfect sky castles, they don’t want our hacky fixes.
So many open source contributions get held up with “yeah but is this the right general solution?” “we’re planning a major rewrite of this”.
And my favorite. ”Can’t land this cause you didn’t write a test and update docs and pass the linter and...”
Nah. I’ll be keeping my coins.
Nah. I’ll be keeping my coins.
Lastly, a few final thoughts about potential payment structures.
I don’t want to pay every individual project. There needs to be some kind of aggregate umbrella.
Non-profit foundations sort of have this model. Except they don’t charge people directly by usage or service. They ask for donations. Nah my g.
Foundations don’t have the pressures of profitability, so they can also get bogged down and distracted. I don’t look to them for good customer service.
Here’s a completely different, radical idea. Union dues. Charge developers dues. Find some way for them to report on what they use and distribute the funds.
This has a few potentially nice qualities. Companies get access to open source through the devs they hire. If your devs aren’t union, you also lose OSS value.
Companies don’t have to worry about budgeting. They just pay salaries as usual. But they can also see how much money they contribute by counting their union members.
