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.
The word underpaid seems too friendly. In the case of the open source developers I know, they don't get any money
I wish there was a better solution to this crisis. There should be a corporate tax on software companies solely to fund open source efforts or something of the like.
Some open source projects have merged into one umbrella organization to better solicit sponsors but I don't think that will lead to anything fruitful, more so lead to a thing like Canonical where there's a profit motivation to the project
For the only significant upstream OSS contributions I've made I was being paid a pretty good tech salary to make them. The first major contribution happened because the organization I worked for was still in an incubation mode and wasn't on the hook to show any P&L results. The second and third major contributions happened because they shipped on a proprietary platform that was built on top of an OSS release, and the company didn't want to maintain those parts in the proprietary layer.
After I left the first company they provided zero maintenance support for the major upstream feature, and I won't maintain it on an individual basis because life moves far too quickly for that. Occasionally I'll happen to run across exasperated posts along the lines of, "Help this is broken for me! Why is my patch being ignored? Is this even maintained any more??"
Chances are you only care because the company you work for wants it done so they can make some money off a product they're building with it. Half an hour dealing with some random patch and/or bug report for an upstream thing I did 10 years ago is half an hour less I have to spend with my daughter before she leaves for college. Nope, fuck you and your bug report.
No, you can't even pay me to fix it. I have a job, and it's already taking more of my limited time on this earth than I want it to take. The email address I used to submit those patches was deactivated when I left that company, and I ain't handing out my current one.
For any minor contributions I made I had to convince the powers that be that there was no real IP or competitive value in maintaining the patch ourselves, and they after much hand-wringing and delay they finally let me push a few small patches up to the maintainers. In hindsight that ended up being more trouble than it was worth because I wouldn't have been the one needing to maintain that patch in the long run. I should have just submitted it to the company-internal repo and let the next guy deal with maintaining that patch.
In spite of all that, my contributions are still there. They still provide tons of value to both individuals and to companies that mooch off of them. Somehow at some point someone somewhere steps up after enough time has passed to review and merge a patch or fix a bug. Or the bug report just withers and dies, and the world keeps turning.
Creating value doesn't automatically get you any money no matter how much value it is. You might get some by adjacency. Businesses capture and raid value.
I think it’d be interesting to see what happens in a few decades if the population of open source maintainers continue to dwindle. Will companies step up to maintain libraries in a sort of cooperative manner? Will everything become closed source and bespoke?
Now this probably won’t happen but it’s still interesting to think about.
As it became common during the last years, public domain and shareware return under other marketing names like open core.
To answer the question of the future of open source, I think it's important to consider why open source has existed for so long. Is it a matter of cost to develop closed-source systems? Or is it about the quality and efficiency enabled by open source development?
From my experience it’s about cost and efficiency. By cost I mean my cost to them for my time coding. Customers don’t factor in the hidden costs of open source.
I preferred developing minimal-dependency software but my customers demand fast good-enough results. The only way to deliver that is to glue together open source dependencies.
Are there even unions for open source maintainers/contributors?
Those unions would bargain with...whom, exactly?
Companies that avoid the AGPL, but are open to suggestions ?
I’m a young(-ish) dev who used to care a lot about open source but never managed to break into a community. In recent years my view of the whole thing has soured a lot. There seems to be few compelling incentives to actually develop or participate in open-source software.
Open source first felt to me like a way to give power back to ordinary people, and it still is, but it seems like those who get the most benefit from free labor are large corporations. Open source feels increasingly corporate and companies like Microsoft dominate and reap enormous benefits. I’ll work for Joe Neighbor for free but not Sataya Nadella.
Open source was always political but in 2010 (around when I started getting into Linux) it felt like dumb arguments over things like programming languages or “the UNIX philosophy”. Now it feels like a vicious Red vs Blue culture war where not picking a side is just as bad as picking the other side.
Contributing to open source is a thankless job and even if your project is really good, most people won’t care and the few that do might still treat you like crap. I’ve submitted a handful of pull requests and I’ve already run into the classic “Your patch works and provides a feature some people might like, but I don’t like it, go away.”
I’ve donated plenty to organizations like Mozilla, Wikipedia, and GNOME. I then email them with my opinions on what they’re doing. In nearly every case not only am I ignored completely, I see those projects (Mozilla especially) continue to drift in a direction that I disagree with. So, I stopped donating.
For me, the Linux kernel is probably one of the few big open source projects where 1) the project is technically interesting enough that I would learn a lot by contributing, 2) It seems like politics and infighting is kept under control, 3) it actually seems possible to get a patch in while having a 9-5, 4) I use the product myself every day and enjoy it, and 5) the technical direction feels positive in that it is getting regular features & bug fixes that I like
> but it seems like those who get the most benefit from free labor are large corporations.
I feel one thing to remember on this front is that large corporations will ALWAYS get the most benefit out of open source / “free labor”, simply because they have the ability to bring massive amounts of resources to bear on using that open source product towards their own ends. Consider the world of hardware emulation. Sure the community has benefited massively from the efforts poured into reverse engineering and understanding old systems and games and preserving what was there. And the big corporations reap huge benefits in the form of continuing nostalgia, awareness of their back catalog of IP, test markets and information about the viability of re-releases and in some cases the licensing (or outright theft) of emulators and emulator code for selling their own retro consoles.
Burn out is absolutely a concern, and the approach of some open source devs (like IIRC the curl dev) of essentially “f you, pay me” to support requests is probably an important thing to have. But for me as an individual, the fact that Atari’s current owners have reaped massive rewards from the fact that the emulation scene keeps their brand alive means nothing to me. I’d rather have the world we have were things are open and the community is there, than one were emulation is closed and insular and getting into it is even harder than it already is just to keep Atari from “winning” the most. And selfishly part of that is because Atari winning also benefits me. Their re-releases and re-masters and dumping of money into manufacturing by hardware retro clones puts money into the market, gives new hardware to tinker with or build on. And their dependence on the open efforts of the community also means bending that hardware to my own needs is much easier than if they had for example just rolled out a bunch of new proprietary SoCs to replace the old hardware with a single blob chip
> I’ve submitted a handful of pull requests and I’ve already run into the classic “Your patch works and provides a feature some people might like, but I don’t like it, go away.”
This might be because they’re not interested in maintaining it for future iterations.
As someone born in the 70s, I don't have high hopes for the Linux kernel after our generation is gone.
It is hard to keep something rolling after the founders, that managed to steer a project under their ideals are no longer around to steer the boat into the right direction.
Something else will eventually take Linux's place, in some form, it might even be a Linux based like Android, WebOS, ChromeOS.
Don't worry, the generation before you didn't have high hopes for your generation too. We are nothing special, the magic will keep on going.
> Something else will eventually take Linux's place, in some form, it might even be a Linux based like Android, WebOS, ChromeOS.
So Linux won't be the same because it might start being marketed under a different brand name?
Anyone that has developed for Android, WebOS, ChromeOS is fully aware that the Linux kernel is an implementation detail, full of features not available upstream, like drivers written in Rust, which remains to be seen if it will ever have them.
Also given that the kernel is not exposed to userspace, other than to device vendors, it could be replaced by something else with very little impact to userspace, other than whatever people are doing in their rooted devices or via ADB shell, both meanigless for common consumers.
> it could be replaced by something else
It is not really Linux-based if something else. Assuming the Linux-based property remains, though, isn't that still Linux, even if it ultimately is taken down roads Linus would have never approved of?
As I have said before, here and elsewhere:
I would love to work on open-source full time, but I exist in a capitalistic hellhole which employs violently coercive methods - the constant threat of homelessness, destitution, and even death - to ensure I remain “sufficiently profitable” to someone else who already has more wealth than I could ever spend in a dozen lifetimes.
If UBI is ever implemented in Canada to an effective degree, you can bet both your left titties that I would be able to afford the time to contribute. But now? I’m just struggling to tread water.
Rewrite of https://news.ycombinator.com/item?id=41591692
To summarize the programming toolsets used in open source are not the kind that hobbyists and volunteers should be using, that is why I mentioned the use of Lisp and other higher level languages compared to C and C++. If you are going to write non-trivial programs and utilities in C and C++ then you should absolutely be paid for it.
Most of the people burned out developing for Linux should focus on a cut down system developed for end users and small businesses in mind.
You clearly haven't heard those jokes about C and C++ have you?
You are talking about unpaid people using the wrong languages to do difficult things and that never works.
http://harmful.cat-v.org/software/c++/I_did_it_for_you_all
The Unix Haters Handbook
http://harmful.cat-v.org/software/c++/linus
http://harmful.cat-v.org/software/c++/
Not only that the industry ignores the foundations and principles of correct and sound programming in low level programming languages like C and C++ which makes things even worse.
Going back to the start Linux did not start of with an intention to be an operating system developed for enterprise users.
Commercial interests took it over and decided to go with it. The simple truth is unpaid hobbyists and hackers should not be involved in the development of enterprise operating systems. Windows NT was not developed by unpaid hackers. Why should that apply to Linux? Unix was developed by corporate employees, and it is only when the GNU project decided to replicate the toolsets it used that Linus decided to build a kernel of his own, after which corporations jumped on it because of the GPL.
Those days are long gone and hobbyists and part timers shouldn't be really involved. It is OK to write and develop programs to gain a sense of achievement and pride from them, but when the need to maintain it kicks in that is when it all starts to south.
Their involvement should on simple utilities for end users. That is what free software development which is not well remunerated should focus on.
Solution is to stop using languages low level like C except where it matters, and improve the UI capabilities of higher level like Lisp.
Most of the problems in free software are rooted in the use of C and C++.
When you are doing things for free you shouldn't be using slow, difficult tools.
What on earth are you talking about? The problem is corporate interests thinking that 'open' means 'free labor'
The choice of language has absolutely nothing to do with it. Not in the slightest.
This cargo cult nonsense does not make you smart or cool. Try some critical thinking next time instead of regurgitating catch phrases you haven't even bothered to analyze.
Most of the people burned out developing for Linux should focus on a cut down system developed for end users and small businesses in mind.
You clearly haven't heard those jokes about C and C++ have you?
You are talking about unpaid people using the wrong languages to do difficult things and that never works.
http://harmful.cat-v.org/software/c++/I_did_it_for_you_all
The Unix Haters Handbook
http://harmful.cat-v.org/software/c++/linus
http://harmful.cat-v.org/software/c++/
Not only that the industry ignores the foundations and principles of correct and sound programming in low level programming languages like C and C++ which makes things even worse.
Going back to the start Linux did not start of with an intention to be an operating system developed for enterprise users.
Commercial interests took it over and decided to go with it. The simple truth unpaid hobbyists and hackers should not be involved in the development enterprise operating systems. Windows NT was not developed by unpaid hackers. Why should that apply to Linux? Unix was developed by corporate employees, and it is only when the GNU project decided to replicate the toolsets it used that Linus decided to build a kernel of his own, after which corporations jumped on it because of the GPL.
Those days are long gone and hobbyists and part timers shouldn't be really involved. It is OK to write and develop programs gain a sense of achievement and pride from them, but when the need to maintain it kicks in that is when it all starts to south.
As unpopular as this opinion may be:
AI solves this problem long term
We need way more code than anyone is willing to pay for.
OSS has been a great foundation for GenAI progress in coding.