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

What I mean when I say that machine learning in Elixir is production-ready

cigrainger.com

Post reply on HN
Print view View original

101–108 of 108 posts

Re: What I mean when I say that machine learning in Elixir is production-ready

#101

Post by neonsunset »

ambicapter wrote: post #3

I see so much breathless adoration for Erlang and Elixir, how come it hasn't taken over the world in all this time?

They are slow, have clunky syntax and soundness issues due to lack of strong type system. Now, swap Elixir and F#, and HN crowd, usually signing praises to the former, could stand to gain a lot by using productive language with performant runtime and much better standard library that does not force you into writing code in the exact specific way (especially when it does not fit the use case).

Re: What I mean when I say that machine learning in Elixir is production-ready

#102

Post by vvpan »

OutOfHere wrote: post #89
vvpan wrote: post #57

Wanted to add to the usual "why is BEAM not more popular" conversation that there is also Gleam [1], which is an up and coming typed language on top of BEAM. It has the more common C-like syntax and a growing ecosystem of libraries, for example Lustre [2]. Unlike Elixir it has a much tighter syntax without all the macro magic. It also does not wrap around Erlang processes and you have to use those directly. To myself…

Gleam is a tooling development language, not a concurrency language. It runs away from Erlang's concurrency ecosystem. Also, its documentation, especially wrt concurrency, is sorely lacking. In its current state it would be a mistake it to use it for any distributed work, but perhaps this can change in time.

Ah, I did not realize that. What is the point then? Why have a language on top of Erlang that "runs away from Erlang's concurrency ecosystem"?

Re: What I mean when I say that machine learning in Elixir is production-ready

#103

Post by OutOfHere »

vvpan wrote: post #102
OutOfHere wrote: post #89

Earlier quoted context omitted.

Gleam is a tooling development language, not a concurrency language. It runs away from Erlang's concurrency ecosystem. Also, its documentation, especially wrt concurrency, is sorely lacking. In its current state it would be a mistake it to use it for any distributed work, but perhaps this can change in time.

Ah, I did not realize that. What is the point then? Why have a language on top of Erlang that "runs away from Erlang's concurrency ecosystem"?

Gleam has some third-party package(s) for concurrency, but I wouldn't consider them mature, and they wouldn't compare to anything like what Elixir has. It is not the primary or even the secondary focus of the Gleam language. I guess there is no point for me.

Re: What I mean when I say that machine learning in Elixir is production-ready

#104

Post by dns_snek »

asa400 wrote: post #81
dns_snek wrote: post #63

Earlier quoted context omitted.

Great write-up! I've been using Elixir for a handful of years now, primarily working on solo/consulting projects and in micro teams. I'd like to think I have a pretty good grasp on the language and OTP fundamentals, but I'll admit that some the things you mentioned in your last paragraph about lacking knowledge of operational concepts resonated with me. Over the years I've tried to learn more about releases, deployme…

> My impression is that it's difficult to learn about these topics without real hands-on experience (something that seems hard to replicate in small/solo projects). Maybe. Unfortunately, I don't think there's a cheat code there. Often the single best thing you can do is to put yourself near people who know what you want to know. > Would you mind expanding on that last point? I'd be curious to learn more about these k…

Thank you for your insights!

Re: What I mean when I say that machine learning in Elixir is production-ready

#105

Post by namaria »

ambicapter wrote: post #3

I see so much breathless adoration for Erlang and Elixir, how come it hasn't taken over the world in all this time?

I see so much concern and worry about climate change, how come it's not solved after all this time?

Re: What I mean when I say that machine learning in Elixir is production-ready

#106

Post by kamma4434 »

ajmurmann wrote: post #46
ToJans wrote: post #2

I'm a big fan of Erlang, Elixir, OTP and the BEAM vm. However, I only have very lightweight server needs for my SaaS, so it just doesn't make sense to switch to it, especially as my code is mostly client-based/typescript. If my core value prop would be around real-time messaging or streaming data, BEAM/OTP would be my first choice. Slightly OT: I'm still in dubio about Cloudflare pages, but I'm sure that the platform…

I am really intrigued by the distributed systems capabilities that BEAM brings. I built a chat app using it and am also using Mnesia DB to store recent messages. However, my experience coding it has been terrible. I am using IntelliJ with an Elixir plugin and there is essence is no auto-complete. I used to write a ton of Ruby, but there RubyMine was really good at guessing what's available. Not so much here. On top o…

Vs Code with the default Elixir plugin is pretty solid. Not wonderful but you get all you need, including the debugger.

Re: What I mean when I say that machine learning in Elixir is production-ready

#107

Post by kamma4434 »

OutOfHere wrote: post #89
vvpan wrote: post #57

Wanted to add to the usual "why is BEAM not more popular" conversation that there is also Gleam [1], which is an up and coming typed language on top of BEAM. It has the more common C-like syntax and a growing ecosystem of libraries, for example Lustre [2]. Unlike Elixir it has a much tighter syntax without all the macro magic. It also does not wrap around Erlang processes and you have to use those directly. To myself…

Gleam is a tooling development language, not a concurrency language. It runs away from Erlang's concurrency ecosystem. Also, its documentation, especially wrt concurrency, is sorely lacking. In its current state it would be a mistake it to use it for any distributed work, but perhaps this can change in time.

From what I understand, you get all of OTP as it is now – You only use the experimental Gleam packages if you want types. So you should be able to use them right out of the box, with the same level of guarantees (or lack of) that you get in Ex/Erlang. Still, I have not yet found a compelling reason to switch over – or at least try – Gleam.

Re: What I mean when I say that machine learning in Elixir is production-ready

#108

Post by kamma4434 »

ambicapter wrote: post #3

I see so much breathless adoration for Erlang and Elixir, how come it hasn't taken over the world in all this time?

Because it’s an excellent fit for some problems, but not so much for others. The ecosystem is definitely smaller and the flower quality compared to say, JavaScript, go, or Java. The performance of the virtual machine is definitely way off the JVM For general computation.
Post reply on HN