Stanza: a new optionally-typed general-purpose programming language | Hacker News Reader

Stanza: a new optionally-typed general-purpose programming language

lbstanza.org

44 pointstrishume20 comments

We moved most of the comments from this thread to the creator's Show HN, which was posted a few days ago and didn't get much attention: https://news.ycombinator.com/item?id=11717561.

We'll roll back the clock on that post so it gets to the same place on the front page that this one was.

Post from the original author of the language-

https://news.ycombinator.com/item?id=11717561

I like what I'm seeing so far. The mix of ease of prototyping (e.g. not fighting the type system if you don't want to) and powerful constructs from functional programming land (map and friends, currying) and powerful function overloading is nice. And the syntax is concise yet readable.

One comment on the website: The only font-family specified for code is Menlo so I see an ugly variable width font on my PC :-(.

Can we port numpy/sklearn to this?

How many more new languages do we need? This is getting ridiculous. What does this thing do that Python doesn't do?

How many pointless comments do we need? This is getting ridiculous. What does your comment add to the conversation that a million other whiners don't do?

It points out that we don't need more and more computer languages that no one uses. If you can come up with a language that really makes new contributions or really works better than the existing ones, great, but instead it seems we just have a bunch of me-too languages doing almost the same thing. Languages aren't useful unless there's a critical mass of people developing them, making libraries for them, making good compilers for them, and providing mindshare for them, so that there's enough people competent in them to use them. This is why languages like C and C++ are still used even though there are many valid criticisms of them and supposedly superior alternatives have been developed: the alternatives just haven't been good enough, in many/most cases, to justify switching. C++ may have its warts, for instance, but with all the libraries that exist for it and all the existing code and all the workers familiar with it, it usually makes sense to stick with it rather than switch to something like D or Rust or Swift or Go unless you have a really good reason (like you're developing on iOS where Swift is really well-supported).

These days, Python is the de-facto standard for general-purpose type-optional programming, and has a huge amount of libraries and mindshare, and is growing. So why do we need a clone?

> Python is the de-facto standard for general-purpose type-optional programming

Type hints were added to Python only in the most recent version (somewhat controversially), and the interpreter only knows about the syntax; there are no semantics defined. (You have to run an external program if you want to check the types). Why would you say Python is the standard for optional typing?

The alternatives to C/C++ are so popular that most programming is not done in C/C++ anymore, so your premise seems off. The alternatives have done stunningly well.

Rarely, RARELY is someone going to make a language that's radically better than what's currently available. But every new language is a chance to step in the right direction, or confirm that nope, no one wants something like this. We need MORE languages. Hundreds, a thousand more. Mostly they won't be used, but sometimes things will move forward.

We need as many languages as we want to make, otherwise we would have stopped at C.

Please don't post unsubstantive dismissals to Hacker News, especially in response to new work. If you have a substantive criticism, fine—please make it respectfully. If you don't, then don't post in these threads.

On HN, no one need ever apologize for creating something new, regardless of whether someone else thinks "this thing" is "needed". Most new creations won't end up becoming a huge deal, but that's no reason to trample on them. Indiscriminate trampling just ensures we end up with (a) an ugly, damaged garden, and (b) a culture of the status quo. It's also a vulgar behavior that violates the civility requirements of HN.

No one, least of all the trampling internet commentariat, knows how to distinguish what will prove important from what will not prove important. Therefore we should receive new things with respect for the creative spirit, even if we don't personally intend to use them.

Reply on news.ycombinator.com