You use more open source software than you think
github.com
github.com
Back then, there was this enormous revolutionary fervor around free software, and it was intensely idealogical. You used the GPL because you wanted to fuck Microsoft and the entire concept of proprietary code. Linux was punk, not a good “solution.” And every time someone hacked up a free version of something, we all rejoiced, not because it “didn’t cost anything,” but because it was another way to make the point that the entire concept of making money off binary blobs was inherently evil. Install fests! Bring us that old 486, and we’ll make it YOURS! What a party.
Were we hopelessly naive? Maybe. Probably. But I find it depressing to read threads like this. It’s like everybody is just doing some kind of cost benefit analysis, but nobody is asking serious questions about whether the world we built is actually the world we want.
I’ll say it: If many of you had walked into some scrappy LUG in 1996, you’d have been laughed out of there as some kind of tool.
Are you a tool? I don’t know. But the resignation still bums me out. Y’all want to “support OSS devs,” not throw a sledgehammer at the screen.
I think I would describe that as spite. Is it not? By long term strategy I mean strategy for building things in general. Like, I get it, I don't particularly enjoy using certain Microsoft products either. But once you get that off your company's machines, then what? There has to be somewhere to go from there.
So really, you think “building things in general” required . . . what? Calm acceptance of restrictive licensing and software patents? Or maybe we should have been nicer about it? Again, it’s like “building things” must be properly done the way we currently do it. With lots of lawyers and lots of EULAs.
This strikes me as a profoundly conservative position. How rude those Jacobins were! If only they were more polite. If only they had recognized that we live in the best of all possible worlds . . .
I miss those times, and places like HN are far from it. Join mastodon if you want to breathe a similar atmosphere again.
> Were we hopelessly naive?
No, Linux won hands down. Also, a lot of people warned about the coopting of FOSS from software freedom into unpaid labor. We saw it coming and couldn't stop it.
Also, what were people naive about?
I think what was naive was building OSS around service, support and consulting. Instead it could have been built around mutual benefit as a governing principle.
Stallman (and others e.g. Lessig) had clear visions of where they wanted to be but were thoroughly stuck in an end-of-history mindset - "we'll do the work to get this small part of society from A to B, and then we'll be at B." Well, no, then someone is going to drag parts back to A, and another is going to drag a few over to C, and after a few steps of this your rhetoric around digital freedom is going to be half fueling the propaganda of the largest, most powerful corporations in American history, and half the weak intellectual justification for a right-wing populist riot.
https://thebaffler.com/latest/capitalisms-new-clothes-morozo...
https://www.e-flux.com/journal/70/60567/the-vectoralist-clas...
Free software was built around the principles of mutual benefit. OSS was coined as a "corporate friendly" label, and it is from that culture that the whole service and consulting thing comes. Service, support and consulting is how you make money from it. It wasn't naive at all - it was a cynical co-opting.
(And if you don't see the problem with the maintainers of a piece of software to also charge for support, may I direct your attention to The Shirky Principle [0] - "Institutions will try to preserve the problem to which they are the solution". Software developed under this model will, over time, become byzantine, baroque, and frustrating - because that brings in the consulting cash.)
[0] https://en.wikipedia.org/wiki/Clay_Shirky#Shirky_principle
To geezers like us, the elephant in the room is this article being hosted by m$. And honestly, I still can't fully trust that this isn't the Embrace phase of a long game. For the ultimate cognitive dissonance, I'd even work there because I like how they treat their employees.
On the other hand... whose screen should we be throwing sledgehammers through? The internet enabled an absolute explosion of tech, but these days it's nearing impossible to live without a smartphone. Governments tacitly support Apple and Google's monopsony to distribute apps to access government services.
Back when it was just nerds on the net, we controlled our digital lives. But today, the rug has been swept away and... the impact of our sledgehammer only isolates us from the non-nerd friends and family.
And I'll say it... if I brought a modern phone to a LUG in '96 with Candy Crush on it, y'all wouldn't get anything done for a month. It's not just resignation. Contentedness is a problem too. So I support the OSS devs I can.
Me too - it's as if the entire scene has "sold out". I'm younger than you are, but I entered the Free Software(TM) scene as a zealot and a true believer - you probably had to back in the day because the overall software quality of the entire ecosystem wasn't close to parity to proprietary solutions (some parts of it were clearly superior, e.g. Apache vs. IIS). I think this high bar resulted in a self-selected group being involved in LUGs and similar groups. These days, the bar is much lower, thanks to Github and the large number of high-quality FOSS projects that are market leaders - far better than proprietary equivalents.
My theory is that with this self-selecting filter gone, a lot of developers are now part of the community for self-serving reasons, so their attitude is ...mercenary. Its now all about "How does this project help me and make my life convenient as a developer" rather than "How does this project help software users like me": it's a subtle difference, but with far-reaching consequences.
I don't think we were naive - we thought we had won when corporations defected, but our victory was subverted and it feels more and more like most corporations are a fifth column. "Open Source Software" is slowly suffocating "Free Software".
I don't think this is a new thing, or something that is necessarily correlated. I've seen plenty of old open source projects that are totally obtuse and obscure to everyone but the developer -- having that attitude is not mutually exclusive with being a FOSS enthusiast. Probably more people are developing this stuff for money now, but that's a different thing from the "selling out" you were talking about.
Github didn't change much here if you ask me, in the old days it was Sourceforge, and you can still find a lot of those obscure projects still hosted there.
Agreed - it's not new, but its certainly related as it is at the root of the philosophical differences between BSD-style licenses and GPL, and the respective licenses popularity over time support my thesis.
Not only are more people developing stuff for money, there are fewer who are developing it to scratch an itch and giving it to the world for free. For a large number of people (too large, if you ask me), the process is now:
1. Author popular open source project
2. ???
3. Profit
Free software used to be a way to avoid exploitation, now developers (and corporations!) are tripping over themselves to exploit the popularity of their FOSS projects - that's what feels like "selling out", because money/control is the goal now.
> Github didn't change much here if you ask me, in the old days it was Sourceforge, and you can still find a lot of those obscure projects still hosted there.
Git, Github and turn-key CI lowered the bar by a lot! Forking an existing project or getting a new one off the ground is much easier now, which is a blessing and a curse - but mostly a blessing. Building and packaging software by hand (ok, using Makefiles) would filter out all, but the very committed - and that was what you had to do in the Sourceforge era. I didn't say it was impossible in the past, only that the technical (and commitment) bars were much higher
I feel like all of this is trying to escape economic realities that someone has to do the work to make those binary blobs and that person deserves some sort of compensation. There's a line somewhere between supporting Bill Gates and supporting rank and file developers, and I don't purport to know where that line is, but I don't understand how "all software should be freely written by volunteers" is sustainable. Perhaps that's not the position advocated?
If your goal is sticking it to the man and demolishing capitalism or whatever then it's different of course.
Copyleft, having to release code diffs, is a stipulation / restriction.
MIT/BSD is the one without any stipulations / restrictions:
* https://en.wikipedia.org/wiki/MIT_License
Depending on one's goals one may be better than the other.
> My work on free software is motivated by an idealistic goal: spreading freedom and cooperation. I want to encourage free software to spread, replacing proprietary software that forbids cooperation, and thus make our society better. [1]
Free as in freedom for the users of the software to control, audit, and modify what runs on their hardware.
The stipulations are there to prevent people who aren't aligned with those goals from benefiting from (and working against) the work done by the community that believes in those goals.
It's easy to give the lie to this, because the restriction they are unhappy with is the restriction against restricting others. Anyone complaining about that is just masking a demand of "free for me, but not for thee" - definitely _not_ "totally unrestricted."
For Stallman, and the rest of us inspired by the copyleft movement, writing GPL software wasn't merely a way to impact the profits of capitalists. He would have been scarcely more satisfied with closed-source software written by a democratically organized cooperative and released for free. The goal was and is to ensure that users have the freedom to know what the code running on their machines is doing, and to alter its behavior if they wish.
That dream seems so distant today that people worry it may never be satisfied, but I have a more optimistic view. Someday, hardware generations will not be so rapid or represent such major improvements. That will create an opportunity for the GPL ecosystem to close the gap in functionality and provide alternatives that run on the hardware most people own. Mostly, the GNU project is always running behind on the hardware treadmill; keeping up with software functionality is easy by comparison.
(I guess in many ways this is a data point to support the “OSS work becomes just a source of free labor for large tech companies” thesis up thread.)
No one at any company I've heard of or been involved with cares if you simply use, i.e. run, Linux or other GPL software. That doesn't affect the licensing of your own software.
What they do care about is if you take the GPL software and incorporate it into your own product.
Often we are the end users. Because of permissive licensing phones, routers, iot devices have a lot of closed or otherwise locked own components that I cannot trust nor modify.
The same apply to SaaS.
Of course that is not the case for the majority of projects like curl for example.
The code I do release is probably not that useful to many and in most cases is something i've written for personal use primaraly.
I always use a permissive license for anything I release. It's usually served it's initial purpose to me and if it can be of benefit to another, then they should be able to use it in any way they want
Back when I was just a Linux hobbyist, I loved the GPL. Then I got a job as a proprietary software developer. Now, GPL just means "stuff I can't use at work".
I would love to try and convince my boss and coworkers to implement a project in Scheme. But because the majority of the packages for most Scheme implementations are GPL'ed, it's a non-starter.
So Scheme (and lots of other pockets of FOSS) shall remain a weekends-only pipe dream :(
Especially if it's a hosted web app business?
Providing paid support a la Red Hat, paying for hardware/cloud space, or even paid binaries if you think your customers won't bother to build from source.
(Yes, one can build from source then distribute free binaries, but maybe those same customers won't even go to the trouble of finding them?)
Of course if you accepted external contributions to your code under a copyleft license then you can't use that part freely, but that's kind of the point.
The irony is that non-copyleft project author/maintainers want to reserve the right to exploit other people's work when they "grow the project into a business", but bellyache when Amazon "exploit" their permissively-licensed work. Freedom for me -- but not for thee.
That there is not a good standardised way to give preferential treatment to small companies does not mean it's unreasonable to want to.
Well yes. But if you:
1. Want to accept external contributions 2. Think you might want to use the entire project (which may still be mostly your own work) in proprietary code in the future.
Then a permissive license is a lot easier than the alternatives.
Maybe I'd feel differently if I made an entire service instead of small focused libraries. But that's an entirely different world.
But one person can't do much, so it at least depends on what one thinks of that.
It's not surprising that they invested a lot into spreading FUD around GPL and spreading the idea that weak licenses are "practical" while labeling GPL as "ideological".
I write with other humans in mind, not companies.
I occasionally like to read other people's code to learn. However, in order to avoid any problems, this means that I stay away (much to my regret) from reading any GPL licensed code, no matter how interesting the project might look to me.
I admit to have been occasionally disappointed in the past when large parts of some of my work were plagiarized with no attribution (including by one commercial venture) ... but I prefer for my work to be free of barriers so that anyone interested can either learn from it or (more likely) get in touch with me with suggestions as to how to improve upon it.
Because they can just... stop maintaining it if they wanted. They are completely free to do so.
Many developers like to feel useful, feel like they are solving an important problem in a suitable way. Some even want the fame associated with it, or the doors having such (a) project(s) opens (you use X? I wrote it, and maintain it - or people hunting you specifically because you are known to have created X).
The point is, the tradeoff between the feeling of being taken advantage of and feeling useful and fulfilled because of solving an important problem in a suitable way is controlled directly by the author / developer! If they feel it is not worth the effort, they can just stop!
Having a permissive license helps you gain popularity. Many people won't touch GPL derivatives for their business / side-project-that-might-turn-into-a-business so such a license can have an inhibiting effect in adoption. Developers generally aren't stupid, they know that and they choose more permissive licenses, because they want to be seen, they want their solution to be used. They maybe want the fame and open doors associated with having such a project take off. It is a trade-off under their control. They can stop with maintenance any time. Hell they can even change the license of future versions (if they have not started with a GPL-like license) if they want. So I don't see the problem here. Everything is under the developer's control. You can't have it both ways: The boost to popularity by having a no-strings-attached license, but reaping proportional profits from such an endeavor as if you had a restrictive license protected by a business entity.
If you give money (value) away for free, you'll be pretty popular. "I want to be easily popular as if I'm giving money (value) away for free, without actually giving money (value) away for free" does not make much sense.
But if it is a choice, then it's a choice that undermines fellow workers. By working for free, you're taking away jobs from people who might genuinely need them and be unable to work for free.
In this way, OSS work can serve to entrench inequality by giving a hiring advantage to those with more time and resources, while reducing the salaried jobs available for others in the industry.
Which top jobs require you to be a widely-publicized OSS author?
I've only ever heard this to be a plus (which I believe is fair), not a requirement.
That said, RH hires key contributors all the time (as do many companies), so being the maintainer of a project is definitely a plus.
[0] Or whatever, if it's a small library maybe $50k or maybe something that scales with market cap, etc. Throw lawyers at it.
But that's not what that's defined as.
The OSI defines open source one way, and I just defined open source another way. This is an inconsistency that trips up lots of people on HN and elsewhere. There isn’t one definition, there are many, and calling “anybody but the mega-rich can use this for free; they have to pay” open source falls within some definitions but not others.
The OSI's definition is the widely accepted one. If you mean something else, then make up your own word. To redefine open source like you're trying to do just makes communication more difficult for no benefit.
Why shouldn't they be the ones required to "make up their own word"? "Open source" has a reasonable plain english meaning which doesn't automatically include all the requirements of OSI's "definition". I don't see why the simplest plain english interpretation shouldn't be considered "the real one".
You say OSI's definition is "the widely accepted one", I'm 100% certain that's not true if your sample group is representative of "english speakers" instead of "hacker news readers".
(And I'm pretty sure your linked xkcd supports my argument here more than yours.)
That's "source available"
I think you're confusing open-source with visible-source.
Other "open source" companies are somewhat more realistic and honest about the term and it's history: https://www.redhat.com/en/topics/open-source/what-is-open-so...
I'm not saying the OSI's objectives and motivations are bad, but they did not invent the term "open source", they were not the first to use it, and their ongoing attempts to gatekeep it's usage are - at least to me - tiresome.
(Having said that, I also find rms's hardcore free software stance tiresome, but I'm glad he chooses to dig his heels in so hard there, as an opposite-end-of-the-bellcurve outlier to counterbalance the vested commercial interests trying to extend and embrace the free software movement for commercial greed...)
low complexity, low decision making overhead. it wouldn’t help for small or very specialized open source projects but the big ones that keep the world running are probably used in some way in all the big companies anyways so the total should come out somewhere in the range of 1-2 developer salaries (adjust the 10k above to match. no idea how many trillion+ companies there are right now)
edit: I overestimated how many there are. With 10k one would get to 50k this year. Numbers must be adjusted a bit :)
It's a race to the bottom, largely because developers aren't unionized. Developers who would do work under the GPL get undercut by those who are willing to do it for free.
I really don't understand why there are always comments of this nature every time this is brought up. Why is the size of the company relevant? Do people here not have jobs in the IT field? When people contribute to open source, anyone is able to use that code, including you. That means when you go get a job at a new company, you can keep using that code. It doesn't matter if you go to a small company or a large company.
But the thing is, they often don't. That tends to rub people the wrong way.
Some products like databases are now switching to AGPL but it takes a lot of effort to revert the trend now.
Not a lot of URLs on GitHub sit outside of their regular /owner/repo pattern - the only other section I can think of that does that is https://github.com/about - things like documentation tend to live on separate subdomains, like https://docs.github.com/
[0] https://readme.tumblr.com/
[1] https://web.archive.org/web/20090709114358/github.com/readme
[2] https://web.archive.org/web/20160218095949/github.com/readme
How does Microsoft come to own all of open-source? That's the plan, right?
I wouldn't be surprised if they end up picking up Docker for a song now that Docker has utterly failed to find a workable revenue model. (Microsoft sponsored Docker's surprisingly-complete Windows support, IIRC.) The extremely-widely-used Docker Desktop products are already nonfree, as well as the dockerd that runs on Windows.
Coax people slowly over time to give up other options until there's nothing else to choose, then flick the switch and fleece them harder than hardness itself.
Based on past performance I just assume we're in an elaborate version of the 'fleece tourists with inordinate drinks bills' con trick.
#GetOffMyLawn
“Even if you have funding, there’s always more work than you can possibly do,” Van Slyck says. “And the bar is always being raised. It used to be enough to just put some code online. Now you’re expected to foster a community. You have a lot more hats to wear.”
EDIT: I glossed over this part when I was reading the article, but indeed there are many open source developers who do hold this attitude:
Halliday—better known by his handle substack—takes an uncommonly hands-off approach to his work. “I have all my GitHub notification emails turned off,” he says. If someone finds a problem with his code or wants a feature added to a module he’s no longer working on, they’re free to fork his code. That is, after all, the open source way. But he doesn’t put effort into issues or pull requests for packages he considers finished. “It’s not my job to keep tabs on every little thing I wrote years or decades ago very often,” he says. “I am always busy with new projects and if I was always looking back at old projects I wouldn’t have enough time to move forward.”
I fully agree with this viewpoint and I think it's good for the overall software ecosystem if engineers did more due diligence on the packages that they use.
If that person is actually motivated enough to write a patch (which I have had!), I welcome reviewing it, because it empowers that person to feel like they can contribute to my project (and hopefully other projects too!), and makes others feel like they can contribute to my project as well.
But then I'm employed by $MEGA_CORP and part of the job description is "helping to build and maintain the OSS community", so helping novices create their first patches is a big part of that.
That really bad support didn't lead to any actively maintained fork, though. So I guess it's all stable enough for people to still depend on it in their new projects.
It's time our governments start taking OSS seriously.
Like academia, OSS deserves state sponsoring.
Most of the research papers I read were made possible by OSS one way or another.
Ironically it can be really hard to donate to OSS from government entities.
It's actually easy to pay someone for software, but for all these libraries and technologies that are used it's really hard bureaucratically to say, donate $2K a year. Which is still a fraction of the value gained from open source for most places.
If you stop and think about just how much stuff humankind built around stuff like the mentioned curl, or linux, or gcc, or numpy, or netscape.
We all stand on the shoulders of giants.
"Did you know there are less than 2100 people in the inner GitHub open-source maintainer community? I often wonder what the future would look like if these high achievers that our digital society is built upon were empowered to become independent artists."
https://www.gitpod.io/blog/devxconf-wrap and https://www.gitpod.io/blog/gitpod-open-source-sustainability...
Anyway, just to get it out there, this is what [I think] should happen:
A small group of really really good programmers write software that works (is bug-free). The rest of us use it, we are technicians who configure and deploy the code written by the elite. This (plus technological innovation in general) results in an "economic supernova" where science and technology are applied to solve our basic problems in an efficient way. (Bucky Fuller calculated that this would be possible by sometime in the 1970's.) This results in the collapse of the profit motive precipitating a post-historical secular utopia (e.g. Star Trek).
All of the technology is there, right on schedule, but we as a species continue in our old technically-obsolete patterns, unmindful.