Comment by phendrenad2 2 months ago
Maybe open-source as the load-bearing infrastructure of the world will never be sustainable, and maybe that's okay. I think open-source is best when it's empowering people to modify and remix their software, and have free alternatives to expensive commercial programs. It seems like open-source has become so captured by corporate interests, that people's main motivation for contributing is to add it as a bullet-point on their resume.
It would probably be more sustainable if the companies that depend on 100s of FOSS OS'/libraries/applications/etc to generate billions of dollars in profit would contribute more significantly.
That's not how companies work.
Individuals have agency over discretionary spending, in other words you can wake up and decide to donate money, hire a OSS dev, and so on.
Because you personally have that agency, and because companies have lots of money, you project that agency onto companies.
But that's not how companies work.
People at companies have very little personal agency when it comes to spending money. Spending has to be approved, justified, and some value has to be received.
FOSS basically is (mostly) incompatible with this model. Some companies do pay staff to work on OSS but it's rare (and exclusively tech companies with a motivation.)
So while your statement is completely true, it's also not possible.
It's important to recognize that donations and corruption are indistinguishable, and company finances tend to be set up to avoid corruption.
> It's important to recognize that donations and corruption are indistinguishable, and company finances tend to be set up to avoid corruption.
Other than the humor in this sentence, I'm not sure why it would be limited to donations. They can hire devs to work on FOSS.
> People at companies have very little personal agency when it comes to spending money. Spending has to be approved, justified, and some value has to be received.
Approved by... people.
>> , I'm not sure why it would be limited to donations. They can hire devs to work on FOSS.
These FOSS we personally use. And FOSS we don't. Most of the FOSS we use already has plenty of paid devs. Think Linux, Firefox etc.
But what about small projects? How does that hiring conversation go? How do I explain my hiring request to my supervisor? Why am I spending 100k a year on a text editor? Why have I mandated we use my cousin's text editor and we're paying him to write it? When there hasn't been a significant update in 2 years? When we had to layoff staff to make mandated cost cutting?
> They can hire devs to work on FOSS.
And they do. Most big OSS projects have at least some of the maintainers and contributors be employees at a company where that OSS is used.
People make choices all the time. Choices that are overseen by management, and shareholders.
People can of course choose to spend their political capital, and discretionary budget on random OSS projects. Or they can choose to spend it on their project, their goals, the outcomes that make their walk in the company easier, the actions that will get them promoted and not fired.
These are choices people make, and frankly OSS funding delivers very very little bang for the buck. (On a buck by buck basis.)
> People at companies have very little personal agency when it comes to spending money. Spending has to be approved, justified, and some value has to be received.
seems to all hinge on the justification part, for which ppl that do it for the lulz don't really care
Absolutely, There’s hope in initiatives like https://osspledge.com/ & https://thanks.dev
Also https://polar.sh/, which lets users of the project "vote with their wallets" on which issues should be prioritized. You can see it in action in Starlette's GitHub issues, e.g. https://github.com/encode/starlette/issues/649
This would be nice, but since it hasn’t happened so far, hard to see why it would start happening.
No idea what the future will look like in general in 5, 10, or 20 years but I am reasonably confident that donations to OSS won’t be drastically more than they are now.
Fair enough, but having worked inside a lot of tech companies I think I also have a pretty good sense of why tech companies don’t monetarily contribute more: no incentive to do so and because OSS is often chosen specifically to avoid costs.
Hard to see why those things will stop being true.
Why would you say that? I believe the GP is correct. Unless something drastically changes, why would we expect companies to start getting generous, spending money they don't have to? Especially in the context of donations! If we're talking about a licensing shift that requires companies to pay, then sure. But for donations? I doubt it.
Actually no, historical data is the best indicator of future probability.
They do. A huge chunk of open source software is maintained by companies.
Companies that depend on FOSS would contribute if the license did not explicitly tell them that they don't have to.
MPLv2/EUPL come to mind: they are compatible with proprietary products, but they make it mandatory to distribute changes/extensions of the library, not the whole product.
FOSS authors have a responsibility when they choose a permissive license.
> Companies that depend on FOSS would contribute if the license did not explicitly tell them that they don't have to.
No they won't. They'll only contribute if they're required to, or if doing so will be beneficial to them, and they'll do that regardless of whether the license says they have to or not.
When I've worked at companies that use FOSS, and have needed to modify those sources, I'll contribute back (regardless of license) if I think that change is likely to be accepted upstream, because I'd rather not have to maintain a fork. This would fall under "contribute if doing so will be beneficial to them".
At any rate, no FOSS license (that I'm aware of, or is in wide use) requires users to contribute. At most, they require that changes be made available. There's nothing that says the changes need to be submitted (or accepted) upstream. Often getting a change into a state where it would be accepted upstream is a significant amount of work beyond what the company has already done for their own purposes, so they don't bother.
> No they won't. They'll only contribute if they're required to
Did you actually read my comment before you answered? Because I said that copyleft licenses "make it mandatory to distribute changes/extensions", which means that companies are required to contribute if the license is reciprocal.
> At any rate, no FOSS license (that I'm aware of, or is in wide use) requires users to contribute. At most, they require that changes be made available.
Making changes available is a form of contribution. If you work on a proof of concept for a month and at the end your company decides not to use it in a product (thanks to the learning from your work), do you say that you did not contribute, so you should not be paid?
Feels like you're being pedantic just for the sake of the argument.
Permissive licenses come with strings attached (that most companies ignore): attribution.
With copyleft licenses, nothing say that you have to get your changes upstream: you just have to distribute them to the users. It's not a whole lot more complicated than attribution: set up a repo and put your fork there publicly.
I am not here to shill for Google, but they publish a staggering amount of liberally licenced software. We can much less of that about Microsoft, Apple, and (my personal most dreaded for open source) Amazon.
Also, I stand by my previous comments from other similar discussions: Almost all big corps use Redhat. They are indirectly funding open source. Redhat probably employs more programmers that contribute to a base Linux install than any other company on the planet. (Yeah, I know they were bought by IBM, which gets no love around here.)
> you're not a good person, you don't fool me. Fund open source, it would support young people who were just like you were
Or maybe he knows he's not a good person and has no intention of multiplying people who are just like he was, because he knows people as himself are bad and the world is better without them.
> It's probably too much to ask corporations to dump money into it as it would not be a legitimate business expense.
Um, excuse me?
Ok, let's suppose you've got a product that depends on open source project X. For simplicity let's say it's a direct dependency, though I think everything here applies to indirect ones as well.
Let's consider the options.
Option 1: never pay a dime for it. This works in so far as someone else picks up the bill. So really there are two sub-cases:
Option 1(a): the project is successful enough that it's self-sustaining. What this really means is that someone else (or multiple someone elses) picked up the bill. Congrats, you lucked out.
Option 1(b): the project is insufficiently funded and either dies or has a major security breach. Now you end up paying either for the security breach fallout and/or to replace the component, possibly on short notice, with something else. Or you maintain it yourself and start paying that cost, again possibly on short notice.
Is that really worth it? Do you think so? I'm betting all those costs are higher than it would have cost to maintain it in the first place. Because anything you do in an emergency is more expensive, and you're paying the cost of losing all the context in the development of the project itself (if someone leaves before you start maintaining it).
Option 2: pay for the software in the first place, making the cost predictable and avoiding a low-probability high-impact failure mode. Honestly, given all the risk management companies do, this seems worth it to me. At least if the dependency is critical enough.
Obviously you won't do this with any random open source project. But that's sort of the point: companies are making economic decisions all the time about what they really care about. If they aren't paying, that means they're happy with the inverse lottery[1] of the failed open source project model.
[1]: An inverse lottery is one where most of the time you get nothing, but rarely you lose big.
There are other options.
Option 2: Fork the code and do whatever they want with it.
Option 3: Directly employ open source project maintainers instead of donating to the project. They can exert at least some control over project direction that way.
Most enterprises don't even have a budget line item for open source project donations.
Companies will continue to take advantage of free and cheap labor just as long as those people continue to serve.
Perhaps after this generation of LBIP moves on, the next wave won’t be so generous.
Also, open source doesn’t have to mean free labor. One could be paid a wage to work on open source.
I think there might be a little different.
New developers with higher level tools don't seem to go low level enough.
And as a result some very talented developers are caught up in the crosswinds of complexity to make things like React deliver html/css, or get people to click on ads, and that same intellect and talent could definitely go towards the meaningful open-source that contributed to them getting access to a lot of opportunities.
The beauty of open-source is that it doesn’t matter what people say, or do, or decide.
The rate of new contributions may be dependant on these factors, but not for what’s already out there.
I don't think there's any other sustainable format to make software that becomes load-bearing infrastructure. Open source is the only player existing.
What isn't sustainable is just the model where some random person takes a huge part of the responsibility for himself without any structure around it. We need different organizations.
Privately owned infrastructure is perfectly sustainable in other fields, and so is publicly funded infrastructure. I think that, if necessary, either of those could be a viable replacement for open source being the infrastructure for tech. Obviously each comes with its own pros and cons compared to open source, but they still ought to be viable.
I haven't yet seen the evidence that it's unsustainable. People are still doing it. It seems to be working. There's no reason to believe people are going to stop doing it or that it's going to stop working.
It might be important to you to get paid for your work. Okay, charge for your work then! That was always allowed. There are many people who are okay with doing certain work for free. Maybe they just want to put it on their resume, or they already have enough money, or they just really like that work. Open source is working fine.