Email or username:

Password:

Forgot your password?
Baldur Bjarnason

“The slow evaporation of the free/open source surplus”

baldurbjarnason.com/2024/the-s

Where I try to explain, as succinctly as I can (which isn’t that succinct), why I’m worried about where FOSS is heading

66 comments
Chris Gioran 💔

@baldur

"[...]dollars of value for the conomy[...]"

I choose to believe this is not a typo.

Well done.

Baldur Bjarnason

@chrisg Thanks. The fix should be live soon 👍

ohmrun

@baldur There's odd branches of development are increasingly difficult to pull back into quote OS distros.
They can't package Deno for Debian because the Rust toolchain has spurred off to the degree that 600 packages would need to be set up for the dependency chain.

Baldur Bjarnason

@ohmrun That's a good point. That definitely has an effect.

ohmrun

@baldur The nix/guix//flatpak/snap/rust/npm toolchains depend on the distros but how much is getting fed back in?

Daniel Barlow
@baldur FYI the link to Ben Werdmuller's post doesn't work for me
xmlns="Dan"

I blame the decline of the #FSF. Users are using OSS because it's free as in free tech support, Devs are making it OSS to increase their userbase for VC.

Ask either group why their software isn't #AGPL3 and they would recoil with horror at the idea, but wouldn't really be able to articulate why not.

xmlns="Dan"

Just pressed send and realized I told a lie. If you asked Devs these days most would reply "GPL? What's that?".

Tim Lavoie

@divclassbutton On the other hand, the "Why AGPL" post makes for insightful reading, including the limitations of licenses against Service as a Software Substitute (SaaSS).

AGPL means that the behemoths have requirements if they want to use software licensed that way. This might be Kryptonite to them, but I don't know that that is a bad thing.

gnu.org/licenses/why-affero-gp

xmlns="Dan"

@tim_lavoie For sure AGPL3 is the best license, even if you don't care about the four freedoms.

1 - Unlike MIT, you are compatible with most other stuff, no need to worry about virality or derivative works, just set it and forget it.
2 - Because the only people using your stuff will be other hobbyist coders, you don't need to worry too much about backwards compatibility or support cycles or anything like that, tinker away breaking compatibility at will.

Audun

@baldur Good post. This has been a flaw with FOSS for a long time. You can't build a sustainable model on pure voluntarism. FOSS is a great idea, but it either needs public funding or a model where the users pay for the development. There are a few companies that rely on FOSS and also contributes to it's development, (ie the user pays model), but way too many just treat it as free (as in beer) software.
Public funding is probably a more reliable route. It could be a variant of user pay along with the "public money, public code"-principle, but (some) software should be treated as infrastructure that needs public investment.

@baldur Good post. This has been a flaw with FOSS for a long time. You can't build a sustainable model on pure voluntarism. FOSS is a great idea, but it either needs public funding or a model where the users pay for the development. There are a few companies that rely on FOSS and also contributes to it's development, (ie the user pays model), but way too many just treat it as free (as in beer) software.
Public funding is probably a more reliable route. It could be a variant of user pay along with...

Eaton

@audunmb @baldur the conflict between “free as in beer” and “free as in puppies” is real

Sven Slootweg

@audunmb @baldur There exist more models than just those two that are sustainable, including ones based on volunteer work - but you do need deliberate and sustainable (collective) management of the commons, and that is what is currently missing.

Cy
I've been worried about where FOSS is heading ever since they coined the term FOSS. The whole nattering on about software licenses is so tiresome. We need power, not permission. By giving permission ourselves, we tacitly agree to allowing them to withhold permission, giving them open license to unbridled exploitation.

No I don't have to declare my software open source, because all software is open source. You're just illegitimately bullying people into ignorance when you close it. Neither of us should be able to close our source code, whether we want to or not.

CC: @audunmb@todon.nl @baldur@toot.cafe
I've been worried about where FOSS is heading ever since they coined the term FOSS. The whole nattering on about software licenses is so tiresome. We need power, not permission. By giving permission ourselves, we tacitly agree to allowing them to withhold permission, giving them open license to unbridled exploitation.
Troed Sångberg

@baldur You're absolutely right. Just look at the age of open source developers and us GenX:ers who were early inte home computers pop out.

However - I think there is a solution and I think it'll happen for a host of societal reasons outside of open source needs. We will see a lot of countries implementing some form of Basic Income scheme in the somewhat near future.

More than a few people will be happy to live off that while poking away at their favorite piece of code.

Chris Johnson

@troed @baldur Pssst: open source predates GenX’s interest in computers.

Troed Sångberg

@cxj

O'rly? ;)

You're of course right, don't want anyone to believe I meant we aren't standing on the monitors of our ancestors.

@baldur

Chris Johnson

@troed @baldur 👍 Agree that UBI if implemented will be a big factor.

Sven Slootweg

@troed @baldur Out of curiosity, have you seen any concrete pointers on basic income (notably, *not* just welfare) being implemented in the near future anywhere?

Because I know that lots of countries are doing 'trials', but they've been doing that for several decades now, and then it never gets talked about again until the next 'trial' that has the exact same outcome.

ShadSterling

@troed @baldur why do you expect to see a lot of countries implementing UBI?
Any specific countries?

Phil Nelson

@baldur I work for OpenCV. It is tough out here

J Λ M Ξ S

@baldur Using AI to generate software also seems to push harder in the other direction because it seems likely AI companies will eventually claim ownership of whatever we create too.

mike bayer :python_logo:

@baldur "Most of the time, it seems to be a never-ending parade of angry demands with very little reward. "

I used to answer 100% of SQLAlchemy emails / questions. these days I try to answer as few as possible, hoping someone else can pick it up. when your project is new, your users are all enthusiastic, motivated people who picked you. but in 20 years, most of your users are people who were told by their bosses to learn it

matt wilkie

@zzzeek @baldur "when your project is new, your users are all enthusiastic, motivated people who picked you.  but in 20 years, most of your users are people who were told by their bosses to learn it"

^^^ this

matt wilkie

@zzzeek @baldur I puzzle recurringly on the nut: how big is too big, how small is too small.

Both ends of the scale are sucky. This truth seems to occur in every domain I look. It also seems to be very hard to impossible for people to know for themselves when they're at the threshold.

Hen Gymro Heb Wlad

@baldur Sounds realistic to me.

Good to see explicit recognition that "free" OSS has always been subdidised by developers (their time, their investment in education and skills) and industry (directly supporting FOSS projects, providing jobs where developers can maintain/extend their skills, paying enough for developers to have free time they can spend on FOSS, etc).

All of that ecosystem is being scoured away. I really see little future for most FOSS projects without strong industry support.

Garonenur

@baldur that sounds like a good point for general unconditional income for me.
Arts and many other things that make humanity and living in it nice would also benefit.

netsensei

@baldur A thoughtful post. Here's some good research: Roads and Bridges. The unseen labor behind our digital infrastructure. By Nadia Eghbal for the Ford Foundation. Written in 2016, it still holds lessons you mentioned.

fordfoundation.org/work/learni

Baldur Bjarnason

@netsensei Thanks! That looks interesting 🙂 👍

Chris Johnson

@baldur Really interesting topic. I find myself mostly agreeing with your points. A couple we might argue/discuss.

One, high interest rates, seems wrong. They are low (in the USA) and are heading lower. There were significant OSS efforts in the 1980s and the interest rates then were far higher than today.

Baldur Bjarnason

@cxj The context for that is that US interest rates used to be effectively zero until a couple of years ago which drove a lot of investment in tech. The change in behaviour in tech after 2022 when they were raised was really noticeable. It's "high" insofar as it's higher than zero

Sven Slootweg

@baldur You're certainly correct about the npm ecosystem as it currently is, not being sustainable.

It *could* have been, if a few different choices were made (notably ones rooted in governance, that have nothing to do with the technical decisions that people like to complain about), but they weren't, and so here we are...

Juho Mäntysalo

@baldur

Thank you, I'm not into coding at all, but your ideas seem applicable to other sectors of life.

I appreciate the help!

Donald Ball

@baldur @toxi This seems undeniably true and likely to only further aid the concentration of software development into the handful of megacorps that will fund their own internal ecosystems, will often rent their fruits to the wider world “as a service”, and occasionally offer a landmark project as actual open source, though whether these are gifts or traps is a question worth asking.

Donald Ball

@baldur @toxi The other color I might add is that, in addition to the rich California coder cadre, a couple of decades back at least, there was a cohort of developers doing open source as a hobby alongside college work, or even work outside the industry, back when the cost of living crisis didn’t require full-time professional employment and there was room for, well, slack.

In the US at least, that path has narrowed to a knife’s edge.

Karsten Schmidt

@donaldball @baldur I agree, but it's not just so in the US. Also here in Europe/DE it's changing, rampant inflation is ongoing and the knife's edge is approaching! Never felt that gloomy about FOSS before (in almost 25 years)! It's not just the financial side[1], also socially, interest/excitement/contribs/sharing/discussion with others has massively slowed down, and I'm worried for how much longer we/I can keep this all up... What will come next?! Proprietary and/or Copilot are not a basis to form a new culture of shared knowledge/learning & reciprocity, especially not long term... So I'm wondering, once defunded (monetarily, politically & socially) what will this all morph into, what else is there/can there be? Will there be FLOSS winter before the next AI winter? Sometimes I think the role of contempary AI, its ubiquitous hype and the redirection of funding/attention are just strategic means to an end to divide, conquer and ensnare...

@donaldball @baldur I agree, but it's not just so in the US. Also here in Europe/DE it's changing, rampant inflation is ongoing and the knife's edge is approaching! Never felt that gloomy about FOSS before (in almost 25 years)! It's not just the financial side[1], also socially, interest/excitement/contribs/sharing/discussion with others has massively slowed down, and I'm worried for how much longer we/I can keep this all up... What will come next?! Proprietary and/or Copilot are not a basis to form...

Héctor The One And Only

@baldur@toot.cafe

As open source software and operating systems gain traction I do have to wonder how long it will be before corporate interests take control and ruin them all. We have already seen how IBM has changed the way Red Hat does things and since the the licenses in question don't forbid what they have done, it makes me think that eventually all the open source will be behind paywalls.

Sherri W (SyntaxSeed)

@baldur Awesome article.

My hot take is that a decline in OSS participation will be a good thing. People need to re-evaluate what things are worth spending unpaid labour on. And companies need to stop receiving an endless supply of free labour via using FOSS software, tools & libraries.

Because as great as the philosophy is, it just results in a massive transfer of labour-wealth from skilled coders to savvy corporations.

Sherri W (SyntaxSeed)

@baldur The philosophy of FOSS works best when all parties are cooperative. I'll share my OSS library & you share some free sound effects you made, and that guy shares an icon set & that girl shares her plugins. And the companies who use these things give support in terms of $$ or staff who are given paid time to contribute.

But this isn't the world or economic system we live in yet so FOSS becomes a spinning top that is too heavy on one side. That wobble will just keep growing.

Don Marti

@baldur Good points. In the days of "create more value than you capture" a lot of startups were so profitable that they could afford under-utilized developers. Then came zero interest rates and "employee hoarding," when it was cheaper to for a big co. to keep someone on staff than to watch them go develop a competing product and have to buy them out along with their VCs. Now both productivity-boosting innovation and money are scarce, so developers are time-crunched at best, or laid off

Graham Spookyland🎃/Polynomial

@baldur another factor I have observed is that COVID and lockdown changed people's attitudes toward how much nonsense they are willing to tolerate. this has an impact both on projects and the software itself.

projects previously got away with onerous procedures and odious personalities being involved in the contribution process. people won't slog through it now.

the change in user attitudes has, for the most part, not been matched by a deeper focus on QoL bugsquashing and low-friction UX.

Graham Spookyland🎃/Polynomial

@baldur the demographic who are deeply invested in FOSS as a philosophy are generally still willing to make that trade, as they always have done. in practice, however, the majority of users are more casual fans of FOSS, who will seek out alternatives more quickly when things frustrate them, and who will accept commercial alternatives that are more comfortable to use. when the bar for frustration shifts, but the design philosophy of the software does not move to match, those users leave.

Graham Spookyland🎃/Polynomial

@baldur there are examples of projects that have made a concerted effort to solve these problems (KiCAD comes to mind, running user surveys and then acting on them rapidly) but they are a minority.

my personal view is that this is an endemic cultural issue that has plagued FOSS for decades, but I don't have the energy to get into this because it inevitably leads to tedious reactionary defensive responses.

Sven Slootweg

@gsuberland @baldur Don't know that it's worth much, but if it helps: I agree on all of this, and it has been frustrating me for over a decade.

(And I've experienced the same active resistance to changing that culture, too...)

Graham Spookyland🎃/Polynomial

@joepie91 @baldur the core problem, as I see it, is a culture of code being lauded as the be all and end all. writing code is directly equated with creating software. if you're good at writing code, you're smart, you're useful. if pushed to explain what else is involved, folks will mention documentation and maybe community management. if design is mentioned, it's exclusively about architecture, not about experience and usability. as a result, well-designed user experiences are vanishingly rare.

Graham Spookyland🎃/Polynomial

@joepie91 @baldur this also led to the long-standing cultural problems around generally awful people being tolerated, accepted, and even glorified, but that subject has been talked to death.

TSRBerry

@gsuberland @joepie91 @baldur could I ask you what I as a contributor/developer for OSS projects can do to improve things (even if that change might only affect a tiny amount of people)?

Graham Spookyland🎃/Polynomial

@tsrberry @joepie91 @baldur the answer to that heavily depends on what kind of project it is, who the users are, and what scale it operates at. if it's a small CLI tool developed by one or two people and used by a handful of linux sysadmins you probably aren't running into many UX design pitfalls. having really good docs is nice, having error messages provide context and guidance helps, being welcoming towards contributors and not needling them with endless style changes is very helpful.

Graham Spookyland🎃/Polynomial

@tsrberry @joepie91 @baldur if you're operating at a larger scale (especially on gui tools) then running annual community feedback campaigns is a huge winner. make it a simple survey form - minimum friction, maximum response rate. people will tell you all the things that have been bugging them but never got around to opening an issue for. commonalities in responses will help you direct your efforts and make a lot of users happy with relatively low effort.

Graham Spookyland🎃/Polynomial

@tsrberry @joepie91 @baldur but ultimately at that scale you need input from actual designers - people who really know how to make a good UI/UX, and who can tell you what you can do better. that can be direct, by getting designers to come help with the project and help come up with better approaches and workflows, or it can be indirect feedback. and be humble. don't get defensive because you're attached to the project or it'll be work to fix the issues. it's hard work to do it right!

Sven Slootweg

@gsuberland @tsrberry @baldur One thing I'd want to add to that is that if it's not feasible to find experienced designers (funding reasons, community politics, whatever reason), it *is* possible to just learn how to do this stuff yourself on-the-fly as a maintainer, but it makes the "be humble" part all the more important, and to make that work, you *need* to have an attitude of "if a user complains about something, that means there's a bug, even if it's not where the user thinks it is" (eg. it might sometimes be a documentation issue instead of a UX issue, but as long as people complain, there's *a* bug somewhere).

@gsuberland @tsrberry @baldur One thing I'd want to add to that is that if it's not feasible to find experienced designers (funding reasons, community politics, whatever reason), it *is* possible to just learn how to do this stuff yourself on-the-fly as a maintainer, but it makes the "be humble" part all the more important, and to make that work, you *need* to have an attitude of "if a user complains about something, that means there's a bug, even if it's not where the user thinks it is" (eg. it...

Sven Slootweg replied to Sven

@gsuberland @tsrberry @baldur Also, related to the 'user surveys' point - it's surprisingly helpful to just enter your project name into a search engine every once in a while, and read every blog post, Reddit thread, and rant you come across. Those tend to be *full* of things that never made it into 'properly filed' issues, and give a good impression of the general atmosphere around your project, because people tend to speak more freely and directly.

Of course that does mean that you should refrain from responding to them unless you are absolutely 100% certain that you can do so non-confrontationally, because seeking out confrontation with people talking in their own spaces is (deservedly) a very quick trip to the trash pile. Better to just take notes and work on fixing their complaints in the background.

(Bonus points: you can do this one as an 'unaffiliated' contributor with no formal power within the project. Whereas a proper user survey, although it can have better coverage, tends to require speaking formally for the project.)

@gsuberland @tsrberry @baldur Also, related to the 'user surveys' point - it's surprisingly helpful to just enter your project name into a search engine every once in a while, and read every blog post, Reddit thread, and rant you come across. Those tend to be *full* of things that never made it into 'properly filed' issues, and give a good impression of the general atmosphere around your project, because people tend to speak more freely and directly.

TSRBerry replied to Sven

@joepie91 @gsuberland @baldur Thank you both, this helped a lot! :D
I'll note these down and try to do my part!

Fiona

@baldur@toot.cafe I see your point, though I feel the article leaves out an important counterpoint: When the goal isn't to sell software (aaS) but to have well working software for whatever a business does, it's simply the most efficient way (at least in the mid- to long term) to cooperate with others with similar interests. A client I'm working with right now on embedded system firmware isn't worried about someone else benefiting from Buildroot, kernel, etc. patches I'm writing, they're (rightly) concerned about the long term effort of maintaining a growing patch set and having to continually rebase it on new releases, so they explicitly want me to upstream as much as possible. Sharing reduces the local maintenance burden, and spreads the general one on more shoulders.

That's unlikely to save npm, but it's a much more positive outlook for those of us who work on solid infrastructure rather than flashy stuff and the latest hype. ​:neocat_floof:​

@baldur@toot.cafe I see your point, though I feel the article leaves out an important counterpoint: When the goal isn't to sell software (aaS) but to have well working software for whatever a business does, it's simply the most efficient way (at least in the mid- to long term) to cooperate with others with similar interests. A client I'm working with right now on embedded system firmware isn't worried about someone else benefiting from Buildroot, kernel, etc. patches I'm writing, they're (rightly)...

stilescrisis

@baldur @mcc I feel like there's a missing link in articles like these; I worked on open source for a number of years that was fully funded by FAANG (Skia, the 2D rendering engine). This isn't an isolated incident--Google, Apple, Microsoft, etc. actually fund a lot of OSS work that is truly free in both speech/beer contexts. It's not all hobbyists by a long shot.

Paul Cantrell

@baldur
Compelling.

This side branch deserves its own writeup:

“Some are reaching for LLM-generated code before they even look for an OSS project, both disconnecting those projects from opportunities to grow a sustainable community and nullifying the strategic advantage of having made an OSS solution for a problem. Note that the models are originally trained on that OSS.”

Baldur Bjarnason

@inthehands It’s still purely anecdotal on my part, based on conversations and observations. It’s still to early to be sure, but I think that for some people, LLMs are fulfilling demand that OSS used to. The question is just how big that group is.

Paul Cantrell

@baldur
That part is anecdotal, and it’s an open question how long lived it is. There’s also a family of partly theoretical questions: What are models trained on once people start using models to write code? Do they propagate expertise from experts to newcomers? Do they create a kind of model in breeding and degenerate? Do they •prevent• the creation of new tools?

I have serious doubts about the sustainability of even the current very mixed success of LLMs for code.

@baldur
That part is anecdotal, and it’s an open question how long lived it is. There’s also a family of partly theoretical questions: What are models trained on once people start using models to write code? Do they propagate expertise from experts to newcomers? Do they create a kind of model in breeding and degenerate? Do they •prevent• the creation of new tools?

acb

@baldur @glyph Not sure about the point about laid-off/unemployed engineers not coding. I have spent periods of funemployment between contracts/jobs on side projects (a few apps/frameworks, some playing around with various technologies/languages and such). Looking at the fediverse, I don’t seem to be alone in this. Perhaps better social security such as a UBI would lead to more open source contributions.

Glyph

@acb @baldur I think the truth is nuanced (and I would love a heavily footnoted version of this article with stats! But that is probably a huge amount of work) but I am an interesting exception that proves the rule: my independent crowdfunded maintenance work pays about 0.01% (yes that is the correct placement of the decimal point) of the commercial/proprietary work that I do. That number has been climbing but it is a *steep* hill before this makes any sense whatsoever

Sheogorath 🦊

@baldur I would argue the industry didn't overspend, but overbuilt on FOSS. FOSS was free, is free and that encourages usage. But as all infrastructure, even free one, there is a maintenance price tag and the people with the bills are starting to knock on the door.

As a result: panic.

Not to mention that governments start to pick up on the transparency FOSS provides for regulations.

Go Up