This is going to be a draft of a blog post I intend to write. I want to post here first to see if anyone resonates with what I'm talking about. Please comment if you identify with this, or know another name for what I'm going for.
You may already be familiar with that XKCD comic. Number 2347, "Dependency."
It captures a peculiar phenomenon of the internet. Much of it is developed by big companies. From the Cisco switches rerouting packets to the YouTube javascript running on hundreds of millions of devices right this moment, the internet is dominated by corporate tech.
But the thing is, despite it's massive impact, big tech is not the whole internet. In fact, we know that the internet wouldn't work properly if it was only the product of big tech. You need the wikipedia volunteers. The open source programmers. The hobbyist web-tool makers. The free ad-block maintainers.
These are the municipal volunteers of the internet. The roadside litter cleanup crew. The mural painters. And I think we all know that the internet would be a horrible place to live if it weren't for these people. The profit motive that fuels big tech leads to greed, closed-source code, and walled gardens. Every company would need to write their own crypto libraries. There would be no ad-free online encyclopedia. There would be no neal.fun.
I'm going to suggest a name for these people. "Netkeepers." A person who sees a problem on the internet that needs fixing, and fixes it.
But what motivates the Netkeeper? Not the profit motive, obviously. So what then?
Let's do a case study from my own life. I would describe myself as a Netkeeper. One example of something I've done is Blamscamp, a bandcamp-style player that you can embed on itch.io to sell your albums. I made this when it was announced that Epic Games bought Bandcamp. I saw that the internet had a problem—that bandcamp was a single point of failure in the selling-albums-online space—and I did my best to fix it.
I did this because I want a better world. I don't want it to become impossible to buy albums anymore. And if I can do my part to steer us away from that future with two days of gratis hypomanic programming, then I'll do it. I don't need to get paid for it.
But there's the problem. I do this because I believe in it, AND I do it for free because I don't need the money. I don't need the money because I already have a job that pays well. I don't need to follow the profit motive because I'm already comfortable. Capitalism isn't actively threatening my existence, so I have the luxury to spend time doing stuff I find interesting and important for free.
If I didn't already have a stable income, or I had a job that took all my energy and left nothing for myself, I wouldn't have been able to create Blamscamp.
And this is why this post is titled "The Netkeeper Needs to Exist But Can't." I'm not actually a Netkeeper. I'm a hobbyist Netkeeper. A true Netkeeper would be someone who could do it full time. Who finds problems on the internet, fixes them, and is paid for it. Not paid for each problem they fix, but paid unconditionally. Paid because the community has trust that you'll do the right thing with the energy you have.
So, can a true Netkeeper exist? Maybe. The only obvious option to me is Patreon. However, how do we describe what we're doing? If our Patreon says we "do random stuff for the internet," will people actually donate?
And this is where I reveal this is actually a manifesto. I feel like Patreon could work. We could make the Netkeeper real. But we need a better phrase than "I do random stuff for the internet." I suggest we popularize the word "Netkeeper."
Imagine a Patreon where someone proudly describes themselves as a Netkeeper. Every month they post updates of what they've been up to, like a Tom Scott newsletter. You pay them because you believe in them. You believe they can steer the internet in a better direction with the time, money and energy.
There could be directories of Netkeepers. It could be your job title. It could be how you spend your career break, gap year, or retirement.
But the Netkeeper can't exist. Not right now. The concept is knocking on the door to our reality. It wants to come in, it wants to become real. The Netkeeper needs to exist. I think we can all agree on its value. Maybe not on the execution, that could use some ironing out. But the concept, the concept is perhaps something we can believe in.
the really fun thing about the “free software is load-bearing” is knowing all the places it pops up that you really wouldn’t think of.
There’s the common knowledge ones, that the PS4 and PS5 run on top of a BSD variant (I think FreeBSD but I could be wrong), or that Mac OS is derived from a BSD, or that netflix’s servers all run BSD, and I would be remiss to mention Android being Linux-based, but then there’s the things you don’t think about.
Linux has been present in things like set top boxes and DVD players as far back as the early 2000s. Every consumer router/modem/wifi AP runs a variant of Linux (some even come with literally just OpenWRT and maybe some vendor customisations), and most of the enterprise ones do (Cisco Meraki products all run a derivative of OpenWRT, even). Juniper equipment runs their FreeBSD derivative JunOS.
Any Intel-based machine that has the management engine coprocessor built in? That runs MINIX, which gave Andrew Tanenbaum a huge ego. You may also be surprised to learn that regardless of what your phone itself runs, the cellular modem inside it is probably running Linux! Hell, the PinePhone demonstrated this wonderfully when someone found a vulnerability in software running on the baseband, and was able to get root on that. My understanding is that modern networking cards (WiFi/Ethernet) all also run some form of embedded Linux, though I don’t remember where I found that out.
That’s just conventional operating systems. In microcontrollers, there’s FreeRTOS, Zephyr, NuttX, RT-Thread, a bunch more, which all can be found in microcontrollers used in commercial products, from the immediate thought of IoT devices, to stuff like dishwashers. A market study from 2019 done by “AspenCore” showed that 21% of respondents were using embedded linux, followed by 19% using an in-house or custom OS, then 18% using FreeRTOS.
The very funny part comes in when an open-source product is used as part of bootstrapping something commercial and proprietary. In a past job, I worked on the software security team for some widely-deployed Windows-based embedded machines, but those machines needed disk encryption beyond what was available in Windows. The solution? A linux-based preloader environment that boots first to do all the setup.
This was just an overview of operating systems. The number of commercial products and proprietary software out there using free software as a component is nearly immeasurable, and you might not find out until something breaks or a license change comes up and someone from a company using the software opens a bug report to complain.
if I were paid a fraction of a percent of the value that my volunteer labor has produced for private corporations, I would easily be wealthy enough to never need to work again. and what would I do with my spare time? write more software to benefit everyone! but instead I have to split my time between open source and internal work at an evil corporation, and even that is a pretty lucky state of affairs relative to most maintainers
i just woke up and i'm still horizontal but lets catch everyone up: this is part of a long tradition of healthy conversation about the sustainability of FOSS funding and the volunteer maintainers that keep the internet running. the paradox is that where the very same economic system that benefits from free software fails to provide for those who create it.
the conversation around FOSS funding
foundations like Linux and Apache Software provide another model to consider, but 2/3 of the Linux Foundation's $6M/year budget comes from 8 individuals. don't get me wrong; i like the NetKeeper concept. but we live in an expensive society that works very hard to bog us down.
along with the software sustainability conversation we also have to talk about the maintainers-- how they're getting paid and how they're avoiding or healing from burnout. @zkat had a really good post some years ago summing up a system involving licenses and Patreon (or similar) to bootstrap your opensource work. unfortunately it doesnt always pan out like this.
other platforms like Open Collective and Github sponsors exist but you have to do all the legwork required to actually attract and persuade funders; finding the kind of deep pockets that can actually make your FOSS income survivable is another beast.
the conversation around getting companies to fund FOSS
if you want to get companies who benefit from FOSS to pay for it (good luck!), you have to consider the mechanisms through which you can do that: previous work trying to deal with this has focused on incorporating FOSS funding into corporate accounting, and using licenses as enforcement (both pretty lofty goals given the current landscape imo). licensing also requires a lot of work to uncover loopholes. policy work could help change the landscape but i haven't seen any meaningful considerations in that direction.
the open source projects that have been the most successful in drumming up funding from big tech are the ones that have had entire companies, often VC-backed, built around offering services and enterprise support. given the current economic situation, in which borrowing money with leverage is no longer practically free, it is quite likely that the current era of VC-funded open source technology is over.
so what's left? often it's literally just having to carve out time to ask people for money. i currently participate in a volunteer software project to provide data tools to tenant organizers and we're in the middle of begging funders for money; our work is inherently adversarial and if landlords decided to assemble against us, their capacity of financialization with their collateral means a wildly magnified resources would be at their disposal. and have you ever been or seen a core maintainer of a project that caught the receiving end of a conspiracy theory or misguided online moral crusade? or literally just any popular media software? a thing about working on the commons is that, despite how rewarding it can be, having to deal with the public also has the capacity to really suck.
the visibility of digital labor
which also brings me to a related topic: unseen digital labor. the people doing the moderation work to maintain their discords, the cleanup work people running a community have to field after instances of harassment. the posts people write to educate the ignorant. the people writing committee letters to try and ask members of a project for accountability, or the glue work required to get your project's headass BDFL to sit the fuck down.
[edit: a comment on the OP mentions the similarities between this conversation and the digital labor that content creators do. this of course is a whole topic in and of itself]
it's probably not a surprise to anyone here that this work is and always has been deprioritized. it's disproportionally carried out by people in marginalized groups. the first step to addressing any of this is to make sure this on-the-ground community labor is seen and valued alongside the technical contributions that's needed to make all our digital commons infrastructure happen.
prior work
- 2007 review of business models https://eprints.soton.ac.uk/263925/
- kyle mitchell's licensing experiments: https://medium.com/licensezero/the-license-zero-manifesto-fecb7aaf4c0a
- zkat's FOSS business model post https://dev.to/zkat/a-system-for-sustainable-foss-11k9
- anything lisa nakamura has written or spoken about digital labor https://youtu.be/LN9h6ldeVdI
- the Free Software Foundation Europe (FSFE) https://fsfe.org/news/news.en.html
I know I just got through complaining about people not putting any thought into where their money is going and I stand by that sentiment but also:
it's maddening being aware of the gaping void in all these conversations where things like "state funding" or "worker organization" used to be
not to be like "we used to be a real country" but, didn't we? we used to have funding for both science and the arts. even if you're not as much of a sicko for central planning as I, a communist, am, surely it's apparent that there's a bunch of critical goods for society that go beyond the liberals' beloved "firefighters and roads" that The Free Market just can't handle on its own? why shouldn't that be handled through a state authority? and tech companies get state funding, which, to quote The Grapes of Wrath, "ain't that relief?"
or hell if that seems too inconceivable, here's another perspective: if FOSS is really so critical to the infrastructure of everything... that kinda sounds like leverage to me. it kinda sounds like an arsenal. like if there was an actual conception of netkeepers as workers whose labor is being exploited by the corporations that depend on FOSS, you would immediately have some pretty impressive power to shut shit down.
"this is naive for x y z reasons" yeah sure fine, it's as naive as saying "comics creators should unionize". it's more complicated than that, obviously, particularly because a lot of these are fields that CAN'T unionize in a traditional way per US law. this isn't me producing One Simple Trick to fixing all the problems, that'd be silly.
but 30, 40 years ago, the possibility of state intervention and worker action was part of the conversation on issues like this... and now it's just completely absent. I think that's also pretty silly!
if FOSS is really so critical to the infrastructure of everything... that kinda sounds like leverage to me. it kinda sounds like an arsenal.
This is where it boils down to me more and more as I get older.
Programmers successfully liberated the means of production, but then got convinced that the noble goal was to then go back to work for the bosses for free.
It's insanity. "Free software" is one of the greatest pieces of doublethink to come out of the 20th century, and also one of the most successful. It's literally some "freedom is slavery" level rhetorical doublespeak, and it fucking worked, and now it's literally goddamn killing people.
But if you put out your code with a non-commercial license, a million nerds will firebomb your website.
