Comment by scottrogowski | Hacker News Reader

Comment by scottrogowski | Hacker News Reader

Parent Full thread

This article takes a purist stance driven by a rigid adherence to ideology. Let's look at this another way...

Open-source and proprietary licenses are at two ends of the software development spectrum. The open-source model maximizes ease-of-adoption but doesn't provide much incentive for the developers. Proprietary software provides a lot of incentive but adoption can be slow and burdensome.

Let's assume that a good goal for society is to maximize the rate of innovation in software. To do that, you need a mix of BOTH ease of adoption and suitable development incentives. Source-available licenses are an attempt to accomplish this.

Is this a perfect solution? Probably not. I think better licensing models are still waiting to be discovered.

However, my sense is that these new licenses will accelerate the development of software with limited downside for the user. After all, they are designed only to impact companies attempting to sell a SaaS.

In addition, they have the potential to weaken the tech monopolies which, in my mind, is a Very Good Thing.

> After all, they are designed only to impact companies attempting to sell a SaaS.

Well, that’s the PR message associated with the new wave of source-available licenses (source-available licensing is not, itself, new; its long been an established form of proprietary licensing.) But it doesn’t hold up: you can’t harm competing services providers without harming end users. There is a reason why the very different ideologies of the Free Software Movement and the Open Source Community nevertheless have stably settled on definitions which are virtually identical in practical applications (and even though those communities have very different preferences for licenses within the scope meeting their similar definitions.) It is because the space is not a continuum, and there is a minimum needed in each of a number of axes of liberty for the whole structure not to collapse into something which either community prefers free/open licensing. Particularly, without robust freedom that protects what other people can do with it (including there ability to sell you services built around the software that the original maker might also want to sell), you are not insulated against future actions of the copyright owner restricting the software or its or others services around it.

And this isn’t opaque to the people issuing these licenses; the overt motive is to enhance monetization by preventing licensees from competing with them to sell services: it is to create a moat enabling monopolization and monopoly rents. That’s the explicit idea: to create lock-in that free/open licenses would not support.

The benefit that the licensors seek directly depends on the harms that extend beyond competitors to end users.

> you can’t harm competing services providers without harming end users.

> the overt motive is to enhance monetization by preventing licensees from competing with them to sell services: it is to create a moat enabling monopolization and monopoly rents.

Here's the problem: Amazon, Microsoft, and Google are already able to charge something approaching monopoly rents because they have built a moat that others can't dream to cross. Heck, try to even build two services that talk to each other between the moats and you will get hit with massive data-egress charges. And they do this by monetizing free products with no compensation to the developer. If you are Redis, your hosted product, whose core you developed, is less competitive than that of the cloud providers who didn't contribute anything to the project. At the same time, the parent tech behemoths are engaged in all sorts of other bad behavior that only massive tech companies can get away with.

This state of affairs seriously threatens the open-source ecosystem and the source-available licenses are targeted to righting this imbalance. Again, are they perfect? No. If you are making money selling the software without modification, is it reasonable to ask for royalties to go to the developers? I think, on balance, that is reasonable.

If not source-available licensing, what is the better solution to the problem?

> Here's the problem: Amazon, Microsoft, and Google are already able to charge something approaching monopoly rents because they have built a moat that others can't dream to cross.

Even if that was true and they acted like a coordinated oligopoly rather than actively competing, including on price, more monopolies in adjacent markets make things worse, not better, for everyone (customers especially) except the additional monopolists.

> If you are Redis, your hosted product, whose core you developed, is less competitive than that of the cloud providers who didn't contribute anything to the project.

Yes, that F/OSS licensing advantages the firms that have the best infrastructure for monetizing services has been widely observed since it was just Free Software because OSI didn't exist yet, before cloud dominance that was mostly about professional services. And, yes, proprietary (and sometimes source-available) licenses which don't deliver the value of open source to end users have always been the “solution” to that.

What's novel recently is a bunch of firms ignoring well-known facts about the economic incentives around F/OSS, trying to build rapid-growth startups around F/OSS with no plan at all to address the strengths necessary to monetize F/OSS, then whining about the well-known advantages of established firms to pivot to proprietary licensing, then having the unmitigated gall to pretend their proprietary licenses provides the customer benefits of F/OSS licensing (which center on the fact that F/OSS doesn't create a vendor monopoly), despite the fact that not providing those benefits is central to the business reason for which they have adopted the proprietary license.

If these licensing models were worthwhile, they would be able to stand on their own merits. I've no objection to someone coming up with a new licensing model and trying to show that it's a good way to make software, good for society and all that. What I object to is freeloading on the success of the open-source model by obscuring the differences between your model and that model. If you're genuinely trying to pursue a new model that you think is worthwhile, own it! Give it a name that doesn't sound like "open source", that could never be mistaken for being the same thing as open source.

https://commonsclause.com/

Second FAQ: "Is this “Open Source”?" Answer: No.

I don't think they are trying to hide anything and if anything are being generous because the source-available licenses do have most features of the open-source model with the narrow SaaS restriction.

> After all, they are designed only to impact companies attempting to sell a SaaS.

That might be the intent, but the problem is that integrating any non-free software to a bigger project becomes problematic because the combined project likely becomes non-free. Which is most apparent in that these non-free components will not be appearing on any major distros main repositories.

So the non-free software remains an isolated island rather than melding smoothly into the larger ecosystem. And this has impact well beyond some SaaS providers.

It doesn’t seem all that purist? It describes the situation, but in the end it says to decide for yourself what you want to do.

There is some mild caution about knowing the differences between these licenses, and mild support for open source projects that aren’t controlled by a single vendor.

I disagree. I personally would prefer to license the code I write myself with a GPL copyleft or a 'no commercial use' type of license. However, I license it instead under MIT, specifically to make sure that your average corporate user will be ok using it because:

1. I would prefer that it be widely used. Not because I am seeking clout or advancement, but because that's why I share it. 2. Sharing benefits everyone, including me. Fragmentation and bureaucracy harms everyone, including me. 3. I don't support monopolistic practices by large tech, but this is not the way to stop them. What we had before widespread free software was worse than it is now, arguably held back human progress for years, and didn't stop Microsoft one bit.

Reply on news.ycombinator.com