From one of the articles the article refers to: "[According to Harvard something] value of OSS to the economy is 4.15 Billion USD."
That seems ludicrously low - Linux itself, if you value it at $1 per install is considerably more that 4 billion USD. The various open source web projects powering all the large sites are more than billions. Banks may not pay for SSL but they receive economic value. Etc. etc.
And it's not a labor surplus except for those poor souls that are trying to make their fortune with an OSS "side project." It is a method of doing things where results are shared and improved upon, fundamentally the scientific way brought into applications. Unlike physical devices, it's essentially free to share software. Just don't put access controls on the git repos.
There is a tendency to overly productize things, but that's not the heart of open source; the heart is just to share results, perhaps in a hope that ones labor can contribute to the progress of humanity. I don't need to promise to fix issues ever for my source code to potentially have value to someone solving a similar problem.
Now, it's harder to achieve this way of working in corporations, and things have to split into "open source with some modular hooks for private things that aren't open source" (even just in-house logging or metrics systems), but it's not impossible, and does have the same advantage science has - it grows exponentially and eventually will figure out how to do anything that is doable.
'We first estimate the supply-side value by calculating the cost to recreate the most widely used OSS once. We then calculate the demand-side value based on a replacement value for each firm that uses the software and would need to build it internally if OSS did not exist. We estimate the supply-side value of widely-used OSS is $4.15 billion, but that the demand-side value is much larger at $8.8 trillion.'
The methodology used in the underlying paper is, to put it generously, not even wrong. Way past “assume a spherical cow” territory.
It supposes you could simply hire programmers to build OSS from scratch.
If you have ever worked on a large project in a corporation, you instantly know how shockingly ignorant this is.
Hint: many of them end in failure and are never released at all.
Then there are the massive amplifications that happen due to the mere existence of open source: learning, spreading of ideas, reusable tooling, and more.
Has any business school ever produced a paper worth a damn?
As you say, whatever the real cost is, it's much higher than one that supposes every company could reproduce F/OSS. How amazing then, that even in this hypothetical universe where everyone could, we still wouldn't want to due to the enormous cost.
Providing a lower bound on value, and furthermore one that is astronomically high, is extremely useful as an eye-opener. This is a useful result for policy-makers.
A plausible lower bound would be useful, but $4B is a joke. That’s less than the annual budget of single major university (Harvard is around ~$5-6B).
Gartner says the world spent $4.5T on IT in 2022. To pick some numbers out of thin air, let’s assume half of that is on software, and half of that is on new software (not maintenance). And the let’s assume that software is sold at ~20% net margin. And open source powers an enormous fraction of it at some level, but let’s be conservative and say it’s 10%. $4.5T * 0.5 * 0.5 * 0.8 * 0.1 = $90B per year just for the new stuff.
Recreating the existing stacks… if we multiple that number by 25 years, you are at over $2T to rebuild OSS in some kind of manhattan project.
But even that is a massive underestimate. For one, it would be competing with other development objectives—there aren’t millions of principle-level engineers just sitting idle. And more importantly, formally run projects have radically different levels of passion and “giving a shit” compared to how people often start OSS.
You simply couldn’t do it at all. It would be like asking “okay, but assuming we really had to replace all the metal with wood, what would it cost to launch a manned wooden rocket ship to the moon?”
> We then calculate the demand-side value based on a replacement value for each firm that uses the software and would need to build it internally if OSS did not exist. We estimate ... that the demand-side value is much larger at $8.8 trillion
But details like a 1000x change in estimate aside, saying it couldn't be done is not quantifiable. At least a dollar figure can be reasoned about.
Saying it can’t be done is still an actionable guide to policy. If it can’t be done any other way, and it has massive benefit, then policymakers ought to be careful to avoid actions that harm it.
For example, tax law changes requiring multi-year depreciation of engineering labor probably have an extremely detrimental effect to small open source projects with commercial potential.
The policy effectively significantly raises working capital requirements for an early stage OSS company that might be focused on services revenue instead of licensing fees.
We're missing the forest for the trees - the estimate was large ($5T), and the estimate is indeed a lower bound, and it seems like pretty damn important to save F/OSS.
Spot on, and it forgets the creative aspect, it‘s like estimating the cost of reproducing all of the world‘s widely read literature. Who could recreate a Shakespeare from scatch, who a van Rossum?
"We estimate the supply-side value of Shakespeare's works by estimating the cost of paying an author to write the approximately 884,000 words in his collective works. We therefore calculate the value of Shakespeare to society at $320,000, or approximately the same as the value of 3 Tesla Cybertrucks."
nobody because the times and places either of them lived in no longer exist. surely the places and the geographical locations and buildings are still there, but the people are all different, so it's not the same place
That's important in literature in a way.. But does it matter to the economy if POSIX were created as it is or if some OS slush fund created a different computing model obsessed with some other primitve besides the file?
In our values a lot of literature is valuable because of how it came about. If Shakespeare never existed and someone wrote like him today we would just lock them up.
it's not about any of that "reasoning"... what matters it the conclusions: the results.
and even more than that, what matters is the capability of the paper to sway the opinion of real power.
our experience dealing with power may vary, in mine, power does not respond to "reasoning" nor any of that stuff.
nonetheless, I agree with your sentiment. why does science demand replication? IMO, a key underlying consequence is that ideas must be transferred, given away before they can be science.
the ideology of "real" human-centric science is equivalent to open source mindset. so then, my question is what to call all that research that happens, privately, in secret and under a lot of NDAs.... it is not science; that is product development.
> To reproduce all widely-used OSS once (e.g., the idea of OSS still exists, but all current OSS is deleted and needs to be coded from scratch), using programmers at the average developer wage from India, it would require an investment of $1.22 billion. In contrast, if we use the average developer wage from the United States, then reproducing all widely-used OSS would require an investment of $6.22 billion. Using a pool of programmers from across the world, weighed based on the existing geographic contributions to OSS as discussed above, would lead to an investment somewhere in between the low and high-income country, $4.15 billion.
Note that they just assume that you can basically hire programmers to type that out and extrapolate from lines of code. So I'm guessing this is off by at least one order of magnitude.
Granted, I'm not sure if I have a better option, but I'd probably start with sampling OSS contributors to figure out how much time they actually devote to OSS.
That's still a terrible underestimate. Meta alone spends more than this in VR yearly with development and they are not recreating more software than the whole OSS every year.
Agreed. When it first came out in January, I did a write-up and emailed Frank for comment. It wasn't until both my critique and his paper appeared together on HN for the first time that he responded ... with a total non-answer. :rolling_eyes:
I do agree that the open source ecosystem seems worth more than that. But $1 per Linux install seems like a kind of arbitrary pick that might not meet our intuition, if you want to hit 4B installs.
Like, there are only ~8B people, and most of them don’t use Linux on the desktop, so we must be including other uses of Linux. Routers, IOT. I’m pretty sure my lightbulbs run Linux. Is Linux supplying $1 of value per lightbulb? I’m not sure, I suspect smart bulb manufacturers could have gone with some proprietary stack if Linux didn’t exist, and it might cost less than $1 per lightbulb… actually, I have no idea, but I don’t think $1 is an obvious lower bound.
Okay so lets say $10 for every Android cellphone that puts you at 30 billion. Not sure how you want to estimate server installs but a quick google puts forth 96% of web servers run Linux.
Collectively, we spend an awful lot of resources on air. Clearly it must be of some value, otherwise we wouldn't bother. Unless it's a scam.
Sure, we're not being metered and charged by volume of air used. But someone, somewhere, is spending money to make sure it's suitable for breathing. And it's a pretty penny too.
If corporation x uses my thing to cut labor costs and doesn't pay me anything while actual employees get to sit around and make six figures adding fields to jsons, that's a shitty arrangement for me
If corporation x didn't exist, I probably wouldn't worry about it. But the second money becomes involved it changes the calculus
And I do think some projects like ad blockers are worth it because corporations don't benefit much from them, but regular people do
> If corporation x uses my thing to cut labor costs and doesn't pay me anything while actual employees get to sit around and make six figures adding fields to jsons, that's a shitty arrangement for me
Then perhaps FOSS isn't for you.
There seems to be a large segment of the community these days that doesn't fall into the MIT/BSD or GPL camps and I'm not sure why they care about FOSS if they don't agree with one or both of those philosophies.
Not to say that those are the only true FOSS philosophies- I just don't understand why this third contingent isn't making their software closed source.
> Not to say that those are the only true FOSS philosophies- I just don't understand why this third contingent isn't making their software closed source.
This perspective pre-supposes the _the_ binary to analyze the world is open vs close source software (binary 1). But there are other binaries.
There are at least two reasons why people are attracted to FOSS. One is the "freedom" angle. But the other is that people like working within a gift economy - a system of value creation in which everyone gifts value to others to the best of their ability. FOSS licenses explicitly only care about the freedom value, but clearly there is some overlap with the gift ethos as well.
But for-profit companies using OSS to make crazy amounts of money and then gifting nothing back destroys the gift economy. This is another binary: systems that create gift economies vs those that don't (binary 2).
Many people have decided that they don't want to make a fool of by these companies, and they stop using FOSS licenses. Using binary 1 analysis, it seems these people prefer or should use closed source licenses. But from the binary 2 analysis, FOSS licenses create a non-gift economy. So there is a need to create other licenses or systems that create and enforce a gift economy.
I feel like in your last part you are also pre-supposing another third binary (binary 3): the copyright license is the only relevant part to enforcement of a system vs OSS is a community and is bound by ethics as well as laws.
The distinction between ethics and laws is a fascinating dichotomy. "What is the culturally right thing to do?" versus "What is the allowed thing to do?"
The gift economy is already an ethical system like "Take a Penny, Leave a Penny". There are no laws that tell you what to do, you are encouraged to do what is right for you and hopefully what you think is right for the larger culture (the next person into the shop; the next user of the open source software).
Sure, it is hard if not impossible to "enforce" a general ethical system. To some extent that is what laws are for in the abstract ideal, but in regular practice there's a gulf there and we find that some of the beauty of the ethical system is broken when attempted to be constrained by legally enforceable laws.
If you watch someone take all the pennies, stuff them into a Gucci bag, cackle like a villain, and never leave a penny and think "Someone should do something about this", you start to find how slippery it is to add rules to who must leave a penny and who can take a penny and how often and how many variables and edge cases there can be just to try to figure out who is "right" to leave a penny and who is "right" to take one. The open "you know best for yourself in that moment" invite of "Take a Penny, Leave a Penny" disappears exactly in that moment you try to constrain it with rules.
As software developers we love rules and coming up with them. Software is about plumbing for all those variables and building the big charts of the edge cases and solving them. Binary 3 is an easy trap here: people aren't playing the game right, lets add more rules to the laws (software licenses). But it's a break to Binary 1 (forced sharing licenses aren't "fully open") and just as with TAPLAP, it starts to sour the "everyone gifts value to the others to the best of their ability". By adding rules you have to define "value" and "best of their ability" and in doing so you lose that fragile idea of "try your best to do what is right, only you really know what that is".
In the real world TAPLAP is enforced via social contract: reputation and gossip, trust and distrust, dislike and shunning. For enforcing a "gift economy" culture in open source, there are options more like the social contract (industry ethics boards, unethical software company boycotts/reputation smearing). Some of those pieces exist (OSI is an ethics board; ACM has an ethics board; IEEE has an ethics board; FSF likes to think it is an ethics board, but also clearly trusts rules/laws more than open, unconstrained ethical systems) they just may not have enough teeth, especially with respect to enforcing anything like the "gift economy".
So I agree there's a need to better enforce a gift economy, but I don't think it drives a need for other licenses or systems as much as it shows a usefulness in giving more enforcement to the existing "social contract" of open source. Take a library, leave a library. I don't think adding more rules solves enforcement. (And it certainly endangers our principles for "freedom" such as binary 1.) We can and should be able to enforce it even in its "not very explicitly stated" current form as an "ethical obligation" (versus a "legal requirement", very different things). It just is a harder (improv) game to play with less "rules" to play by. (As much as we love rules in software, sometimes it is nice and freeing to take a break and go play an imaginative game with fewer rules.)
You cannot have soft enforcement of ethical norms without hard entry requirements. That soft enforcement can only be achieved in the context of high trust social groups, which exist in the context of relatively insular cultures.
I agree in broad strokes with almost everything you say. I don't think adding rules is the correct way to build societies with some desired ethics.
That said, open source software licenses are extremely simple as far as contracts or licenses go. At least an order of magnitude shorter than closed source software EULAs, for example. What I see people advocating for is not tackling on additional complexity onto the existing licenses, but replacing the fundamental principles of existing licenses with new simple principles that slice the world into acceptable and unacceptable behavior in a better fashion.
Of course, licenses alone are not sufficient for change as you say. A lot of other social structures will indeed need to be built to change large scale behavior. That's a hard long slog.
> but replacing the fundamental principles of existing licenses with new simple principles that slice the world into acceptable and unacceptable behavior in a better fashion.
Feel free to come up with a different set of principles, but don't call it FOSS.
What frustrates me about this current FOSS discourse is that the are a bunch of people who want to change FOSS principles without (seemingly) understanding what those original principles are or why they were chosen. These critics live in a world in which FOSS has been uproariously successful but don't appear to appreciate what was necessary for it to be successful in the first place.
this is why AGPL is perfect for the job. Solidly FOSS license, used without problems and usage increasing, temporarily embarrassed billionaires go into a screaming fit about it.
Even mentioning the AGPL on HN gets a flurry of not-even-wrong FUD in the style of early 2000s Microsoft about the GPL, c.f. the recent story about Elasticsearch: https://news.ycombinator.com/item?id=41394797
They are not paid 6 figures to add fields to JSONs. They are paid 6 figures to brainrot into obsolescence instead of going out there and creating a competing product.
The last part is working very well. Most of the innovation these days is coming from China.
I would guess they are referring to the idea that large established companies (usually monopolies) have little motivation to compete of features/technology, especially when that would cannibalize their main offering. Generally, an engineer working on something at Microsoft or Apple is going to be much less motivated to create new tech than one at a (Chinese) small business. But it gets worse than this: they sometimes hire engineers just to keep them from the competition.
That said, China has its own unique category of underutilized staff: those who are hired to work for state-owned enterprises.
Here is the one core aspect about open source. There are people who really want to open source code without expectations of money at all. Immaterial if users are corporate or not. Perhaps for love of the art, career building or even just prestige and so many myriad motivations. I respect their choice. And it's not a shitty arrangement from their p perspective at all.
Yes, of course there are people who really want to do those. But op was presenting their main point of view related to "the slow evaporation" part.
Like op, more and more people can't get past the idea of starting or contributing to FOSS is no longer worth their time, due to various reasons, such as: their free labor of love was turned into another corporate money making tool.
I use ads on my OSS projects' websites so to the degree ad blockers block low key, non intrusive ad networks favored by OSS projects like Ethical Ads and Carbon Ads, they are hurting regular people
Ads are unwanted. If you serve ads, you are doing something unwanted to your visitors. So they block the ads, and you get the unwanted effect of not making money. The pain point is not that your ads are being blocked — it’s that you don’t have a way to make money that doesn’t start with doing something unwanted.
It's the way ads are delivered that is unwanted. I would not care about a static inline banner ad at the top of a page. Slightly possible that I'm interested in it if it's relevant in the context of the page, easy to ignore if not.
The ads that pop up when I scroll, or ads between every paragraph (or more) or ads that are animated or use fancy CSS to overlay the main content or slide in from the side and float around -- those are what make me run an adblocker.
> We don’t yet know which parts of the FOSS system is sustainable,
Probably the parts that have been around for decades.
As someone who's created open source software in the past, I'm put off mostly the explosion in software complexity over the last two decades, the constant churn in libraries and tooling which places an unreasonable maintenance burden on a spare-time hobbyist, and the bad taste left in the mouth by Microsoft using copilot to license-launder the github corpus.
I still produce open source software and gateware on a recreational basis, but in quiet little backwaters of the noosphere which aren't going to change the world (or be tainted by AI) any time soon!
> Microsoft using copilot to license-launder the github corpus
Isn’t the game with licenses that you must specify what can and cannot be done with the code? It’s a big issue that no one prior to now had the foresight to forbid ML training on code without attribution. If the licenses going back 30 years had that stipulation then it would be easy to take down Copilot and ChatGPT. But the licenses simply don’t cover the use case of training a neural net, so it’s probably going to slip through the cracks just like SaaS slipped through the cracks of the GPL by not distributing code, hence the need for AGPL. So I’m sure we’ll see these kinds of clauses added to licenses going forward, but they can’t be applied retroactively.
The irony in all this is that from the start, open source licensing has been a game of wits where software creators try to cleverly use copyright as a lever of control. Well, they weren’t clever enough. They missed ML training and didn’t forbid it. As a result they’ve basically lost the whole game.
> Isn’t the game with licenses that you must specify what can and cannot be done with the code?
No, not at all. Microsoft's argument is that training an LLM on code is fair use, and thus doesn't trigger copyright-based licensing at all. That's why they include unlicensed code in Copilot, which under a training-triggers-copyright theory they have no right to at all.
> OSS burnout. Very few FOSS projects are lucky enough to have grown a sustainable and supportive community. Most of the time, it seems to be a never-ending parade of angry demands with very little reward. When the software labour market was growing steadily, burned out maintainers often got replaced by fresh-eyed graduates or coders who relied on the project at work.
This is big not just because of angry demanding "gimme free stuff slave!" users but burn-out due to the realization that open source in 2024 is often just free labor for SaaS companies.
Open source should not be considered "free." It should be a kind of gift culture where one is expected to give something back. Showing up with angry demands for something someone made for free makes you a parasite and an asshole.
Yeah you used to be able to get nerd street cred as an OSS maintainer, and the cost was your time and willingness to work on A) the code and B) whatever contribution processes you/others came up with. The downsides were known.
Now, you have this yawning chasm opening of large companies taking your work and selling it back to you, or LLMs remixing your work and diffusing it out, both demotivating factors that combined with the economic insecurity and COVID lockdowns forcing people to re-evaluate their hobby priorities, I'm not surprised there's less maintainers.
FOSS is an activity taken up by people who want a particular solution to a problem. When the tech and IT industry were vastly smaller than it is today, there was actually a larger number of high-quality FOSS projects. Today there are more FOSS projects than ever by number, but a good chunk are what I would call "throwaway" projects. Another large amount are projects that just exist to fill gaps in new ecosystems (a new language/framework/platform/etc appears, and now they need a lot of projects to provide libraries to fill needs). And there's lots of immature engineers who are growing up in a culture without documentation, and too quick to reference snippets than learn a tool completely. Tech itself is changing, and that will (and does) affect open source.
But the existence of open source is self-perpetuating at this point. There isn't really a bubble to burst. FOSS will continue to be here as long as the computer allows anyone to create and share a work for free that can be reused and there is a need for. It is not sustainable, because it does not need sustaining. Anybody with an internet connection can decide for themselves to create open source. As long as we provide an unimpeded means to share their work (newsgroups, mailing lists, FTP, mirrors, etc), there will be open source. And as long as somebody has leisure time, and there are nerds to be interested in programming during their leisure time, some will spend it on open source. It's like "sports": people will still play sports regardless of a business motive.
I think people are shocked by the whole "business source" license thing, but I'm not. Open Source was never a business model, no matter how much some of its proponents wanted it to be. It's taken a while for people to learn that lesson, but it's starting to sink in. The fact that fewer startups will be open source now isn't going to change the open source community at all. The community will be here regardless of what any business (or individual government) wants.
As far as "our reliance on open source", that's just a happy accident. It's free of charge and free of use, so business uses it. If for some reason it stopped being free of charge or free of use, then business would just pay for proprietary software like it used to. I don't see this happening anytime soon, though, unless somebody passes a law banning Copyleft.
I think the author thinks the primary motivation for open source is companies making a donation out of the goodness of their heart.
While that does happen, i dont think that is the main cause. Most companies participate in open source for their own selfish reasons. Sometimes that is to attract talent, sometimes it is to attract people to their propreitary platform, sometimes it is to commodify the parts they dont want to compete on, sometimes its something else. The companies do get benefits here. They are probably going to continue so long as the benefit > costs.
As for non corporate open source. It may not be a healthy hobby, but i dont see anything changing here. If people have been doing it as a hobby for the last 20 years, why would they stop now?
That's the thing, while there is value in investing in open source (for corporations and individuals) that value is often speculative and hard to quantify. When times are tough, people tend to focus on short-term, concrete benefits.
This, I don't think open source can be properly characterised through the lens of capitalism. There will always be students, academics, and people who don't work in software at all, willing to try their hand at developing something for their own use, which may or may not get traction in the grander scheme. It doesn't have to be an "unhealthy" hobby if you pace yourself and draw lines around what you're willing to do or not.
I also think there will always be people who really like software development but aren't able to fit into the corporate world. The article mentions rich devs with too much time on their hand, but i've also seen some open source devs who are actually quite poor and underemployed, that simply, for whatever reason seem incapable of thriving in traditional corporate software jobs.
> Short version: my mental model of FOSS is that it’s a function of industry and labour surplus
I believe that originally it was more a function of college students spending their free time on it. Of course, this results in addressing fewer enterprise-level use cases.
Even if you look at industry contributions, this is not actually why code is open sourced.
The individual incentive is to open source code to show the rest of the industry what you have been able to build, with personal returns to you if it actually gets meaningful adoption.
And companies are willing to go along with this because infrastructure projects are not actually a source of competitive advantage and they would rather co-operate on keeping input costs low, and also not get stranded on a tech island with totally custom tooling for everything. You also derive some marginal benefit if the OSS code actually gets used by others since you still control the project direction.
And then separately you have OSS first companies that are hoping that being OSS lets them gain a marketing/sales advantage with developers who care about that.
These incentives remain the same even if there is no meaningful surplus.
Exactly. For individual, hobbyist developers, just showing their work might already be a strong incentive but for profit seeking companies, the motivation to open source their code hinges on tangible benefits like you said. We're developing a digital platform with a marketplace for 3rd-party developers to contribute and make money from their contributions, at one point we had to decide whether we should open source our code, the pros and cons in doing that etc. And we came to conclusion that, in our case, by proving both cloud and open sourced version will definitely help drive more attention to our cloud offerings and attract more developers to provide/sell their own applications on the marketplace. However, we have to keep our focus on our cloud version first and delay the release of the open source version for at least 1 year or until our cloud gets enough traction/attraction. I don't know if this will work out well in the end but we have got to try.
I was around when it started and at that time it was ideological, ala Stallman. Those participating were some of the best programmers that have ever lived. Much later college students got involved and their motivations were totally different - they were looking to do something that would help them land a job, build a reputation, or make money.
It still is, for some, ideological.
I know plenty, whose day to day job isn't even software, doing this.
Now, just as food for thought:
I believe the mere existence of FOSS promotes a more sustainable economy!
It regulates, in a way, commercial software.
You cannot be more user hostile / be less convenient and/or provide lesser of a service than FOSS alternatives! Convenience is at the end of the day, the reason for someone to choose a tool. Software is a tool. Greater convenience of use generally promotes greater productivity when using the tool.
Many of us already make the logical decision to use FOSS to not have to deal with the rising user hostility of consumer grade software. Maybe because, we as developers, have the know how to minimize the rough edges or have the same type of brain of the creator of the software in question, and think: "Indeed! the decisions made on the interface, the flow of using the tool, whatever else, are sane and make the most sense!!!". Hinting here, that the incentives of FOSS do not necessarily align with the preferences of a consumer, which is sometimes a good thing.
Nevertheless! The freedom, as in free, not as in beer, represents a bar of sympathy, of respect for the user. One that has the potential to minimize user hostility of commercial software for fear of losing that customer. Users are beginning to be more savvy than ever, they've been burned a lot before, and/or were born in the midst of the digital age. The barrier of FOSS, its arguable lack of convenience, is shrinking every day. And, in such a way promotes an economy where you have to respect your customer. One that is then, more resilient and sustainable.
I mean like Linux, and Unix tools like vi. “College students” is maybe too limiting, but a large number of early open-source projects came out of university environments.
I generally try to stay away from responding to these posts but it's starting to feel like it's an astroturf concern trolling strategy.
I'm not sure I've properly digested the article but, just from a cursory look, it doesn't even mention the exponentially decreasing costs of compute. I'm not sure my model is correct by I suspect the main driving factor of why FOSS exists and is successful is that the tax imposed by current copyright/IP laws is too usury and FOSS provides a mechanism to circumvent it and provide/capture the value from the dropping costs of compute.
Is the FOSS surplus evaporating? Maybe, but my default position is no. I'd be willing to change that position if some evidence were provided. The article provides none.
Articles like "Is the Open Source Bubble about to Burst?" are click-baity titles that seem on the razors edge of bad faith. The author of that article claims to be a FOSS advocate so I don't know.
Maybe I'm reading too much into things and this is really fallout from some EU politics I'm not fully looped into.
Is there anything here that isn’t purely speculative?
What are the signs that OSS is actually in decline?
BTW, I’m not sure there’s anything wrong if OSS is in decline. Overwhelmingly, people do it not to get paid or help industry but because they love building things and solving problems. If people are generally finding better ways to do that, then good for them. Or if people have less time for OSS due to general economic circumstances, then that is a sad thing, but that’s not specific to OSS, so it’s a bit weird to frame it that way. That is, if people have fewer opportunities to find personal satisfaction in their lives, that’s bad across the board, not just for OSS.
The whole article is evidence-free nonsense. Free software has been around for at least 50 years, and will still be around as long as programmers want to scratch an itch or just enjoy (as you put it well) finding personal satisfaction from it.
I think the extent to which one could say open source software is in decline is driven by the fact that FOSS was recently considered by many big VC firms to be an “investable category” and now that view is less popular.
True FOSS has always been uninvestable, healthy, and small. But we’ve had a lot of “eventually fake” (non-MIT/GPL) FOSS in the last decade. Elasticsearch, mongodb, cockroach, and so on.
True FOSS (maintainers not beholden to rent-seeking investors), as always, trudges along. Because it’s growth is not correlated to market returns.
When 100 companies have an identical problem, and it's not a problem that's in the ___domain of what they're competing over, then it's a tremendous waste of labor for them to come up with 100 separate solutions. It's just cheaper for them to allow enough wiggle room that the problem gets solved in the open so they can all get on to focusing on whatever it is that makes them different from their competitors.
lots of bullets in this blogpost but here are some gems
> OSS burnout. Very few FOSS projects are lucky enough to have grown a sustainable and supportive community. Most of the time, it seems to be a never-ending parade of angry demands with very little reward.
> People who are unemployed or jaded by the software industry have fewer side projects, because – let’s be honest – there are healthier hobbies available.
> There’s less funding for non-AI software startups, who are usually very heavy OSS users.
> Declining surplus and burnout leads to maintainers increasingly stepping back from their projects.
For what it's worth: Thank you for your work on SQLAlchemy and Alembic!
On large projects like those, what ratio of community interactions would you say are angry/demanding vs supportive/helpful? And any advice for dealing with the negative ones?
For me, probably less than 5% of interactions on PRs/issues/discussions are negative, but even that small amout sure does have a way of draining one's enthusiasm and motivation!
if you put angry/demanding vs supportive/helpful on a 0-10 scale, things hover in the 6-7 range typically. i think computer programming for whatever reason correlates a lot with a certain kind of impatience that we all have. I have gobs of it. We're all fighting it each day to not piss each other off because when youre in "flow" we all know how it is when whatever OSS project suddenly surprises you.
truth be told I spend my OSS maintenance day being more and more pissed off all day really and a lot of it comes from the desire to recognize when people are either subtly or not-so-subtly asking of you to make sacrifices for them. The person who didn't read the docs, the person who didn't read your "new issue" template asking them to please open a discussion since they likely didn't find a bug, to write clear self-contained demonstration code and to not assume your well defined and documented behavior is a "bug, let me know when this is fixed", the programmers who are asking you to upend your whole project for what they in a very dunning-kruger sense think is a good idea, these are all things that someone can respond to in a patient and friendly manner. Heck anyone that works in the service industry has to have an iron-clad patient and friendly manner with all forms of idiots and jerks, including when it's me. But for me personally, doing the open source thing, and also quite obviously for a lot of other folks doing it, man it's hard to keep the fireballs in check while at the same time giving each of these users a clue that there's something they could be doing to make life easier for the maintainers of the project that they are using for free.
> anyone that works in the service industry has to have an iron-clad patient and friendly manner with all forms of idiots and jerks
> giving each of these users a clue that there's something they could be doing to make life easier for the maintainers of the project that they are using for free.
The intersection of these two things is something I've been mulling over lately. In other words, how do we respond to the implicit proposition of "I'm going to pay you $0 and expect the effort and patience of a well-compensated service industry employee"? Putting in the mental energy to deescalate conflicts and be diplomatic with challenging personalities is something I'm willing and able to do... but not for a hobby/unpaid volunteer project!
But then there are many different types of people behind those interactions, including but not limited to:
* Someone who's genuinely a sociopath, or at best an entitled ne'er-do-well who just wants free labor
* Someone in full "socially inept engineer mode," with terse communication that comes across as curt or demanding via text, but are otherwise reasonable if you spend more time talking with them; in their own mind, they're probably just being "direct" or "efficient"
* Someone new to open source etiquette, who doesn't yet know what a good bug report or feature request looks like, but are willing to learn with a little guidance
Sometimes it's hard to tell the difference, and all of those cases can have the same effect of being rather draining, at least in the short term. Longer-term, I've had at least a few cases of someone who seemed like an entitled jerk at first blush, but turned out to be a valuable power user who stays engaged with the project and provides useful feedback. Of course power users come with their own challenges, like wanting to cram every feature under the sun into your project without appreciating the long-term maintenance costs. But at least they're more fun to work with!
I've been doing and following OSS for over a decade now and this is the first time I feel like a piece on OSS excludes me. I started in the maker communities in Europe, and most OSS creators I know of are small hobbyists all over the place, not particularly "California rich devs with spare time". In fact I believe that demographic might be (relatively!) under-represented, they have startups to work on after all!
However, what I did not expect is how hard it is to get new open source to be used by people. You can have wonderful features but without a really large amount of marketing effort the pickup can be really slow. Unless you have help from the right influencers. And without traction it will either die or go in another direction, possibly close source/commercial.
What's also sad is the legal system can change, making the cost to entry for commercial use of software for small players much higher or risky. Some people will have worked for years on open source, possibly thinking it can be something they can do commercially after they retire for example, then governments bring out a law with a 15 million euro penalty if you are not compliant (Cyber Resilience Act).
I guess this is a big kudos to those whose persistence have made it succeed for them!
The article seems to deal only with FOSS coming straight from the enterprise world, also as a byproduct of professional developers' free time. This is sometimes the case, but it also misses completely what FOSS is for non professional developers: their best possible resume; and that is not going away anytime soon.
> In giving companies a free pass to enter the “open source community,” however, certain hackers said “take what you want and give what you want” to a bunch of organizations built around maximizing the ratio of the former to the latter.
Like the author says, investment by these entities can balloon the (F)OSS ecosystem, but when they contract (as is inevitable in a boom-and-bust economy), the nonrenewable resources that actually write the stuff (humans) will burn out.
What about Blender? Is it not a canonical success story? I feel that science, art, and tools that provide personal agency are all compelling landscapes for open-source fruition.
One of the reasons Blender is so amazing is that it's an outlier. It's also not in the class "generating billions of dollars of value" unlike some basic components like, say, Linux kernel or SSH to name a few (in an ocean of extremely valuable software components).
> It's also not in the class "generating billions of dollars of value"
Clearly not as much as Linux, but almost certainly it's already been a part of asset creation or use within film and video games that topped billions of dollars. The movie RRR alone is 160m+ worldwide (See in part https://www.blender.org/get-involved/user-stories/).
Love Baldur's writings in general. Smart & insightful, great topics.
I do feel like the mental model here - industry & labor - defines away a lot of the currents that arose open source. These factors are the industrial/capitalist view of open source!
Perhaps I was just young and naive, but man, it was such a an exciting time to be alive, and it was so clear to young me that computers already could do anything you asked them to do, and that they were getting soaringly better all the time, and that we were building new world spanning interactive mediums and revamping how society communicated with itself.
It was a grand project, and the term I heard recently & love vocational awe barely contains the magnitude of interests & awe I had & still have.
And it was such an open accessible time. Reading ajaxian.com was a constant torrent of regular folk inventing futures. We were harkening always towards new better protocols and libraries.
It felt so much more than industrial. And the scale of activity was on such a more personal, small innovator level. There weren't huge teams maintaining React or Nextjs. There were a couple folks creating Comet, or BEEP, and sharing not just the output but more engaged in an arena of ideas, discussing on blogs and comments the tradeoffs & nitty gritty.
Lately, tech hasn't been accessible, and it hasn't been cool. Open source used to be the connection for cool and accessible, used to be endless blogrolls that would suck folks in and let them see & become part of that project of building the new world. Open source has, as it's gotten more economically important and larger scale, become just a taken for granted tool, not something that was going to alloy & enhance our lives or expand our thinking.
The motivations matter, the image matters, the cause matters, the drive matters. We can look at industry and labor from a rationalist economic view, to assess open source, but the philosophy and motivations to me are more interesting, define the energies or lack that create the awe & passion that drive this amazing open source world, that impel people in.
There's also uneven concentration of the open-source effort.
In the past 15 years, companies, or at least the companies I worked for, loved open sourcing their systems for better reputation, for larger community support, and for not getting left behind by competing solutions.
I sense that the tide has changed. Now we have a lot more mature solutions than a few years ago. Few companies are growing so fast on a new ___domain that calls for brand new systems, either. So, the cost of open-sourcing a system may outweigh the benefits. Nowadays, it is more often than not that a system startup open-sources their systems to gain user traction, and soon makes their systems only source-available. And rightly so. They have a business to run anyway, and they certain don't have any obligation to open source their bread and butter.
There is another factor: the fun is always in new challenges, yet a lot of software development has more chores than challenges. I remember many years ago a pretty popular blog explained why Linux did not have a nice UI compared to Windows or MacOS. The main reason, the author argued, was that tweaking UI was not necessarily fun, and we achieved great UI from thousands and thousands of incremental improvements, more often than not painstakingly. An engineer who worked on open source in her spare time most likely did not have incentives to make such improvements. I mean, what's the fun in that? In contrast, people work for free to create amazing libraries in the field of deep learning. I'd venture to guess that's because there's tons of fun in working the "sexy" part of the engineering: deep understanding of optimizations and healthy dose of mathematics, chances to apply or even invent all kinds of engineering tricks, very cool results, potentially huge impact, cheering from a large community, and etc. Such projects are deep, are personal, and are super fun to work on. On the other hand, I'm not sure how many new fun projects (at least in the area of distributed systems) are out there that have outsized impact to the community, compared to 10 years ago.
A test to try out whenever these blog posts show up describing the burnout that is purportedly endemic to "open source":
If you ignore GitHub, how many of the problems described go away (and of the ones that remain, how potent are they)?
The problems are not intrinsic to "open source". Conflating GitHub culture with open source is like conflating being on Twitter with using the Internet.
Could you elaborate a bit? Do you think said "GitHub culture" is just because it's the largest platform, or more intrinsic qualities of the platform? Anecdotally I see many of the same problems elsewhere (GitLab, and even independent/self-hosted), but some things I can think of that might be more unique to GitHub are:
* The larger presence of commercial OSS projects setting a tone of "projects here are products to be marketed"
* Social media-like features encouraging popularity contests, curating a personal brand, etc.
* Both of the above leading to more users acting like you owe them something ("you're competing for my attention, right?")
I'm not 100% convinced that those or other GitHub-specific factors are primary causes of maintainer burnout, but I think it's certainly possible.
> Do you think said "GitHub culture" is just because it's the largest platform, or more intrinsic qualities of the platform?
Neither. Everyone and everything develops a culture and/or a set of norms. This is the one that emerged on GitHub. Looking for the reason inside the machine is misguided. It's the people.
> Anecdotally I see many of the same problems elsewhere (GitLab, and even independent/self-hosted)
Putting your repo and bugtracker somewhere that isn't github.com isn't sufficient to neutralize the GitHub culture. (Do people from country/culture X stop being affected by it when they go off and spend the summer in country Y?)
For many or maybe even most people doing open source on GitHub today, they're not even going to know which direction to drive towards in order to achieve the norms of pre-GitHub culture because they never experienced it.
I think the factors listed in the article will result in less commercial OSS contributions. Also as investment dries up, software folks will have more free time to do hobby development.
Reduced commercial interest may also allow new GPL projects (a.k.a Free Software) rather than the recent trend toward "commercializable" open source.
No I think open source has its place, especially for foundational stuff like compilers and OS kernels.
I’m less sure about the marginal benefit of more open source in the future. IMHO it tends to create an expectation that all interesting and challenging work should be uncompensated, and the role of most software engineers should be to just cobble together open source components to make some clueless SaaS founder rich.
> IMHO it tends to create an expectation that all interesting and challenging work should be uncompensated
The more significant contributor to that expectation is large commercial vendors using open source as a loss-leader or goodwill marketing. Look at the all the assumptions these days that maintainers are customer service, not just people doing their own thing in a workshop with the doors open.
You can add whatever you want in your license and none of it matters unless you are prepared to go to court. Even then, you might have a hard time getting the company to respond if they aren’t based in the same country as you.
For example, Onyx is a Chinese company that makes a line of e-ink tablets (Boox) that are based on open source software. AFAIK, they have refused to honor the terms of the GPL and release their modifications.
> You can add whatever you want in your license and none of it matters unless you are prepared to go to court. Even then, you might have a hard time getting the company to respond if they aren’t based in the same country as you.
It still matters for all practical purposes, IMHO. "Willfully violating terms of a license" results in very very different reputation for the brand than "Adhering to the terms of the license".
In practice, Amazon isn't going to take your "free for non-commercial use" software and try to sell it back to you even though they know you won't sue them!
Sure, some companies will do that, but lets be honest, if they are prepared to violate the license for FLOSS products, they'll violate it even if you didn't release it as open-source.
IOW, pirates gonna pirate; the license terms are irrelevant because they are pirates.
I think the main discussion here is big SaaS companies using peoples' FOSS without giving anything back, either feature or cost-wise. No solution is perfect, but at least this would give you the ability to push back on a company that is trying to "sell your own product" back to you.
Things would be better in the sense that markets would operate to determine the value of things. This is how the meatspace world works outside of software, and also how the software world worked prior to Linux and OSS. I remember paying for DOS and paying for Borland Turbo C and paying for WordPerfect and paying for Lotus 123.
> Things would be better in the sense that markets would operate to determine the value of things. This is how the meatspace world works outside of software, and also how the software world worked prior to Linux and OSS. I remember paying for DOS and paying for Borland Turbo C and paying for WordPerfect and paying for Lotus 123.
In an alternate reality, where there isn't, and never was, free software, the tech space would look very very different: for one, you wouldn't have a trivial product pulling in a tech stack of 20+ other products, because it would be too expensive to deploy if your TODO-ish application had to pay license fees for an OS, a VM inside that OS, a container, a manager for containers, a RDBMS, an interface to the RDBMS (few devs use a RDBMS directly, anymore) a VM (JVM/CLR/Beam/V8) and/or runtime for the language (glibc, etc), extra services (S3, OIDC, etc) a dev-env for the language (compiler/interpreter, IDE+LSP), libraries (+5000 npm deps), source control, CI, CD ... and maybe all those crap little YAML-based tools to do what Make does at various stages of development.
What we'd have is tight little software running with very few dependencies and a slick deployment that probably has no dependencies.
Of course, it would take longer to write, so we probably also wouldn't have 300+ similar TODO-ish applications. New Development would coalesce around actually useful products if producing an MVP took longer than 5m.
I think about this a lot. If all software cost money, would users be more willing to pay for it? If everything was a $50-$100 utility app, $300 for a professional tool, and say $1,000 professional software package like Photoshop, would we have better software? instead of having to find some weird way to fund app development, like advertising, you could just write good software and people would pay you for that directly.
I don't know that we'd not see dependencies though. It's sometimes better to pay for a good tool than home grow one, especially if that tool is outside your area of expertise and developing one would require a significant amount of resources. Eg a featureful VM hypervisor.
I use uncountable amounts of open source components that should cost me a minimum of something. If anything, at least enough to fund a project’s VPS hosting or annual ___domain costs.
have you looked into funding platforms for those projects? most projects on github have a "sponsors" tab that links to a variety of platforms most prominently github sponsors itself, but also tidelift, etc
Open source people don't know how to market themselves, and those who do are accused of exploiting FOSS for - GASP! - personal gains.
If every owner of every semi-important open source project set up a business license with support they would not only make money, but they could also leave their current job that holds back their development time and make an even better piece of software.
But no, most open source will always be considered as gratis and side-project worthy because the authors themselves treat it as such
That seems ludicrously low - Linux itself, if you value it at $1 per install is considerably more that 4 billion USD. The various open source web projects powering all the large sites are more than billions. Banks may not pay for SSL but they receive economic value. Etc. etc.
And it's not a labor surplus except for those poor souls that are trying to make their fortune with an OSS "side project." It is a method of doing things where results are shared and improved upon, fundamentally the scientific way brought into applications. Unlike physical devices, it's essentially free to share software. Just don't put access controls on the git repos.
There is a tendency to overly productize things, but that's not the heart of open source; the heart is just to share results, perhaps in a hope that ones labor can contribute to the progress of humanity. I don't need to promise to fix issues ever for my source code to potentially have value to someone solving a similar problem.
Now, it's harder to achieve this way of working in corporations, and things have to split into "open source with some modular hooks for private things that aren't open source" (even just in-house logging or metrics systems), but it's not impossible, and does have the same advantage science has - it grows exponentially and eventually will figure out how to do anything that is doable.