Few people intuitively understand Copyleft. Especially in all peripheral parts of the software development industry, people are (painfully) unaware, but more importantly, I seldom see someone showing a deep understanding of Copyleft.
But Copyleft, despite many now disagreeing, makes the world go round. “It eats the world,” as a loaded baldheaded white man would say. So why is it that the ideas of Copyleft and, more importantly, its sound understanding is lacking in so many knowledge-work-based and creative industries?
In the following article, I want to outline the importance of Copyleft licensing in the software development industry. I’ll purposefully conflate the ideas of Copyleft licensing with other, higher-level, open-source software principles too. I’ll do that intentionally to paint a decisive picture for everyone working outside my industry, towards a better understanding of Copyleft - e.g., in the music industry.
Please “sample” my code!
Contrary to, e.g. musicians, software developers happily upload their painstakingly-created code to public platforms. Consider this; software developers are among the best-paid knowledge workers globally. In places like Silicon Valley, they take home unheard-of paychecks and equity.
Still, if you know one, you’re likely aware that on some Saturdays and Sundays, they’ll cancel your brunch dates to instead sit in front of their computer just to fix this one bug in their open-source library.
If we were to put a nominal value on the amount of free work being pushed online to GitHub; If we summed up all the accumulated hours of software developers dwelling on code in their undies on the weekend - just to give it away “for free,” every business person and their assistants would be losing their minds! It’s anarchy! How can they do that?! Aren’t they afraid of someone stealing their work??
But that’s where I think it is important to understand that coming from another industry, e.g., music, how software developers think when shipping open-source code has to be considered vastly different from, e.g., musicians.
While Kanye et al. may go complete batshit about seeing their work sampled without prior legal agreement and moral attribution, as a software engineer working under the principles of Copyleft, my highest goal is “developing a sample and then promoting it such that everyone uses it.”
Software, being, fortunately, a rather recently developed art, gladly had genious contributors like Linus Torvalds, Richard Stallman, and others who made sure that attribution (git and package managers), licensing (e.g., GPL), and free speech are built into the free (as in libre) software production stack.
In open-source software development, developers want their code to be “sampled” (or referenced) as broadly as possible. It furthers their brand and credibility; it’s their default as proper attribution is always given.
So it happens that when engineers write code “for free” and everyone uses it, a rather direct incentive comes into play: they are invited to join higher-paid or higher-status jobs. They earn more money, are more popular, and sometimes, they even have to work less.
Open-source software developers like mxstbr, the creator of the react boilerplate, have made a professional career out of a single significant open source contribution. That guy wrote code that 1% of the worldwide web now uses, and he probably never made a penny directly from that react boilerplate OSS.
Brand-building and Copyleft are better than Copyright.
Sure, particularly business types in the software industry may now say that with open source software, a company cannot sustain its business. They’ll enumerate countless examples where OSS hasn’t worked out. Or, they’ll list the numerous instances where software is exclusively reserved under copyright laws and generates shittons of cash.
And it’s probably true: More money is directly made from exclusively copyrighted software.
But for me, as an individual, I have particular aims for my career, those being the freedom to explore my curiosity while still making a good salary - building a brand and contributing to the global catalog of OSS works well.
Moreover, I know for a fact, though, that there are many others like me profiting from their work being openly accessible. After all, the world runs on open-source software in 2022. There’s no app, no website, phone, or car that isn’t using the undies-generated “free” code. As much as you cannot escape technology’s accelerationism, for sure, you won’t be able to run from having to use open-source software.
Today, writing open-source software and building an online brand is a well-paid and prosperous career any engineer should consider.
OSS is speech, and speech doesn’t have a price.
But developing open-source software goes far beyond just brand-building and making money as an individual. Albeit cynically (and falsely) derided, Stallman’s Free Software Definition and its proclamation of software being “free” as in “free speech” (and not “free” as “free of charge”) is paramount to creating a plane of freedom for all contributors.
It is here where I reference back to the article’s beginning and when I said that few truly understand the power of Copyleft. On other occurrences when it was clear, I’ve tried to make a similar argument, e.g., when Facebook set out to build the metaverse, and I said that a stock company isn’t capable of creating a public good like the metaverse. Only a FOSS approach could achieve that.
Copyleft software licenses, particularly the GNU General Public License, allow software developers to create a work, or a social architecture, much greater than their individual contributions.
The GPL license allows each contributor to fully retain their Copyright while making their work available under viral conditions. Any other user is permitted to use the GPL-licensed software in other GPL-licensed software, while the license specifically disallows distributing GPL-licensed code in proprietarily distributed software.
The beauty of this construct is in viewing every single contribution of a repository as a GPL-licensed commitment from distinct individuals. It is this viewpoint that makes clear the mandatoriness of GPL’s virality and the incentives it creates.
A GPL contributor retains Copyright, and since their contribution can likewise only be used within GPL-licensed, an old code repository can grow into a convoluted monster of individual code donations.
It is this property that makes contributing to knowledge work under Copyleft so powerful. Because similarly to how an employee might work towards furthering their company’s interest in increasing its stock prices, a well-setup GPL project creates interest in fostering its popularity and technological sophistication for all its contributors.
Just imagine the over-poweredness of a GPL for music. The virality effect of sucking all musical primitive inventions into a global repertoire of sounds and samples.
But back to software and since, e.g., GNU/linux’s repository has accumulated years of work from a maximally diverse set of agents, all licensing their work under GPL; it is maximally difficult to capture and hence a direct threat (and blessing) to the world’s most powerful corporations like Google, Microsoft and others.
Sure, we could ask everybody who has ever contributed to Linux if they’d be interested in selling Linux for a bazillion dollars to Microsoft: But the point is that reaching a sale-positive state in this vetocracy of GPL contributors would literally be impossible.
OSS is speech, and speech doesn’t have a price. For artists, like musicians, whose work is also often speech, their work too - should have no price - and they should be paid like software engineers. So we should work towards making that happen.
Update from 2022-08-07
Vince from the AI Alignment and Rationality meetup in Berlin made me aware of an interview of Lex Friedman with John Carmack where they end up mentioning the same phenomenon too. Timestamped-link to Spotify. Thanks for sending it to me!