The moonbit compiler is available on GitHub | Hacker News Reader

The moonbit compiler is available on GitHub

moonbitlang.com

23 pointsbvaldivielso13 comments

> MoonBit is still in beta-preview. Introducing forks at this point could risk destabilizing the project. We aim to reach a more mature and stable status before welcoming community contributions.

> We want to safeguard against large cloud vendors leveraging our work for commercial purposes in a way that could undermine our efforts.

Normally I don't like the "technically not open source" comments, but this isn't even remotely open source.

The language looks like rust with some interesting type/generic fun added in. Not sure I like the syntax but it looks like a very expressive language.

Other comments are on the mark, but still, very cool to see such a young language leaning so hard into wasm components / component model. Other languages have a lot of work to do to support the poly-language polytopia we hope to be heading for; starting with wasm components as an early objective can hopefully set yourself up for a long beautiful future of being really good at the new standard virtual machine.

> MoonBit has supported the Wasm component model and reduced the code size to 27kB in the http-hello-world example—much smaller than other languages, making MoonBit-wasm ready for production usage.

We have had C ABI etc for a long time achieving pretty good polyglot for a while. Sorry mistyped, meant pretty good polytopia.

Obligatory: the SSPL is widely not considered to be an open-source license, and this variation probably isn't either. I think it's best to avoid the term (or start with a base license like AGPL instead, if you really care.)

Anyway, I don't know much about this language (first time hearing about it) but congratulations nonetheless. Admittedly, I am struggling to figure out what exactly where Moonbit fits in versus other programming languages, and I'm curious what is special about it, but it's not immediately obvious to me reading the front page blurbs.

Oh man I might get sued by their non-lawyers over something that’s never been enforced in the history of law.

Google recently had a press release where they decided to make the silly and largely pointless claim,

> [Willow] lends credence to the notion that quantum computation occurs in many parallel universes

There is nothing illegal about this, but this silly attempt at making the press release seem even more important than it was (which was unnecessary anyways) wound up detracting a bunch of attention for little benefit. (Ultimately, it didn't really matter much: most of the focus was still on the accomplishment, but it's still a pointless own-goal.)

This is basically the same thing. Of course, it isn't illegal to use a term wrong, but it's going to inevitably detract from the things that matter. There's plenty of other words in the English language that can be used in place of the words "open source". Doesn't matter what you choose: source-available, fair source, etc. It's all fine. Just don't use "open source", that's confusing because it isn't consistent with how the term open source is used.

For better or worse, Open source is firmly-established industry jargon by this point. There's little point in arguing about it because right or wrong, we can go 100 comments deep and the status quo will be absolutely identical by the end. It's not worth it.

That explanation meanders too much.

Then don't participate in long-form discussions.

> "Why are we open-sourcing the Wasm backend?"

I remember this debate when the SSPL was debuted, and I think for the most part people have switched to using the term "source available" for licenses like this, but I don't think it occurred to me before now: what's the "verb" form of "source available"? The best I can come up with is "making the source available".

I mostly find this all pretty silly because "available" seems to basically be a synonym of a certain meaning of the word "open". It reminds me a lot of the whole "free as in freedom" versus "free as in beer", except for the difference between the two types of "open" being way more ambiguous. As far as I can tell, the question mostly boils down to disagreements about how "open" something needs to be in order to qualify; in other words, it's about the degree of openness and where to draw the line. With things like this, I always tend to prefer drawing the line somewhere as easily identified as possible, so drawing the line at "is the source literally possible to obtain or not?" feels like a much more obvious question to answer than "is the list of things I'm allowed to do with the source sufficient for me to do the things that I think I should be allowed to do?".

I recognize that this is already a lost battle, but it really feels like the consensus opinion had a lot more to do with posturing than what people actually would expect words to mean, which is disappointing to me as a descriptivist.

To me it's neither here nor there. Personally, I wouldn't touch SSPL software, not because there's anything particularly wrong with non-open source licenses, but because I think the SSPL is just a terrible copyright license anyways.

The OSI and people who supported the OSI, for better or worse, put a lot of effort into branding for open source that I think we can all agree was extremely successful. There's no reason that this sort of thing can't be done for "fair source" or "source available" or whatever you want to call it. That said, it's an issue for proponents of this license model to sort out in my opinion.

Yeah, I guess I just don't feel like putting effort into branding means that it gives them a permanent right to define the term in a way that's not intuitive. If they want to make a term that they can define however they want, the mechanism for that is a trademark, and the mechanism for enforcing that is legal, not social. Outside of that, I'm pretty much always going to be on the side of language being used in a simple straightforward way rather than for the benefit of a singular entity.

Well, they don't have the right to define anything, but this isn't a matter of law. "Open source" is industry jargon, it is well-understood in the industry, and very few experts are going to disagree, especially not by many of the people and projects that popularized the term "open source" to a wider audience, like Linux.

They could've chosen a phrase less likely to be mistaken, that was not already in use, and this might've saved us a lot of arguing. But then again, the term "free software" could've won out instead, and things would be even worse. Personally, I think it'll be OK.

> what's the "verb" form of "source available"?

"We are publishing the source code for our Wasm backend." It's not hard to work out.

Reply on news.ycombinator.com