Hacker News new | ask | show | jobs
by pfraze 2108 days ago
Something which people don't always recognize about open source is how "value-capture" works for creators. FOSS eschews direct monetization (selling licenses to the core product) based on an ethical premise (it's unfairly limiting to end-users) and/or a business premise (the ecosystem is a key to the value prop and licenses would slow down the ecosystem growth).

None of that means that the creator isn't capturing value. You capture social credibility and market awareness which you can convert in a variety of ways, including monetarily (by selling complementary products/services, by donation models, or by getting a job that you might not have had the career-credentials to get otherwise). As an aside to elaborate on this point: a lot of the recent debate about paying FOSS maintainers has to do with projects which realized all the potential social value-capture, and left creators with an externality of maintenance.

Intuitively, I think people understand that forking and rebranding a project without a really strong motivation can be scummy, but I don't think people can verbalize why. This is why: you're attempting to steal the upside which the creator is the in the process of capturing.

And FWIW, anybody saying that a blogpost is weak action and you ought to be going to court is ignoring that, when the value you're capturing is reputation, then public discourse is the tool you want to be using to manage it.

2 comments

I think that's a perfectly good point. The ZSF wants to build momentum and have the language take off. We also are looking for money in order to pay core contributors.

I can also share that my own role at the ZSF has as ultimate goal of increasing the total amount of effort spent on Zig.

The core problem with what kristate is doing is that, in a moment where the Zig community needs community members to take initiative and build things around Zig, and where likewise there's a tremendous potential for those individuals to capture a good chunk of value for themselves, all while making the pie bigger, he instead chose to cut a slice and run away with it for what amounts to pure vanity.

Right now we are trying to encourage people to start thinking about building a business around Zig, be it programming in Zig, or writing a Zig programming book, or whatever. I personally started https://zig.show before coming on board and I plan for my own, independent, Zig-related activity to become my main source of income one day.

If you add on top of that, that the guy creating this useless fork has a history of offering a free wifi service that steals personal information and rewrites amazon affiliate links, then you can see why we want to put as much distance as we can. For context: https://internet.watch.impress.co.jp/docs/news/496423.html

I've dealt with this kind of thing before, it's unpleasant, and I wish you the best.
Also the back and forth seems to be

Andrew, et al: "you're being an asshole"

Kristopher: "we're not doing anything illegal"

Which to an outsider just reads like a tacit admission of being an asshole.

If being an asshole is something the Zig community does not want people to do with its code, why not just license Zig in such a way to prevent this type of usage of its code?

They already implement a code of conduct to restrict behavior in their community.

Is it necessary to go through all that work to re-license, though? I don't see the problem with how they're handling it now.

The person in question is free to use the code and be an asshole. Everybody else is free to call him out for it.

So in general yes, if you don’t have copyright assignment it is a lot of work to relicense. But that is not necessarily the case here... things like the MIT license (which ziglang uses) are often so permissive that you can relicense them as GPL.

With that said this sort of thing can result in mass contributor exoduses. One might prefer an in-between license like a per-file copyleft as you see in the Mozilla Public License, which is still GPL-compatible but does not put the full onus of the GPL on the software, solving this problem in a more narrow way (connectFree has to keep their proprietary software separate from ziglang when combining the two together, which forces them to be much more transparent about their value-add—things which they cannot keep separate need to be contributed back upstream).

Because then it wouldn't be open source. The open source definition prohibits developers from placing a lot of arbitrary license restrictions. One of the examples of this is that Gab, an alt-right site, now uses Mastodon, a platform written and used largely by a pretty far-left community.
We can get into semantics on what open source means, but I don't think the definition has to include a prohibition on any licenses. Open source just means the source code is freely available.
Except, we do have a rigorous definition on what Open Source really means. The best place to check is the OSI's definition[0] and the licenses they approve[1]. What you're describing is commonly described as Source Available and doesn't readily meet the definition of Open Source.

Our terms matter. And considering this conversation is really about what it means to be OSS and what the tradeoffs are, we need to "get into semantics on what open source means". Too many people co-opt the term and we need to defend it.

[0]: https://opensource.org/osd-annotated

[1]: https://opensource.org/licenses

I reject the OSI's claim to have authority over what "open source" means. We do not have a rigorous definition of what open source means beyond the source code is freely available.
That sort of licence is not helpful. It would be unenforceable in practice and run afoul of DFSG and similar and prevent inclusion into distros.
Im not a lawyer but I’m fairly certain it’s possible for the Zig foundation to craft a license that reduces Zig’s anxiety w.r.t. malicious actors and that still satisfies DFSG. The GPL is an example.