Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Mojo – a new programming language for AI developers

modular.com

Post reply on HN
Print view View original

121–130 of 272 posts

Re: Mojo – a new programming language for AI developers

#121

Post by neilv »

minhazm wrote: post #91
lairv wrote: post #71

Earlier quoted context omitted.

From their FAQ https://docs.modular.com/mojo/faq.html "Will Mojo be open-sourced? Yes, we expect that Mojo will be open-sourced. However, Mojo is still young, so we will continue to incubate it within Modular until more of its internal architecture is fleshed out. We don’t have an established plan yet. Why not develop Mojo in the open from the beginning? Mojo is a big project and has several architectural differences…

The problem with this is that you don't know what the licensing will be when/if they open source it. So you could start using it now and they might change their mind and choose not to open source it or it might be a restrictive license.

Also, even if fully open source, have to also look at governance.

(A canonical project can be steered in a way that makes it impractical for you, and forking is often also impractical.)

For now, I'd treat it as closed source, which is a non-starter for investing in, when I can accomplish the same in open source ways. And there's no sense in giving away the open source uptake benefits to a company when the software isn't open source.

Re: Mojo – a new programming language for AI developers

#122

Post by pama »

chrislattner wrote: post #20

There are a bunch of questions about Julia, so I'll do my best to give a short answer to a very long and complicated topic. Up front, Julia is a wonderful language and a wonderful community, I am a super fan. That said, Mojo is a completely different thing. It is aligned with the Python community to solve specific problems outlined here: https://docs.modular.com/mojo/why-mojo.html Mojo also has a bunch of technical a…

Congrats on the launch! In addition to the dramatic possibilities for ML, I hope mojo will have an impact in other scientific software through the ease of specialization to hardware capabilities. The current common subset of Python and “pure” mojo (ie not using CPython) is small, but will expand according to your roadmap. Do you envision porting pure Python libraries as part of the effort, or perhaps help other communities do so? For example einsum in numpy or pytorch are great, but how much more work would be needed to have a mojo-accelerated einsum interfacing with numpy or mojo-specialized arrays? Or have a pure mojo numpy altogether?

Re: Mojo – a new programming language for AI developers

#123

Post by mattnewton »

mirekrusin wrote: post #28
chrislattner wrote: post #20

There are a bunch of questions about Julia, so I'll do my best to give a short answer to a very long and complicated topic. Up front, Julia is a wonderful language and a wonderful community, I am a super fan. That said, Mojo is a completely different thing. It is aligned with the Python community to solve specific problems outlined here: https://docs.modular.com/mojo/why-mojo.html Mojo also has a bunch of technical a…

Hold on, are you behind it? ...looks like it! This "detail" shouldn't be hidden away, it means a lot.

As a former apple eng I guess I forgot that some people don’t know who Chris Lattner is haha. There isn’t anything unscrupulous going on here he probably just thought posting under his username was enough of disclaimer.

Re: Mojo – a new programming language for AI developers

#124

Post by mirekrusin »

EGreg wrote: post #89
mirekrusin wrote: post #28

Earlier quoted context omitted.

Hold on, are you behind it? ...looks like it! This "detail" shouldn't be hidden away, it means a lot.

How does it look like it?

From "Team"'s link [0]

[0] https://www.modular.com/team

Re: Mojo – a new programming language for AI developers

#125

Post by bobbylarrybobby »

ummonk wrote: post #93
int_19h wrote: post #76

Earlier quoted context omitted.

Why is "no GC" an advantage for something aimed at such high-level tasks?

Yeah funnily enough I think "no GC" would be a much better feature in Julia, which would make it a great language for real time applications.

You could do a lot worse than modern GC. For instance, in Rust, lack of GC may cause you to reallocate memory or use reference counting, both of which are likely slower than GC. For instance, building a tree in Rust will do a lot of reference counting, which is a lot of heap accesses.

Re: Mojo – a new programming language for AI developers

#126

Post by rg111 »

I would have started using it today for my personal projects and write tutorials if it were FOSS.

It looks sooo good. Once it turns into FOSS (if it does), I will start using it as quickly as I can.

Then, if and when it gets more mature, and delivers on its promises, I might start using it at my work.

Re: Mojo – a new programming language for AI developers

#127

Post by asadm »

theLiminator wrote: post #74
lairv wrote: post #71

Earlier quoted context omitted.

From their FAQ https://docs.modular.com/mojo/faq.html "Will Mojo be open-sourced? Yes, we expect that Mojo will be open-sourced. However, Mojo is still young, so we will continue to incubate it within Modular until more of its internal architecture is fleshed out. We don’t have an established plan yet. Why not develop Mojo in the open from the beginning? Mojo is a big project and has several architectural differences…

Their response to why it's faster to move while closed source seems kinda like a false dichotomy. They can choose to start with it open disallow input from the community or close their issue tracker to the community.

It's closed so others don't catch up quicker or add those optimizations upstream? Then they have no moat.

Re: Mojo – a new programming language for AI developers

#129

Post by singularity2001 »

Wonderful language. Only complaints (so far) :

DynamicVector should be named Array and SIMD should be named Vector and dispatched to whatever SIMD/vector pipeline the host offers, similar to the Flexible Vectors proposal in WASM: https://github.com/WebAssembly/flexible-vectors/blob/main/pr...

And please make Optional[T] as ergonomic as Swift / Kotlin with a simple '?'

Re: Mojo – a new programming language for AI developers

#130

Post by dom96 »

Impressive. I think this has a real chance of being successful. Backwards compatibility with Python and what appears to be a strong focus on improving what makes Python frustrating to work with (let/var, no GIL, compiled) all looking promising. I'll definitely be keeping a close eye on this and look forward to getting access to the playground.

Reading the docs it seems there are still some things that aren't settled and I would like to offer my feedback. From https://docs.modular.com/mojo/programming-manual.html:

> Alternative: instead of using the `&` sigil, we could call this an `inout` argument.

I think you should do this. What makes Python great is its readability so don't compromise on this.

Post reply on HN