4 ms·
A potentially nice idea with several things to think about to real-world proof it. Let's think about nodejs packages because of colors and faker. Some random to
by pmontra 5y ago
A potentially nice idea with several things to think about to real-world proof it. Let's think about nodejs packages because of colors and faker. Some random toughts.
I found colors on my laptop in a project of a customer of mine. I downloaded it as part of a npm install or probably with yarn. It is a dependency of another dependency. My customer should pay for the license, but how?
Among the other things, that "how" means also having a way to know how much they're going to pay if random-developer-1 adds a new dependency to package.json. This leads to asking a manager before adding a dependency, some paperwork, etc. All that we were doing when we had to buy software coming on floppy disks and CD-ROMs (but also downloads from the internet.)
However I must be able to use colors for free if all I'm doing is getting it for a non commercial project of mine. The author and the collective should reasonably feel sure that I'm not cheating. How?
One of the possible outcomes is companies dropping entire FOSS ecosystems and going back to buy whatever web development system Microsoft will hack together, backend and frontend and Windows Server + SQL Server too. Or from Oracle.
Maybe Microsoft already thought about selling a monthly license to use everything that gets uploaded on GitHub or distributed via npm. You stop paying, you rewrite your webapp with other components.
So yes, nice idea, but I'm not sure that the outcome will be good. Maybe we should keep thinking that if a customer or employer pays us to write some of our code as open source, than we'll write a lot of it. If we have to write FOSS instead of going out, maybe we'll write not much of it and we'll make money out of something else.
- phh 5y ago> However I must be able to use colors for free if all I'm doing is getting it for a non commercial project of mine. Nope. IMO it's safe (from the floss side) to assume companies won't be cheating. Don't get me wrong. They will most definitely cheat. Maybe 95% of them will cheat. That should still make tangible revenues for FLOSS maintainers. (though IME most companies really do care about the licenses they use, and they don't mind paying, like I know actual companies who bought Qt licenses) In my understanding, the goal from the article is not really to force companies to pay for FOSS, but rather to simplify it. You could say that putting everything under AGPL forces them to pay. That's technically true, but really, it's more of a way for people who want to pay FOSS inside the company to push to actually pay FOSS. Donations from a company to a someone isn't really a thing an engineer can push down to the finance department. Don't get me wrong, with the same wording, there is a fine line between "companies are stealing from us, they need to give money back" and "we know companies would like to help us, but it's hard for them, so let's make it easier for them" which is very easy to cross. For the moment, FOSS financing has no plausible solution (nor incentives), assuming companies are lawful neutral. Even for lawful good companies it's pretty hard. So let's start by making systems to give those companies way to help FOSS, before thinking about potential cheaters.
- jopsen 5y ago> IMO it's safe (from the floss side) to assume companies won't be cheating. Agreed, plus you now have a collective motivated to enforce license compliance. And "license compliance" is easy, because the collective has a list price. No need for a complicated law suit, it's always cheaper to pay the collective.