You are right, i might mix up here some things, as i don’t have experience in licensing other than MIT or GPL.
Still, my recommendation recently was: Take care, that IF someone makes money with your SW. I’ve seen too many cases when SaaS companies take over and forget to share with the originator.
And (at least) i see no problem in having julia (or dominant packages) as free of charge for non-profit, research etc, but with a reasonable price tag beyond that.
My judgement might be biased, as in my day job i’m in engineering/research and SW prices are just another cost position. If SW costs 50.000€ but solves a 1.000.000€ problem there is an easy decision.
Julia is a programming language - I don’t see how charging for it would work realistically, as there is no big corporate entity backing it. The only examples I can think of where this sort-of has been tried is the Metal API on Mac (where you pay as part of acquiring a device) and maybe Java to some degree? It’s not a very widespread model in the world of programming languages, unlike enduser software.
I think the main problem OP has is, that he has difficulties to pay for something which can be used for free. As chairman of a non-profit association I sometimes have similar problems: I can’t just pay for something which is free available as this would be a misuse of the associations money. But sometimes I want to give something back so I have to be a bit creative to find a way which can be easily justified to the auditors, like paying money for some generously counted support hours.
If some one wants to give something back and this something is money, this is a very nice and good idea. Not everybody is able to give something back with code.
Let me add, as co-creator of Julia but also a co-founder of Julia Computing: it’s totally fine to use Julia for free. That’s the idea! Here are some ways to contribute:
The biggest is to be engaged, report bugs and help fix them if you can
If you are in a position to donate to the project, please do so through NumFOCUS, our non-profit, which is an umbrella non-profit for a large number of technical computing projects (but money donated for Julia will go to Julia, not other projects)
Open source is so weird and wonderful. After spending the day interacting with people who build and fix things just because it’s useful/cool, it can be quite a shock to the system to talk to friends and family and here the usual “no such thing as a free lunch”, “nobody does anything for free” etc.
I wander what the economists have to say about foss. Clearly there is a whole world of things where the financial incentive is not the principal one. I guess there is theory and research on that?
well, top Linux kernel contributors are often paid by companies: Linux turns 25, with corporate contributors now key to its future • The Register , similarly a lot of Julia contribution are done by people who work in companies/labs where Julia is at its core (Julia Computing, PumasAI, Invenia, Beacon Biosignals etc.), (some are doing CS/PL research in the first place)
So I think in some sense FOSS is just part of the chain, but since often the work is not explicitly paid for, FOSS contributors deserve all the applause they receive.
The current license already takes a stance on this. The MIT license explicitly states the licensed product can be used for pretty much anything, as long as copyright notices are preserved. This entails commercial and non-profit use.
In our department, there is professor Nico van Yperen who has done a lot of research about motivation. He said that the strongest motivator is that you “like” something. So, a nurse becomes a nurse because that person likes to do nursing work and a fireman likes to put out fires. I’ve been thinking about this for months about open source and haven’t found a counter argument. Most counter examples are just rationalizations when you think about it. Yes, people do open source work to make the world a better place and all that stuff and do like it when someone compliments the work. Those altruistic motives probably help, but in the end it probably just comes down to “liking it”. Liking it when finishing code which solves a problem, when spending an hour writing some complex logic and the computer flies through it in 0.01 seconds and when GitHub shows a green CI badge.