[HN Gopher] MoonBit compiler is available on GitHub
___________________________________________________________________
MoonBit compiler is available on GitHub
Author : frou_dh
Score : 26 points
Date : 2024-12-18 13:30 UTC (3 days ago)
(HTM) web link (www.moonbitlang.com)
(TXT) w3m dump (www.moonbitlang.com)
| zoezoezoezoe wrote:
| I'm a little confused what this is, look interesting, I like the
| rust-adjacent language (I mean I dont like that fn main cannot
| have parenthesis, but other functions must have parenthesis, but
| pretty interesting), I'm just a little confused as to what makes
| it different. I see they seem to have their own compiler, so no
| LLVM overhead, but also no LLVM legacy, and they seem to be
| targeting small WASM compiles, is that just it? I'm just a little
| confused on what makes it special, I think it's stack based,
| there doesnt seem to be a concept of pointers as far as I can see
| in the documentation, it looks cool, but what makes it different?
| tinco wrote:
| This looks like a fun language. I don't like the idea of wasm,
| but I do like the idea of moving away from JavaScript at some
| point. With big software like Figma using wasm to get proper
| performance in the browser the benefits of such a language are
| obvious in my opinion.
|
| Wasm is a move back to the flash days, though most developers
| nowadays probably weren't there to know why that sucked. Even 15
| years later there still isn't a decent replacement for its main
| use case, it got sacrificed for the cause of the free and open
| web. Maybe the market for it disappeared with a certain early web
| culture. Most flash iconic content was probably created with
| pirated copies of Flash anyway.
|
| Since I'm in the old person yells at cloud rant mood, I'll just
| continue on and rant at OSI as well. This project is yet another
| example of a project that would have been and should have been
| proper approved open source if OSI had not failed to get with the
| times and protect the open source community against the large
| cloud providers, the very same cloud providers that are their
| largest sponsors and donors.
| baq wrote:
| I don't understand the wasm is flash argument. Flash was a
| closed runtime full of security issues. wasm is basically
| JavaScript stripped down to the absolute minimum (see asm.js)
| and then stripped down again - the browser controls all aspects
| of execution in the same way it controls js.
| tinco wrote:
| That it was a closed runtime full of security issues wasn't
| the reason it was rejected by the open web community, or even
| rejected by Steve Jobs from the Apple ecosystem. Closed
| runtimes and security issues have always been warmly embraced
| by the web.
|
| It's just that supplying binaries to the web is less open
| than supplying html and js. It's not an especially strong
| argument, less so given how minified and uglified JS is
| nowadays, but that's what it is.
| jacobp100 wrote:
| A bit of background on this project: it was started by the guy
| leading BuckleScript (now ReScript), the OCaml to JS compiler.
| trollbridge wrote:
| I noticed they chose the SSPL licence. I'm working on a similar
| sort of project and am evaluating AGPL3 vs SSPL. (If my project
| is successful, it has a high risk of cloud providers both large
| and small co-opting it, and I'd like to proactively head that
| off.)
| childintime wrote:
| Discussion: https://news.ycombinator.com/item?id=37174619
| ptx wrote:
| The license seems a bit iffy. The blog says it's "open source",
| but the license is a modified version of the SSPL, which is not
| recognized as an open source license in the customary sense of
| the term.
|
| They also don't say explicitly what modifications they made
| compared to the SSPL. Some diffing shows that the changes consist
| of:
|
| 1. Search/replace of license and publisher name.
|
| 2. Renumbering sections with 1-based indexing.
|
| 3. Adding the words "commercially or competitively" to section 13
| (which is renumbered to 14).
|
| In the process of doing this, they seem to have accidentally
| replaced the publisher name with the license name in one place
| ("any version ever published by the MoonBit Public Source
| License") and introduced an off-by-one error where section 18
| (which used to be 17) refers to sections 15 and 16.
| egonschiele wrote:
| Not to be confused with Moonscript, which is a cool language that
| compiles to Lua: https://moonscript.org
| dartos wrote:
| A friend of mine worked on https://github.com/libriscv/godot-
| sandbox
|
| Which uses libriscv as the sandboxed compute target.
|
| IIRC They tried for years to get a wasm backend to work and
| eventually abandoned it citing, among other reasons, something
| about how it was difficult to wrangle a stack machine with a
| register machine.
|
| I don't really see what benefits wasm runtimes have over
| something like libriscv which is very fast and maps more closely
| to hardware.
| apitman wrote:
| First I've heard of libriscv. What exactly is it? A wasm
| alternative that uses RISCV byte code?
___________________________________________________________________
(page generated 2024-12-21 18:01 UTC)