Do you think Ruby could change something so fundamental as dynamic => static typing and still retain its beauty?
The only static typing solution I've seen for Ruby is Sorbet, and it's... not beautiful.
I love ruby but I do look forward to having a good option for types at some point. The community makes up for lack of types by adding more tests to their test suite, which is nice but I'd love to layer some encoding into the codebase itself and save the CI time.
(Random example): https://docs.ruby-lang.org/en/3.4/String.html#method-i-lines
I'm not sure why they don't link that in the documenation but I beleive it is something the RDoc mainer wants to do.
def lines: (?string? separator, ?chomp: boolish) -> Array[String]
| (?string? separator, ?chomp: boolish) { (String line) -> void } -> self
Later, we adopted static analysis tools and code linters. This helped, but only so much could be done with the dynamically typed languages which were popular in the early internet.
As a self-taught programmer with no college degree, I started with dynamic languages: Perl, JavaScript, PHP, Ruby. I viewed the statically typed languages (and the programmers who could wield them) with reverence and awe. C, C++, Java, these were “real” languages that required big brains, and I should stay in my dynamically typed playpen. But as I matured, my understanding shifted: dynamic languages require hubris, require belief that you can keep an entire complicated codebase in your head with perfect recall. Statically typed languages, it turns out, invite humility.
With a static type system, each step in a program’s lifecycle can be a little contract with itself. Break that contract and the language will tell you exactly how and where. These languages can provide immediate feedback to the developer about a program’s correctness before it is even run. Bridging the gap between all of the existing dynamic code (and the developers who wrote it) and the statically typed utopia lies so-called gradual type systems such as TypeScript.
I like to use a meteorological analogy here: if JavaScript is “wintry mix,” then TypeScript lowers the temperature just enough to build a snowman.
I've been doing Typescript now for about 9 months, so I wouldn't call myself an expert in that, but I also have decades of experience in many other dynamic and static languages (including Haskell), so I can say with confidence that a lot of features in Typescript are to solve problems unique to the Javascript world, and then there are the features in Typescript that are there to solve problems in the other Typescript solutions, and all-in-all while it has great utility and has many fantastic features it just isn't possible to completely overcome the fact that it's a static type system, on top of a dynamic type system.
I've seen a number of posts to the effect of "I'm coming from Typescript and learning Go, how do I do X" and so often the answer has been "Even in Go, even with its very simple type system by static language standards, the answer is that you don't do X at all because X is a feature on top of a feature on top of a feature designed to deal with being on a dynamic language and when you're on a static language you not only don't deploy that solution, you wouldn't even have the problem that's the solution to except that you just created it for yourself by copying Typescript too closely in some other way." A simple example being you don't ask how to type a map/dict/etc. in a normal static language based on the contents of the "type" value, you just declare a properly-typed value in the first place and always use that, at most converting at the edge.
Typescript is a great tool but a very unrepresentative view on how static typing works.
In order to catch problems in dynamic type languages you end up needing a bunch of additional tests to, ironically, verify the expected type. And even then, those tests don't and can't tell you how a method is actually used throughout the program.
Consider the following class
class Human
def initialize(name)
@name = name
end
end
class Human
def initialize(first, middle, last)
@fullName = "#{first} #{middle} #{last}"
@first = first
@middle = middle
@last = last
end
end
Types make refactoring this sort of code somewhat trivial. You simply change the fields and let the compiler tell you ever position which relied on that field.
This extends into all sort of circumstances, like letting the user of a method know what types are expected. For example, the above ruby code doesn't really guarantee that `first` is a string. In fact, it could be an int or even an object and the code would likely handle it just fine. Types expose interface contracts in a quick and easy way for programmers to understand what they are dealing with.
In a team, it is just so much easier to come across a "typed" function as opposed to an untyped one. You need to read the entire function to know what it is about.
I personally think that simpler Rust is a better example of the benefits static typing (or maybe something like Gleam).
These languages are really quite similar in many ways, but their domains ended up diverging. With python becoming the layman / scientific / learning language of choice, ruby has been pigeon holed into mostly web development.
Both are really easy to pick up and learn for somebody unfamiliar with CS concepts and I personally find the ruby syntax far more intuitive.
We have a lot more options now. For a while people tried to use python and ruby as glue / systems programming languages, but with golang and rust you have really good and more performant options in that space. And as you say the tooling has improved massively, so the hurdle of moving on to a more "rigid" language is less than it ever was.
I still really like ruby, and I think rails is still a powerhouse due to solving so many real world problems in a really complete package, but the lack of adoption outside of that niche has left it dwindling in popularity.
VSCode was "good enough" for pretty much every language with LSP at that point, I did't even bother with Jetbrains ides outside of work after that.
And when Obsidian replaced org-mode for me, I deleted my .emacs directory from my dotfiles repository.
It was the opposite for me. Emacs + LSP + many other common conveniences all bind together so beautifully and efficiently that I can't imagine using any other IDE at this point.
Hopefully emacs can catch up.
This is undervalued. So frustrating in ruby that this doesnt exist or at least isnt easy.
I use kotlin nowadays…
I didn't know how to think about the types so I wanted some way to annotate them to help think through it, but went through it. And then the compiler complained at me I was passing in the wrong type to a function. I mean yes thanks? But also give me a way to figure that out BEFORE I try running the code.
If only the Ruby ecosystem had adopted Scala instead of Ruby, with cutesy books and eccentric underscored personalities, history might have been different.
Ruby was a great little niche language until Rails showed up.
Scala did not exist when Rails was written/extracted. When Scala was released, it did not include any of the things that made Ruby a good choice for Rails.
During my first Introduction to Programming course at university, I was taught Java. One thing that I found very troubling is that it wasn't easy, or possible in many cases, to change the programming language. Sure, you can write new functions or methods or classes, but I can't change the keyword for an if-statement. I also remember the TA saying "why would you want that?" I caught myself thinking "if we can program a computer, then why can't we program a language?"
15 years later, I still have this issue a bit, except I made my peace with it. It is what it is. There are some exceptions though! Such as: Lisp, Smalltalk and similar languages. It's in part why I have worked for a company that professionally programmed in Pharo (a Smalltalk descendant [2]). I remember hacking a very crude way for runtime type checking in Pharo [1], just for fun.
I'm not a Ruby programmer, all I know is that Ruby has some things that are identical to Smalltalk. But my question to the author would be: if you long for things like keyword arguments, type hints and namespaces why don't you program it in the Ruby language yourself?
Or is that really hard, like most other languages?
[1] https://youtu.be/FeFrt-kdvms?si=vlFPIkGuVceztVuW&t=2678
[2] Fun fact, I learned about Lisp, Smalltalk and Pharo through HN! So I know most of you know but I suspect some don't.
I think that's why extremely flexible languages have seen limited adoption - if your language is more of a language construction kit where everyone can implement their own functionality, everyone has to implement their own tool support (or, more likely, live without any) and there's a limit to how far you can go with that. The best languages find the sweet spot where they give you enough flexibility to implement most reasonable programs, but are still constrained enough that tools can understand and work with all possible code.
#define BEGIN {
#define END }
Flashbacks to scala operator PTSD.
No. I don't want to use ++<>^^%% operator! I am not a number! I'm a man!
https://www.baeldung.com/java-reflection
https://www.baeldung.com/java-annotation-processing-builder
Then you've got the byte code itself, and there be dragons: <https://aphyr.com/posts/341-hexing-the-technical-interview>
While you rarely see byte code shenanigans in Java code bases, it's how some other languages on the JVM achieve things like runtime metaprogramming.
It can be done, but it is not economical, and therefore not practical.
Is a typical response of someone without the background and without the imagination. It may well be, that doing Java-only for too long robs one of both. An alternative response could have been: "What a fascinating idea! How would you apply that? / What would you do with that?"
I am happy for you, that you found the exceptions and that your picture of the computer programming world is not as incomplete and bleak as that of the TA back then.
Why do people make hating a tool their entire personality? I have noticed this same thing with languages like Go ("oh no Go still bad") and C++. I don't like C++ myself but I don't hate it. It would be like hating a screwdriver.
If you don't like a language simply don't use it, there are hundreds of alternatives. If your employer is making you write in that language you should hate your employer, not the language.
> The reasons behind this choice of employment are very much unrelated to the technology stack.
Programming language isn't the only factor for employment. People don't always get to just change jobs when an aspect isn't ideal for them.
On top of that, python is ubiquitous in some sectors. It's not as easy to avoid as a lot of other languages.
It runs very deep for some people. "That's not Pythonic!" or "That's all unreadable line-noise!"
It becomes a kind of language bigotry similar to English-speakers hating a foreign language. And yet, as you rightly say, these are just programming tools.
I suppose that when humans invest in any type of language, they form and protect the orthodoxy.
I know this is hard to understand for logical genius savants like you. A lot of HNers are like that and don’t understand human emotions.
I think this is tricky. I work in the world of data and, although I like python fine, I'd find it very hard to find a role in my field that doesn't involve, if not working in python, then at least integrating very closely with work data-scientists produce that's in python.
Some languages just have a big dominance in their field. Python has that for data, and javascript for front-end.
This is just one such example and it's similar to how group dynamics with sport teams and politics work.
Likewise, other ones exist that make jobs super easy - be they having a ratchet, or quick change of bits, etc
Programming languages seem to be a pretty good parallel. Though, I don't see why Python in particular would be hated. It has its bullshit, but it's workable.
Modern python is a much better language across the board, even if the packaging/deployment story still needs some love
I hate Philips screwdrivers (but love JIS).
Both Python and Ruby are screwdrivers.
In my highly opinionated opinion, the parallels are:
- Flat head: Perl. Functional, and revolutionary at the time! But awkward to work with.
- Phillips head: Python. Awesome improvement, much easier to work with. But still annoying in frustrating ways.
- JIS: Ruby. A further refinement that resolves the last annoying 5% of work. (Well, OK, resolves 4 of the last 5%) ... (Also Japanese, coincidentally?)
/s
Is it 2007 again already?
Initially, the celebrated feature of python was that it allowed easy and fast development for newcomers. There was a joke a long the lines, "I learned python, it was a great weekend".
As much as I like python's type system (and wouldn't want to see them ever go way!), part of me wonders if moving into a world where hello-world can look like this, is a world where python is no longer the "lean in a weekend" language:
from typing import Annotated
import typer
app = typer.Typer()
@app.command()
def main(
name: Annotated[str, typer.Option("--name", "-n")],
) -> None:
"""Prints out 'HELLO {name}!!' (name upper cased) to the console"""
print(f"HELLO {name:upper}!!")
if __name__ == "__name__":
app()
For the last 10 years, Python's evolution has been directed by professional developers working for large software companies, and they've focused on adding features that help them work on million-line Python codebases.
Even if the language was originally intended to be easy to learn and ideal for small programs, that's clearly not a design goal any more.
Is there a language today that’s as easy to understand as the “executable pseudocode” of Python 2.x? I haven’t found one.
I'm very familiar with pyright and still I start most of my new projects without types and start sprinkling them in once I have a good base working already. This works so well that every time I pick up a static language I just get turned off by the friction of mandatory types and go back to Python. The only exception is Typescript where I can just Any everything temporarily as well.
Nobody's stopping you from manually parsing a couple of arguments. I still do it all the time and it's OK. If anything the magic of gradual typing is that you get to use it as necessity arises.
That said, I think it's true that production python tends to look more like my example, and less `print("hello world")`.
I know that the Ruby community loves its clever runtime metaprogramming, but even the most metaprogrammed codebase is still going to consist mostly of plain old in-out methods. And as anyone who's ever typed a dynamic codebase knows, you pick up so much low-hanging fruit, in terms of edge cases and errors, when you slap some types on those. You don't need to type everything, but there is real hostility in Ruby circles to gradual typing, even where it would make sense and wouldn't impose any major costs.
Personally, I've stopped writing Ruby. Short of any pathway to sensible gradual typing, I just can't shake the feeling that every new line of Ruby is instant tech debt. Which is such a shame, since I find real beauty in the language.
I'm not sure what you're basing that on, gradual typing has been built into Ruby since 3.0 (2020). Sorbet can around a bit earlier in 2019.
There is on going work to improve both with RBS-Inline support now in Sorbet with runtime support hopefully on the way.
IRB uses RBS for autocompletion and Solargraph and Ruby-LSP both support it.
I would say there is no hostility to gradual typing in the Ruby community. Quite the opposite, people are being paid to work on it.
I ended up rewriting my Ruby codebase in Elixir (with thanks to some kind pointers here on HN). Elixir has perfectly satisfactory gradual typing via Erlang's dialyzer, which will happily power an LSP right now, and work is underway on a handmade gradual set-theoretic type system. It's a direction of travel I feel more confident in than Ruby's.
The ergonomics of ruby still have me very much liking the language as it fits how I think, but there are a lot of good developments in the usual neighbors, and I see myself picking up both Python and JS ever more for small projects.
You're not the only person to reply with something like this but just to repeat, Ruby has had official gradual static typing support for 5 years now.
Gradual typing is fundermentally already part of Ruby.
Ruby could do with better static analysis tooling but people are being paid to work on that.
I'm really charmed by ML style languages nowadays. I think python has built a lot of kludges to compensate for the fact that functions, assignments, loops, and conditionals are not expressions. You get comprehensions, lambdas, conditional expressions, the walrus operator... most statements have an expression equivalent now.
it seems like, initially, Guido was of the opinion that in most cases you should just write the statement and not try "to cram everything in-line," so to speak. However it can't be denied that there are cases where the in-line version just looks nice. On the other hand now you have a statement and an expression that is slightly different syntactically but equivalent semantically, and you have to learn both. Rust avoids this nicely by just making everything an expression, but you do get some semicolon-related awkwardness as a result.
I used to be like that. When I did Java. I used to think to myself: "Oh neat! Everything has its place. interfaces, abstract classes, classes, methods, anonymous classes, ... everything fits neatly together."
That was before I learned more Python and realized: "Hey wait a moment, things that require me to write elaborate classes in Java are just a little bit of syntax in Python. For example decorators!" And slowly switched to Python.
Now it seems many Java-ers have come to Python, but without changing their mindset. Collectively they make it harder to enjoy using Python, because at workspaces they will mandate the most extreme views towards type annotations, turning Python into a Java dialect in some regards. But without the speed of Java. I have had feedback for a take-home assignment from an application process, where someone in all seriousness complained about me not using type annotations for what amounted to a single page of code(, and for using explanatory comments, when I was not given any guarantees of being able to talk with someone about the code - lol, the audacity).
Part of the problem is how people learn programming. Many people learn it at university, by using Java, and now think everything must work like Java. I mean, Java is honest about types, but it can also be annoying. Has gotten better though. But that message has not arrived yet at what I call the "Java-er mindset" when it comes to writing type annotations. In general languages or their type checkers have become quite good at inferring types.
You definitely need to have a decent grasp of architecture to make this work - strict FP is very helpful to prevent any early spaghettification - but you ultimately get the best of both worlds this way: rapid iteration for the early stages and type safety once you develop a feel for the system you're building.
I've been doing this in Elixir in the last few months and I've really been enjoying it.
Sure, you can always rewrite to match that style with macros in lisp and generators in python, but they weren't meant to be used that way.
Sad thing about ruby is how they failed to do typing. I love python's typing module. I think it is the single best thing about python and I wouldn't touch python with a pole if it didn't have that module.
Only thing I've come near that gave me as much joy was Elixir, and I simply didn't have time to pick it up more than the most generic basics.
my mind just likes a.any? {|x| x.someCondition? }
But then I really don't know Ruby, so happy to be told why this is wrong...
Steen’s not wrong that Python evolved and Ruby moved slower, but he’s wrong to call Ruby stagnant or irrelevant. Just think what we've enjoyed in recent times: YJIT and MJIT massively improved runtime performance, ractors, the various type system efforts (RBS/Sorbet etc) that give gradual typing without cluttering the language etc.
Ruby’s priorities (ergonomics, DSLs, stability) are different[2] from Python’s (standardisation, academia/data). It’s a matter of taste and domain, not superiority.
[1] I'm stealing a point DHH made on Lex's podcast. https://www.youtube.com/watch?v=vagyIcmIGOQ
[2] I'm once again parroting DHH/Matz
And I've actually used Sorbet with Ruby for 2 years, but that seems like a really bad solution to this problem.
Doubly weird, considering that TypeScript work was inspired by typed/racket, and TypeScript doesn't have a sound type system afaik and the OP's first love was Scheme.
Error handling for instance has always been my pet peeve. With dynamic languages like python, errors were all "exceptions". But then golang came along and decided they'd be values and that only the truly exceptional errors should be "panicked". But the if err != nil syntax became super verbose only after I learned about the Result<Ok, Err> from Rust and the matching syntax associated with it.
A lot of people are shocked when they learn about ruby's monkey patching. I for one never truly groked the packaging of python applications until uv came along to deliver an experience similar to npm.
And I agree, Typescript is the state of the art as far as static typing on top of a dynamic language is concerned. But I never considered it a programming language. More like a tool to assist developers write/manage large javascript code.
In the end, I think the true reason for returning to pythong probably had more to do with getting a python gig. I live in the part of the world where my tech stack ended up being influenced early on by the places I worked. I didn't mind learning Typescript for my first gig or improving my skills with nodejs.
In the end, every language can really get things done. And Typescript helps me pay my bills and I couldn't be more grateful. Learn to love the quirks of your language and stop comparing it unfavorably with others. To date, I've never seen a language as elegant as ruby. Nor do I find an ecosystem better than python's at data science.
I work in Python and PHP every day though.
Keyword arguments and destructing assignments are there.
This brings me to C#. At .NET 10 there will be another option then python. .NET 10 brings the ability to run C# script files without the need for a proj file or a main method. This will bring the full .NET Framework and NUGET eco system with it. I can't wait to replace all my python scripts with this.
In the last 5-7 years I've had to go in other directions. Clojure, Python, Java, even back to C and taking a look at Rust and Zig. I'm now in a role where I don't code so much, but I can see Ruby's problems - performance, the surprises, the fact it allows idiots to do idiotic things (nobody under the age of 40 should be legally allowed to monkey patch a base class or engage in meta programming).
And yet when I want to do something for me, for fun, perhaps advent of code, or a quick mock-up of something that's rolling around in my head, I reach for Ruby. Not elixir which has better runtimes, or C or Zig or Rust which has better performance, not something statically typed which leads to fewer bugs, not Python which has a huge data science community to it...
A few weeks ago I was listening to the DHH episode of the Lex Fridman podcast [0], where DHH talks about Ruby as a "luxury programming language". This matches my own experience.
When I need something to be fast, it's either because I'm dealing with a low-latency problem (and some of my side projects are very latency sensitive - 5ms can make the difference between success and failure), or because I can't afford the luxury of Ruby and Rails and the developer ergonomics.
Ruby makes things fun for the programmer. That's the point. It's beautiful to work with, even if it doesn't do all the things that all the coding books and blogs insist I should be ashamed to not have in my language.
I was slightly embarrassed to be a Ruby and RoR advocate for a while because of the brogrammer BS that emerged around both ecosystems in the 2010s. I then became very embarrassed because it wasn't as cool as a systems language like Rust or Go, or as intellectually deep as Haskell, or as hot on the ML bandwagon as Python.
But I think I don't care anymore. I'm just going to accept it for what it is, and lean into. Life's too short for "shoulds" - I'm just going to like what I like. And I like Ruby.
Languages like C#, Java, C++, Scala, Kotlin, and Python in 2025 feel like industrial computerized bandsaws with twenty different settings and controls. Some more complicated than others. They can be tuned to crank out a factories needs but you could spend days just fussing with a single setting that isn't right.
That being said, modern Python to me feels like the least thought out of these. It has been incrementally changed from one language to another, while being forced to keep many of the worst parts of both. To be honest, I think they should keep up the "breaking backwards compatibility" trend and make Python 4 an optionally-compiled, statically-typed language more like Go, but with more expressivity than Go.
I suppose F# is already like my ideal Python 4. It's possible to run as a script or compiled binary. It's a nice clean syntax, and the type system is a joy to use.
A valid F# program can be a single line script or dozens of configuration files. This let's the developer use it for quick and dirty work, then progressively tweak settings to run in a more industrial scale setting.
Anyway, that said, for new personal projects I like typescript and rust. But recently I needed to stick an admin interface on such a project and rails shines there, you can get something good and secure stood up with less code and faff than anything else. In today's world of LLMs that is helpful too, rails is concise and old and has lots of open source projects to pull from, so AI breezes through it.
class Mess
def chaos(x)
if x > 0
[1,2,3].each do |i|
case i
when 1
if i.odd?
puts "odd"
else
puts "even"
end
when 2
begin
puts "trying"
rescue
puts "failed"
end
else
puts "other"
end
end
else
puts "negative"
end
end
end
class Mess:
def chaos(self, x):
if x > 0:
for i in [1, 2, 3]:
match i:
case 1:
if i % 2 == 1:
print("odd")
else:
print("even")
case 2:
try:
print("trying")
except:
print("failed")
case _:
print("other")
else:
print("negative") class Mess
def chaos(x)
if x > 0
[1, 2, 3].each do |i|
case i
when 1
if i.odd?
puts "odd"
else
puts "even"
end
when 2
begin
puts "trying"
rescue
puts "failed"
end
else
puts "other"
end
end
else
puts "negative"
end
end
end
class Mess
def chaos(x)
if x > 0
[1, 2, 3].each do |i|
case i
when 1
puts i.odd? ? "odd" : "even"
when 2
puts "trying"
else
puts "other"
end
end
else
puts "negative"
end
end
end
class Mess
def chaos(x)
return puts "negative" unless x > 0
[1, 2, 3].each { |i| handle_item(i) }
end
private
def handle_item(i)
case i
when 1 then puts(i.odd? ? "odd" : "even")
when 2 then puts "trying"
else puts "other"
end
end
endThis is a question of style; it's possible to write ruby the way you have but with experience you definitely wouldn't.
e.g.
Why wrap the iterator in an if..else..end block? You could replace that with a one line guard statement at the top of the method. No end required.
Why use begin..rescue..end in the middle of your example? Just replace that with a single rescue at the end of the method. Again, no end statement required.
This is like nesting 20 if..else blocks in python and then complaining you don't have a monitor big enough to view it without scrolling. You just wouldn't do it that way.
My hope is that with the new ruby parser rubocop will be more agressive about automatically refactoring examples like this away.
To each their own. It’s because we all have different preferences that there are so many choices.
class Mess
OUTPUTS = {
1 => ->(i) { i.odd? ? "odd" : "even" },
2 => ->(_) { "trying" },
3 => ->(_) { "other" }
}.freeze
def chaos(x)
return puts "negative" unless x.positive?
[1, 2, 3].each do |i|
puts OUTPUTS[i]&.call(i) || "other"
end
endOf course you can get all this down to a single line with ; demarcation.
And your `.each` could use `{ ... }` syntax, just like C or Java or... you know, everything else.
But sure, whitespace is better, or whatever it is you prefer.