So I think the permissive license like Apache 2 without opening up for upstream contributions would be the way of future open source development. In a very similar way of SQLite. It definitely will help shield off a bunch of AI generated contributions first and foremost and then, aligning mental model with the human contributions with Mojo leader team.
you can but through a python interpreter in the mojo process so you get the same numpy speed with mojo<->ptyhon interop overhead.
NuMojo is native and also is now starting to support features that numpy can't really do like native GPU execution.
I wonder if the Python-to-Mojo overhead is significant enough where language-native bindings make more sense. I'd be interested in seeing some benchmarks.
This is really exciting. I've been using Mojo off and on for side projects over the last two years.
(copying from some previous Mojo threads) It's got an ownership system adjacent to Rust, comptime similar to Zig, and a first class dependent type system. Even more exciting, is that uses LLVM (to the best of my understanding) in some novel ways and for more optimizations.
Hm, in docs they say that dependent type system is not first class seems.
Is it even dependent? More like refinement types, which seems will be in Rust too.Also HN recently mentioned Rust building similar llvm based integration with GPU.
But compared with Python, Julia, Matlab, R, Rust, C, C++ - Mojo feel relief for working with numerics.
The innovative potential of Mojo would be constrained to relatively niche and senior dev circles until they get a Windows release, which probably won't happen anytime soon. Python keeps the throne for now.
In the category of AI/HPC software Mojo aims to displace (which is my industry), I don't know many people who care that much about running Windows. The software I work on may only run as a Docker image or directly in a Linux/WSL. Mojo is hardly the only tool in the space (currently) lacking Windows support.
Technically soruce available now with the promise of accepting contributions (thus becoming fully open source) early next year. But since it is an Apache 2 license (like the rest of the LLVM project) you are already allowed to fork and contribute to your own fork right now, just no upstreaming.
For many the closed source nature of the compiler was a knock-out criterion. We will see if Mojo can gain traction now or if it has missed its window of opportunity.
Maybe it is not listed by the OSI, but it is 'open source' in practice, if the source is also in the public domain. The FSF's definition of free software makes more sense not only ideologically, but also because it does not try to be exhaustive.
You're just making up your own definitions of things.
If the complete source is available and distributable under an OSI-approved license, it is absolutely Open Source. Accepting outside contributions is nowhere in the definition of Open Source.
Technically, the criteria for open source is freedom to freely distribute and modify the source code, not to upstream changes. That's a different for thing.
As long as it was proprietary, I could ignore it. Now, I guess I have to learn a little something about it.
So I think the permissive license like Apache 2 without opening up for upstream contributions would be the way of future open source development. In a very similar way of SQLite. It definitely will help shield off a bunch of AI generated contributions first and foremost and then, aligning mental model with the human contributions with Mojo leader team.
Very exciting. Closed source was one of the main things making me not try it out yet.
Any good resources for getting started? And anything on doing NumPy/SciPy-like stuff in it?
https://mojolang.org is the best spot for getting started generally. numojo is the closest thing to numpy right now. You can check out more packages here: https://github.com/modular/modular-community/tree/main/recip...
NuMojo is the closest thing in the ecosystem to numpy and the final aim is offer numpy-like experience. https://github.com/Mojo-Numerics-and-Algorithms-group/NuMojo
Can't you just use Numpy directly in Mojo?
you can but through a python interpreter in the mojo process so you get the same numpy speed with mojo<->ptyhon interop overhead. NuMojo is native and also is now starting to support features that numpy can't really do like native GPU execution.
I wonder if the Python-to-Mojo overhead is significant enough where language-native bindings make more sense. I'd be interested in seeing some benchmarks.
This is really exciting. I've been using Mojo off and on for side projects over the last two years.
(copying from some previous Mojo threads) It's got an ownership system adjacent to Rust, comptime similar to Zig, and a first class dependent type system. Even more exciting, is that uses LLVM (to the best of my understanding) in some novel ways and for more optimizations.
Hm, in docs they say that dependent type system is not first class seems.
Is it even dependent? More like refinement types, which seems will be in Rust too.Also HN recently mentioned Rust building similar llvm based integration with GPU.
But compared with Python, Julia, Matlab, R, Rust, C, C++ - Mojo feel relief for working with numerics.
Such an exciting time for Mojo, I don't think people realize just yet the innovation being brought to fore here.
The innovative potential of Mojo would be constrained to relatively niche and senior dev circles until they get a Windows release, which probably won't happen anytime soon. Python keeps the throne for now.
In the category of AI/HPC software Mojo aims to displace (which is my industry), I don't know many people who care that much about running Windows. The software I work on may only run as a Docker image or directly in a Linux/WSL. Mojo is hardly the only tool in the space (currently) lacking Windows support.
I'd put you in the "relatively niche and senior" bucket without even thinking.
I would put you in the clueless bucket without even thinking
They announced a partnership with Microsoft today.
That's quite promising - https://www.modular.com/blog/modcon-announcements#:~:text=Wi...
And Julia as well.
Related post with relevant pr: https://news.ycombinator.com/item?id=49347777
Related:
Mojo 1.0
https://news.ycombinator.com/item?id=49261128
"Modular Confidential (obviously), May 14, 2022"
lol, feels like espionage
https://github.com/modular/modular/blob/main/KGEN/docs/Desig...
Compiler walkthrough
https://github.com/modular/modular/blob/main/KGEN/docs/MojoC...
Technically soruce available now with the promise of accepting contributions (thus becoming fully open source) early next year. But since it is an Apache 2 license (like the rest of the LLVM project) you are already allowed to fork and contribute to your own fork right now, just no upstreaming.
For many the closed source nature of the compiler was a knock-out criterion. We will see if Mojo can gain traction now or if it has missed its window of opportunity.
It’s open source under the Apache 2 license, not source available.
Accepting contributions is not required to be open source.
SQLite doesn’t accept contributions from random people either [0] (though you could argue being public domain isn’t being open source).
Anyway, this is great, I was waiting for this to take the language for a spin! Congrats to the Mojo team!
[0]: https://simonwillison.net/2025/Dec/29/copyright-release/
Maybe it is not listed by the OSI, but it is 'open source' in practice, if the source is also in the public domain. The FSF's definition of free software makes more sense not only ideologically, but also because it does not try to be exhaustive.
You're just making up your own definitions of things.
If the complete source is available and distributable under an OSI-approved license, it is absolutely Open Source. Accepting outside contributions is nowhere in the definition of Open Source.
Technically, the criteria for open source is freedom to freely distribute and modify the source code, not to upstream changes. That's a different for thing.
https://opensource.org/osd