hckrnws

For this audience it may be worth noting that Ruby’s blocks are closures and are passed to methods either anonymously/implicitly or as a named parameter, may be subsequently passed around to any collaborator object, or otherwise deferred/ignored, have the same range of argument arity as methods and lambdas, can even be formed from (and treated similarly to) lambdas, and are thereby fundamental to Ruby’s claim to being a multiparadigm language even as they also betray the Smalltalk roots.

In addition they have nonlocal return semantics, somewhat like a simple continuation, making them ideal for inline iteration and folding, which is how most new Rubyists first encounter them, but also occasionally a source of surprise and confusion, most notably if one mistakenly conflates return with result. Ruby does separately have callcc for more precise control over stack unwinding, although it’s a little known feature.

These were some of the best long running sentences I’ve read in a while. A true rubyist!

[flagged]

The hard part is doing without semicolons.

> can even be formed from (and treated similarly to) lambdas

They are also used to create lambdas (even the shorthand stabby-lambda syntax desugars to a call to Kernel#lambda with a block.)

> Ruby does separately have callcc for more precise control over stack unwinding, although it’s a little known feature.

callcc is included in CRuby but has been sidelined from Ruby as a language separate from CRuby as an implementation for a while, with Fibers understood to cover the most important use cases for callcc.

As someone who comes from strict languages (the more strict, the better) Ruby blocks are... not fun.

I've seen them used in situations where they are used like a callback, but due to the nature of how you write them, you have no clue whether the variable you're referring to is a local or a global one.

This makes debugging incredibly hard.

Ruby isn’t strict, no, but that’s by the by, because this doesn’t sound like a problem with blocks or how you write them. It sounds more like a problem with evals i.e. some library or framework misusing them. Blocks are closures and they straightforwardly bind variables and resolve constants/instance variables from the context of their instantiation, and resolve methods similarly because self within the block is from instantiation, when we call them normally with yield or Proc#call. Same goes for implicit contexts used for definition, if your block does a bit of metaprogramming.

If someone plays silly buggers and invokes them under instance_eval or class_exec etc that fiddle with self or definition contexts then some of this goes out the window, but those are special-purpose methods that come papered with red flags. This is typically seen in poorly designed DSLs that are trying too hard to pretend they’re not actually Ruby. If memory serves, the Chef DSL was a prime example in this regard. If the language was stricter, then sure, this wouldn’t be possible. But debugging these cases isn’t super hard either once you know the limited range of culprits, and the fix is always the same: place values in local stack variables to rely on them in a closure.

Using blocks for callbacks is fine. Don’t make assumptions about the semantics of flow control statements that other languages may have imposed on you, i.e. use next and break for explicit block results and local exit instead of return, and don’t eval them.

This is really cute and heartwarming.

Back in the day, a lot of people including me reported feeling more comfortable in Ruby after one week than all their other languages with years of experience, as if Ruby just fits your mind like a glove naturally.

I'm glad new people are still having that "Ruby moment"

Ruby is still the first tool I reach for if I need to do something quickly and naturally. People are always surprised at how readable the code comes out even compared to python.

The three stages of ruby:

1. Someone scoffs at you for writing ruby instead of python / go / javascript / kotlin

2. Then they read the code

3. Then they install ruby

It's an amazing language and remains my go-to for anything that doesn't need protocol stuff.

Thank you for reading this. I'm having fun learning Ruby. I just started working at a company where I use it full time. It's great learning it and I have supportive colleagues who are excited for me. I'm going to write more about Ruby. I have planned about 6 articles in the next few weeks. I hope I get around to them all.

Coming from a language with functions as first class objects, blocks felt a bit limited to me, because it feels as if you almost have functions but not really, and they get inputted by a back door. Used for example to:

let isLarge = a => a>100;

numbers.filter(isLarge)

Blocks let you do the same but without extracting the body as cleanly. Maybe it’s a chronological issue, where Ruby was born at a time when the above wasn’t commonplace?

>When you write 5.times { puts “Hello” }, you don’t think “I’m calling the times method and passing it a block.” You think “I’m doing something 5 times.”

I’m of two minds about this.

On the one hand, I do agree that aesthetically Ruby looks very clean and pleasing. On the other, I always feel like the mental model I have about a language is usually “dirtied” to improve syntax.

The value 5 having a method, and that method being an iterator for its value, is kinda weird in any design sense and doesn’t seem to fix any architectural order you might expect, it’s just there because the “hack” results in pretty text when used.

These magical tricks are everywhere in the language with missing_method and the like, and I guess there’s a divide between programmers’ minds when some go “oh that’s nice” and don’t care how the magic is done, and others are naturally irked by the “clever twists”.

> The value 5 having a method, and that method being an iterator for its value, is kinda weird in any design sense and doesn’t seem to fix any architectural order you might expect, it’s just there because the “hack” results in pretty text when used.

I don't think this is particularly weird, in Ruby at least. The language follows object orientation to its natural conclusion, which is that everything is an object, always. There is no such thing as "just data" in Ruby, because everything is an object. Even things that would just be an `int` in most other languages are actually objects, and so they have methods. The `times` method exists on the Integer classes because doing something exactly an integer number of times happens a lot in practice.

I don’t have an issue with the “everything’s an object” part, because it _is_ consistent, even though it gets a bit trippy when classes are objects as well and they are implementation of a Class class which is an implementation of itself (trickery again!).

The issue is more with this part:

>The `times` method exists on the Integer classes because doing something exactly an integer number of times happens a lot in practice.

It is practical, but it breaks the conceptual model in that it is a hard sell that “times” is a property over the “5” object.

The result is cleaner syntax, I know, but there is something in these choices that still feels continually “hacky/irky” to me.

Perhaps I've been doing Ruby for too long, but it's still not that weird to me. The quantity "5" is very abstract without anything to have "5" of. That is why "5.days" and "5.times" exist, among others. Mathematically it makes just as much sense to start with the amount and add the unit later than it does to start with the unit and add the amount later (ie like `time_to_wait = SECONDS_IN_A_DAY * 5` as you might do in some other languages).

Maybe it is clearer if I explain it in syntactic terms? In my mental model objects are nouns (described entities) and methods are verbs - actions over the noun.

process.start() is the action of starting done by the the noun that is the process.

It's not exactly a matter of naming, as some methods are not explicitly verbs, but there is almost always an implicit action there: word.to_string() clearly has the convert/transform implication, even if ommitted for brevity.

I see no path where 5 is a noun and times the verb, nor any verb I can put there that makes it make sense. If you try to stick a verb (iterate?) it becomes clear that 5 is not the noun, the thing performing the iteration, but a complement - X iterates (5 times). Perhaps the block itself having a times object with 5 as an input would make it more standard to me (?).

But I do understand that if something is extremely practical a purist/conceptual argument doesn't go very far.

It’s not just about practicality. Ruby is using message passing, not method calling. This is fundamentally different and a bit foreign to the larger community. Then ruby layers syntactic sugar on top that hides this.

Behind the scenes everything is a message passed using __send__ and you can do this directly as well, but you generally don’t.

So when you write

5.times { puts "Hello" }

It’s sort of expected by the average programmer that you are telling 5 to call the times method and expect it to exist and do what it’s told.

In reality you have indirectly sent a message that looks like

5.__send__(:times) { puts "Hello" }

What we are really doing is sending a message to 5 (the receiver) and giving it the opportunity to decide how to respond. This is where method_missing comes in to allow responding in a custom fashion regardless if a method was explicitly defined.

So you’re not telling 5 to call the method times, rather you are asking, “Hey 5, do you know how to handle the message times?”

These are fundamentally different things. This is actually super important and honestly hard to really grok _especially_ in ruby because of the syntactic sugar. I came from a C/C++ background originally, then Java and then moved to Ruby. After a few years I thought I understood this difference, but honestly it wasn’t until I spent a couple years using Objective-C where message passing is happening much more explicitly that I was able to truly understand the difference in a way that it became intuitive.

I’m a rubyist, but how is message passing fundamentally different from method calling? I get that method_missing adds a twist, but your comment doesn’t explain what the fundamental difference is.

Especially in the context of Fixnum#times. How does message passing vs method calling matter there? #times is just a method on Fixnum.

Because it leaves it up to the object being called what to actually do with the message. The object you're talking to might be forwarding its messages to another object in another ruby instance on another machine (if the current machine is getting fully loaded, etc), and the caller would be none the wiser. And crucially, the caller wouldn't have to be modified to enable this. The logic for this would be entirely within the object being called.

So the difference isn't just with method_missing.

With "method calling" as you put it, the program blows up if that object doesn't have that method, WHEN YOU CALL IT.

Basically, this smalltalk oo paradigm is about shifting where you put more of your logic. At & around the "call" site, or within the object whom you're calling & entrusting to do something useful with it.

All hearkening back to Alan Kay's original ideas about biology influencing how we organise code, and having a program be 1000's of "little black boxes" where work gets done by all these boxes talking to each other.

Which is why smalltalk (& ruby implements the Smalltalk object model to its entirety) actually has an awful lot in common with Erlang & other BEAM runtimes, even though those are functional languages. Because once you get beyond the techy buzzwords, the main ethos behind them is actually quite similar.

I guess what I’m getting at, is that I don’t understand how the difference actually informs anything concretely, as in the example of Fixnum#times, where this discussion started. Why is it super important to understand this fundamental difference?

Fixnum#times isn’t a great example, I only used it since the parent used it to illustrate their confusion and quite frankly a concrete useful example is to complex for this format.

ActiveRecord has changed a lot over the years, but as an example in the original ActiveRecord you used dynamic finders. None of the finder methods existed initially, but if you passed a message to an active record object for a non existent method rather than fail it would determine if that should be a method and then it would build and persist a method to the process for future calls.

It allows for some really interesting and powerful applications in horizontally scaling as well.

> I’m a rubyist, but how is message passing fundamentally different from method calling?

The difference is fairly subtle most of the time in practice, which is why dynamic OO languages like Ruby that use message passing can mostly look like OO languages that use method calling, though it is significant behind the scenes and opens up a lot of possibilities; one of the more obvious practical differences is that it is why Ruby can capture and handle unknown "method calls" via "method_missing", which works the way it does because the version of __send__ defined in BasicObject looks for an existing method matching its first argument in the object's inheritance chain, and if it finds that calls it, and if it doesn't find one does the same thing with the "method_missing" method.

There's a lot that can be built on top of this, either leveraging method_missing or more directly.

Though I think the more relevant issue upthread isn't message passing vs method calling but narrow vs broad conception of nouns/agents.

In a pervasively object oriented system/language, anything that can be represented in the system is an object (=noun =potential agent of action).

Ruby is almost a pervasive OO system (falling short because blocks are not objects, though they can be reified into objects with some efficiency overhead.) As such, unreified blocks are the only thing in Ruby that cannot be the agent of action.

For an iteration of a block a fixed number of times, the potential agents are the block and the number of times. But blocks aren't objects, so they can't be the agent, leaving only the number.

Hence, Integer#times

I have been doing Ruby for so long that it feels very natural to apply a method in this way on the instance.

false.not

applies the not method on the false instance in the same way that

car.start

in every OO language calls the start method on car as the receiver.

So filter(list) feels just wrong when you are clearly filtering the list itself.

Although I prefer Elixir currently I agree that ruby at least goes all the way in on OO and not having to remember which feature is implemented as a language syntax and what is just a method invocation is a strength not a weakness. It is different in other languages for historical performance reasons really.

list.filter is ok! Filtering is an action that applies to a list

false.not is borderline but if read as false.negate it makes sense (negating is an action that applies to a Boolean value). That wording screws the chaining though.

5.times is where the pattern breaks: times is not an action that applies to a number (nor an action at all). It’s the block the one that should repeat/iterate - but Ruby breaks the rule there and blocks are not an object (!). If they were you could block.repeat(5) which IMO is cleaner.

There is a bit of personal preference in what "applies to a number", but I see what you mean.

As a slight correction, a block is indeed an object! They are received by methods as an instance of the Proc class:

  def inspects_block(&block)
    puts block
    puts block.class
  end
  inspects_block { "foo" }
  # => #<Proc:0x0000000000000000>
  # => Proc

  class Proc
    def repeat(n)
      n.times { self.call }
    end
  end
  ->{ puts("foo") }.repeat(3)
  # => foo
  # => foo
  # => foo

I think I feel you. However, I think you have conceptually loaded "method" with more meaning. I think a less loaded way to think of the object/method is that the object is the first argument to the method that is called on it. So

    5.times { puts 'hi' }

    times(5) { puts 'hi' }

    my_function = -> { puts 'hi' } 
    repeat_times(5, &my_function)

After my foray into functional programming, I actually ended up appreciating Ruby more, because it lets you have it both ways: program your computer directly, and harness functional concepts. Since computer hardware is not functional I don't want the extra ceremony and abstraction over it for the sake of purity.

All that said, going back and forth between Ruby and Elixir really conceptually crystallized for me that the method call receiver is basically just the first argument to the method, accessible with the keyword `self` (which in Python is made explicit for example).

Hmm, but that's not really a "breakage" in ruby, it's more an aesthetic argument over which objects should have which logic. It's like naming things. Smalltalk (& ruby) has 'select'. Everyone else uses 'filter'.

To some, 5.times seems very readable & logical. It's like arguing over the "right" colour scheme to use while coding (BTW, the correct answer is solarised light, but with black foreground text!!)

Blocks are actually instances of the Proc class. There are helper methods to handle blocks passed to methods in a lightweight manner but you can also accept the block as a method argument, e.g.,

  class Integer
    def times(&blk)
      i = 0
      while i < self
        blk.call(i)
        i += 1
      end
    end
  end

> Blocks are actually instances of the Proc class.

No, they aren't.

Blocks can be reified into instances of the Proc class, but they are not objects and reifying them into objects has overhead.

Using a & argument asks for a block passed to the method to be reified into a proc accessible in the body of the method, which is useful if you are going to do something with it that you can't do with a bare block.

Fun fact - there's no Boolean class in Ruby. True is an instance of TrueClass and false is an instance of FalseClass

> that “times” is a property over the “5” object

Maybe here is the confusion, ruby is based on message passing, so the `times` is a message you are sending to 5, not a property of it.

I think you’re right, but I also suspect that doesn’t clear up anything for most people as in my experience they generally don’t grok the difference unless they’ve already spent a significant amount of time in something like smalltalk or Objective-C

When you're new it just looks like a weird-but-reasonable syntax.

When you learn the language you really fall into two camps:

- ah, yes, this is unusual, but it's consistent and now i understand the language

- this is way too clever

I'm more in the first camp.

In Smalltalk there is no if statement, but there is ifTrue and ifFalse methods in a Boolean variable. This is one example how to create custom control flow with non-local returns and blocks. Ruby follows suite.

Times isn’t a property of the 5 object, it is a method of the 5 object.

With similar justification, you could argue that blocks should have a method `times` taking an integer, because repeating a block an integer number of times happens a lot in practice. I’d even argue that it is conceptually closer to blocks than to integers, and therefore the method belongs on the block. Hence you’d write `{ puts “Hello” }.times 5`.

But blocks are not objects in Ruby, so you can’t do that, and everything actually isn’t an object in Ruby.

Also, it’s even more common to do something depending on whether a condition is true or false, but true and false in Ruby don’t have a method to (not) execute a block on them, and you use a non-OOP `if` instead, so what’s up with that?

> everything is an object, always.

I beg to differ. What object does the method `puts` belong to? Why do you not call puts with its objects name?

Ruby has a concept of mixins (Golang interfaces), these are not objects. Neither is `puts`

>Ruby has a concept of mixins (Golang interfaces), these are not objects.

Ruby "mixins" are the affordance of sharing instance methods from a module via a keyword. Modules are objects, and are instances of class Module:

  module ProvidesFooMixin
    def foo = "foo"
  end

  class Bar
    include ProvidesFooMixin
  end

  Bar.new.foo
  # => "foo"

  puts(ProvidesFooMixin.class)
  # => Module

  ProvidesFooMixin.object_id
  # => (some integer value)

Like all methods, `puts` is an object:

  method(:puts)
  # => #<Method: Object(Kernel)#puts(*)>

  method(:puts).class
  # => Method

  method(:puts).object_id
  # => (some integer value)

>What object does the method `puts` belong to?

It belongs to the object you are calling it from within. You don't need to call `puts` with a receiver because it is an instance method, just like you don't need to call an instance method `foo` via `self.foo`. But you could choose to use a receiver, since the `puts` you know and love is just another instance method. You can try `self.puts` for yourself in some context!

Your classes (and their instances) inherit this `self.puts` instance method from the Object class, which includes the Kernel module, which provides `Kernel#puts`. So the only reason you can send it as a message without a receiver is because it is just another instance method (again, the same as calling instance method `#foo` without using `self.foo`).

Caveat: You can build an "alternate universe" object hierarchy by inheriting from BasicObject, and in your alternate universe, you can choose to not `include Kernel`, and you will see that instances of your new objects do not have access to `puts` in their instance scope.

`puts` is just a method of the Kernel module: https://ruby-doc.org/3.4.1/Kernel.html#method-i-puts, just like `p` and many others. Kernel is included in the Object class that is the root of the class hierarchy, so its methods are available in every ruby object.

Mixins are just modules, which are objects, which you can call methods on. (Or rather, send messages to) You can easily verify this in irb calling a method on (for example) the Enumerable module:

    irb(main):001> Enumerable.class
    => Module

Well, “puts” is a method defined in the Kernel module, which is included in the Object class, which means it is available in all contexts (since every object is a subclass of the ‘Object’ class). So, to answer your question, the puts method belongs to the Kernel module object, and the Kernel module object is an instance of a Module object.

Answerable by a few minutes of googling. Sort of:

puts is a method which has a class: the Method class:

  irb(main):001:0> method(:puts)
  => #<Method: main.puts>
  irb(main):002:0> method(:puts).class
  => Method

  irb(main):003:0> 3.puts()
  Traceback (most recent call last):
          2: from /usr/bin/irb:11:in `<main>'
          1: from (irb):3
  NoMethodError (private method `puts' called for 3:Integer)

  irb(main):004:0> method(:puts).class.methods
  => [:allocate, :superclass, :<=>, :<=, :>=, :==, :===, :autoload?, :autoload, :included_modules, :include?, :name, :ancestors, :attr, :attr_reader, :attr_writer, :attr_accessor, :instance_methods, :public_instance_methods, :protected_instance_methods, :private_instance_methods, :constants, :const_get, :const_set, :const_defined?, :class_variables, :remove_class_variable, :class_variable_get, :class_variable_set, :class_variable_defined?, :public_constant, :freeze, :inspect, :deprecate_constant, :private_constant, :const_missing, :singleton_class?, :prepend, :class_exec, :module_eval, :class_eval, :include, :<, :>, :remove_method, :undef_method, :alias_method, :protected_method_defined?, :module_exec, :method_defined?, :public_method_defined?, :to_s, :public_class_method, :public_instance_method, :define_method, :private_method_defined?, :private_class_method, :instance_method, :instance_variable_set, :instance_variable_defined?, :remove_instance_variable, :instance_of?, :kind_of?, :is_a?, :tap, :instance_variable_get, :instance_variables, :method, :public_method, :singleton_method, :define_singleton_method, :public_send, :extend, :to_enum, :enum_for, :pp, :=~, :!~, :eql?, :respond_to?, :object_id, :send, :display, :nil?, :hash, :class, :singleton_class, :clone, :dup, :itself, :yield_self, :taint, :tainted?, :untrust, :untaint, :trust, :untrusted?, :methods, :frozen?, :protected_methods, :singleton_methods, :public_methods, :private_methods, :!, :equal?, :instance_eval, :instance_exec, :!=, :__send__, :__id__]

It's not as much a mystery as you think, the Kernel module is pretty well documented and states right in the first paragraph that its methods are available in all Ruby objects: https://ruby-doc.org/core-3.0.2/Kernel.html

> Everything being a confused muddle in Ruby

It really isn’t a confused muddle, the rules are very clear. Just because it doesn’t match what you expect from your other language experience doesn’t mean it isn’t clear.

I'm not a Rubyist really, but I started using it at work a few weeks ago for a very small script. I'm not allergic to RTFM, so I picked up the canonical reference book on Ruby and occasionally visit the official docs for the language. I agree: the structure is clear and straightforward and there's nothing difficult about learning it.

It's also the opposite of magic; magic is when language features can't be described in terms of the language itself.

Clear, like the glass in Liberace's candelabra.

Modules (mixins) are objects, classes are also objects (of type Class).

The Kernel module is included in the Object class, which means its methods are available to every Ruby object and can be accessed from virtually any scope, including the top-level (global) context where self is an instance of Object called "main."

> I beg to differ. [...] Why do you not call puts with its objects name?

In the first place, I'd say what you're asking for goes beyond "everything is an object".

But I think your questions can be answered in a way that affirms that "everything is an object" in Ruby anyway.

> Why do you not call puts with its objects name?

Because it belongs to whatever object you're working in already; `puts` is identical to `self.puts`. And yes, you're always working in an object: https://bparanj.gitbooks.io/ruby-basics/content/chapter1.htm...

> What object does the method `puts` belong to?

As indicated above, it belongs to the object `self`. It gets added object via the mixing-in Kernel module into Object. Kernel is itself an instance of class Module: https://docs.ruby-lang.org/en/3.4/Module.html

The `puts` in Kernel delegates to `puts` from IO, which is likewise an instance method belonging to the object you can refer to by the name `$stdout`: https://docs.ruby-lang.org/en/3.4/IO.html

> Ruby has a concept of mixins [and], these are not objects.

Sure they are. Mixins themselves inherit from Module, and Modules are also objects (just like classes are).

Some highlights from Chapter 27 ("Library Reference: The Class Model") of the recent edition of the pickaxe book (emphasis mine):

> The Kernel module is included by class Object, so its methods are available in every Ruby object. One of the reasons for the Kernel module is to allow methods like `puts` and `gets` to be available everywhere and even to look like global commands. Kernel methods allow Ruby to still maintain an "everything is an object semantics".

and regarding mixins:

> The Module class is the class of any module you declare with the `module` keyword. Each module is an instance of class Module.

on Object:

> Object is the parent class of (almost) all classes in Ruby unless a class explicitly inherits from. BasicObject. [...] Object mixes in the Kernel module, making the built-in functions globally accessible.

tl;dr: mixins in Ruby are instances of class Module, and their methods end up bound to instances of class Object. Abstract module methods that don't belong to a concrete instance of some class that mixed in their module belong to the Object that is the module itself (the instance of class Module). (The same kind of thing is how class methods work.)

`5.times` is not so outlandish, though it would seem better for that to be in a `Loop` library or something (`Loop.times(5) { do some stuff }`).

The `5.days` example that was posted somewhere else in this thread might be a better example. It is not, as far as I can tell, part of Ruby's core library, thank goodness, but it is the kind of thing the methods-as-hacks-for-syntax culture seems to encourage. My question being "why the heck should the number 5 know anything about days? What does 5.days even mean, given that there are various ways to interpret 'a day'?"

This kind of bad design has made its way all over the place. The Java mocking libraries that my coworkers like to use are full of it. Long chains of method calls that appear like they're trying to 'look like English' but make no damn sense, so you have to dig into each method call and what kind of thing it returns to understand what this chain actually means.

you can view it as serving the same role as universal function call syntax, the latter being a feature in some other languages that rewrites `a.f(x)` to `f(a, x)` if `a` doesn't have a method `f`. in ruby you can just add the method to `a`'s class directly, but the idea is the same - you're adding some user defined function that is closely related to the type of `a`, and you're using method call syntax because that reads nicely.

Blocks are fundamentally different from functions due to the control flow: `return` inside a block will return the outer method, not the block. `break` stops the whole method that was invoked.

This adds some complexity in the language, but it means that it’s far more expressive. In Ruby you can with nothing but Array#each write idiomatic code which reads very similar to other traditional languages with loops and statements.

More specifically blocks (and "proc"'s) return from the defining scope. This is just a minor clarification, but it matters, because if you pass a block down from where it is defined, and the block calls "return" it will still not just exit from the method where it was called, but from the method where it was defined.

This can sometimes be useful: A calling method can pass down a block or proc to control if/when it wants an early return.

Basically Ruby has two types of closures:

* A return in a lambda returns to the calling scope. So basically, it returns to after where the "call" method is invoked.

* A return in a block or a proc returns from the scope in which it was defined (this is also why you get LocalJumpError if you try to return a block or a proc, but not a lambda to the method calling the one where the block or proc is defined).

When you name a block, you get a Proc object, same as you get when you take the value of a lambda or proc.

In practice, that blocks in MRI are not Proc objects already is just an implementation detail/optimisation. I have a long-standing hobby project to write a Ruby compiler, and there a "proc" and a bare block are implemented identically in the backend.

You are right on return (use next in a block), but break uses block scope.

Maybe I explained it a bit imprecise. I was trying to explain the following behavior:

    def foo
      p 1
      yield
      p 2
    end

    foo { break }

WAT? I'm a 12+ years Ruby developer and I didn't know this.

If you are familiar with a true object-oriented language like Smalltalk (rather than the watered-down form of OO in C++, Java, etc.), an integer like 5 having methods makes sense because it (like everything else) is an object. Objects in Ruby aren't just window dressing -- they are its core.

But then Ruby only goes half way, not unlike the "watered-down form" in your term. Why is `#times` a method of Integer, but `#if` (or `#ifTrue`) not a method of booleans like in Smalltalk? Ruby does the same cherry picking from Smalltalk like everybody else, just different cherries. When looking at Ruby, it feels like the simple examples are all nice and clean but then the weird details start to appear and the language feels more hacky than others (like Ned Flander's house in Simpsons S08E08).

#if and #ifTrue are yours if you want them:

  class TrueClass
    def if = true
    def ifTrue = true
  end

  class FalseClass
    def if = false
    def ifTrue = false
  end

  true.if
  # => true
  false.if
  # => false

In Smalltalk those methods don't return `true`. They take a block and evaluate it if the boolean receiving the message

    (a > b) ifTrue: [ "do something" ]

There's no `if` keyword in the language. Control flow is done purely through polymorphism.

I apologize for my lack of Smalltalk knowledge. As you can imagine, you can do similar in Ruby by defining ifTrue to accept a block, even adding ifTrue on other all objects and defining something similar:

  class TrueClass
    def ifTrue(&block) = block.call
  end

  class FalseClass
    def ifTrue(&block) = nil
  end

  class Object
    def ifTrue(&block) = block.call
  end
      
  class NilClass
    def ifTrue(&block) = nil
  end

Certainly possible: add ifTrue as a method to TrueClass and FalseClass.

It just isn't very fast.

problem is not with ifTrue, and not with it's performance, it's easy to do. it is "ifTrue:ifFalse:"

also it is common to do assignments in the "if", and with actual method and blocks scope of the introduced variable would be different and everyone would be tripping on it all the time.

basically it's because of "else" and "elsif". While ".each" works the same as "for .. in ...; end", it's harder to do "if else" as method which will also return value of the block inside the branch. Smalltalk can do it because "ifTrue:ifFalse:" is _one_ message, ruby didn't go that way syntactically.

This is your example in Ruby:

    isLarge = -> {|a| a > 100 }

    numbers.filter(&isLarge)

    isLarge = -> { _1 > 100 }

Interestingly, in the Lambda Calculus, where everything is a function, a standard representation for a natural number n (i.e. a whole number >= 0), is indeed a function that 'iterates' (strictly, folds/recurses) n times.

E.g. 3:

(f, x) => f(f(f(x)))

The "aesthetically pleasing" aspect of blocks is not mutually exclusive with real, first-class functions! Ruby is really more functional than that. Ruby has both lambas and method objects (pulled from instances). For example, you can write:

  let isLarge = a => a>100;

  is_large = ->(a) { a > 100 }
  is_large.call(1000)
  # => true
  is_large.(1000)
  # => true

Additionally, you can even take a reference to a method off of an object, and pass them around as though they are a callable lambda/block:

  class Foo
    def bar = 'baz'
  end

  foo_instance = Foo.new
  callable_bar = foo_instance.method(:bar)
  callable_bar.call
  # => 'baz'

  class UpcaseCertainLetters
    def initialize(letters_to_upcase)
      @letters_to_upcase = letters_to_upcase
    end

    def format(str)
      str.chars.map do |char| 
        @letters_to_upcase.include?(char) ? char.upcase : char
      end.join
    end
  end

  upcase_vowels = UpcaseCertainLetters.new("aeiuo").method(:format)
  ['foo', 'bar', 'baz'].map(&upcase_vowels)
  # => ['fOO', 'bAr', 'bAz']

  (0..10).map(&:even?)
  # => [true, false, true, false, true, false, true, false, true, false, true]

  is_div_five = ->(num) { num % 5 == 0 }
  (0..10).map(&is_div_five)
  # => [true, false, false, false, false, true, false, false, false, false, true]

That is interesting! I haven't explored Procs much, since I use ruby for a shared codebase at work and I was originally a bit afraid of trying to push unidiomatic ideas in the codebase.

In your experience, is it ok to use Procs for example for extraction of block methods for cleanliness in refactors? or would I hit any major roadblocks if I treated them too much like first-class functions?

Also, is there any particular Rails convention to place collections of useful procs? Or does that go a bit against the general model?

> In your experience, is it ok to use Procs for example for extraction of block methods for cleanliness in refactors?

In addition to the efficiency issues of unnecessarily reifying blocks into Proc objects, there are some sharp edges that can cut you doing this. For instance, Ruby blocks and procs [0] have return semantics based on where they are defined (specifically, "return" in a block or proc returns from the method in which the block/proc is defined.) So if you have two identical blocks used in two different methods, and those blocks have return calls, and you refactor them to shared Proc defined somewhere else that is used in each method, you will find that the behavior is broken, and when it returns it will either return from the wrong method or produce a LocalJumpError, depending on the exact structure of the program.

Local return from a block/proc is done with "next" instead of "return" (and if neither a next or return is hit, the last expression evaluated will be treated as an implicit "next"), so a block without an explicit return can be converted to a proc and reused in different places more safely.

[0] using the convention of lowercase "proc" to refer to a Proc object that does not have lambda semantics, whereas those created with Kernel#lambdda or the stabby-lambda syntax are "lambdas".

You shouldn't have much difficulty, Ruby converts blocks to Procs whenever it needs an actual object. Their semantics are intentionally kept the same. This is unlike lambdas, whose semantics are closer to methods.

Pass the wrong number of of arguments to a Proc or block, it will pass nil for missing args and omit extras. Pass the wrong number of arguments for a method or lambda and you get an ArgumentError. Use the return keyword in a lambda, it returns from the lambda, just like if you call return in a method. In a block or Proc, it returns from the calling method.

So I would feel comfortable leaning on them for refactoring as it's as Ruby intended. Just use lambdas when you want to turn methods into objects and Procs when you want to objectify blocks.

You should get ahold of a copy of Metaprogramming Ruby 2 if you find yourself refactoring a lot of Ruby. It's out of print, but ebooks are available.

Just to clarify here: Both lambdas and procs are Proc objects. Blocks gets turned into Proc objects when you take their value.

So just be clear about whether you're talking about a proc or a Proc...

> In a block or Proc, it returns from the calling method

No, in block or a Proc it returns from the scope where it was defined.

Usually this is the same, and so it doesn't usually matter much, but if you pass a proc or a block down to another method, then a return within the block will still return from the method where it was defined.

This can occasionally be useful, as you can use it to pass in a predicate to a method that can control if/when you want to return, irrespective of how deeply nested.

Too late to edit now, and this is what I get for quibbling about casing:

> No, in block or a Proc it returns from the scope where it was defined.

Should of course read:

> No, in a block or a proc it returns from the scope where it was defined.

> Use the return keyword in a lambda, it returns from the lambda, just like if you call return in a method. In a block or Proc, it returns from the calling method.

In a block or proc, it returns from the method in which the block/proc was defined, not the calling method. These may or may not be the same; the proc may be called deeper in the call chain, or it might (erroneously) be called after the defining method had exited, producing a LocalJumpError.

This also means that implicit returns from a block/proc have different semantics than explicit returns, unlike lambdas/methods where implicit returns have the same semantics as explicit returns. (From a block/proc, the last evaluated expressions behaves as if it were passed to "next" rather than "return", resulting in a local return to the calling method rather than a return from the defining method.)

This sounds like a really innovative idea. I haven't seen a dedicated place for "collection of useful procs", but one emerging pattern is to use `app/services` and then have a bunch of single-responsibility service classes that each have call or perform method and then you use the service when you need some shared functionality. It's like a proc, but instead it's a class with `#call` method.

> It's like a proc, but instead it's a class with `#call` method.

It's called the "callable" pattern.

Sorry for the late reply. Some others have responded and I mostly agree, but I will add where I can.

>is it ok to use Procs for example for extraction of block methods for cleanliness in refactors?

As others have said, use Lambdas via `->() { do_thing }` over Procs if you want to pass them around. They are the "true" way to use first-class functions.

Also, if you want to pass methods around, you can do so, just be aware that they carry everything you might expect with them: they are called with the receiver still being the object they came from. Callbacks are uncommon in Ruby, but I've seen a few instances of passing methods of the current object into some other as a callback, e.g.

  some_interactor.do_complex_thing(
    order_id: 123, 
    on_success: method(:succeeded),
    on_failure: method(:failed),
  )

>Also, is there any particular Rails convention to place collections of useful procs?

No convention for a "place" for these. I generally dislike how Rails got people into using directory names for the "kind" of a module/class instead of actually mirroring the architecture. This is kind of why.

If what you really desire is a big bag of pure functions in the global scope, you should still at least put them inside of a namespace (a Ruby module). For example, if you wanted a big bag of type coercions, you might put them into a module:

  module CommonCoercions
    class << self
      def currency
        ->(value) { do_things }
      end

      def coordinates
        ->(value) { do_things }
      end
    end
  end

  module CommonCoercions
    class << self
      def currency
        @currency ||= ->(value) { do_things }
      end

      def coordinates
        @coordinates ||= ->(value) { do_things }
      end
    end
  end

  module CommonCoercions
    class << self
      def currency(value)
        do_things
      end

      def coordinates(value)
        do_things
      end
    end
  end

  CommonCoercions.method(:currency)

However, if you want to use them only locally within a particular class, in a Rails project you are better off just calling them directly on the module, or even delegating so that the method is available in the class as though it were a local instance method:

  module CommonCoercions
    def self.coerce_currency(value)
      do_things
    end
  end

  class SomeAction
    delegate :coerce_currency, to: CommonCoercions

    def some_method(params)
      coerce_currency(params.fetch(:amount))
    end
  end

Ruby is my favorite language. There is no other language with same ergonomics. I dont understand hype around ts and all frameworks that changes every week.

I've been very taken by Ruby and how it uses blocks everywhere! This is an article I wrote just to emphasize that.

I discovered Ruby (through Rails) about twenty years ago on the dot. Coming from Perl and PHP it took me a while, but I remember the moment when I had the same realisation you did.

I still love this language to bits, and it was fun to relive that moment vicariously through someone elses eyes. Thanks for writing it up!

I'm glad folks are having fun reading this. I want to write a few more articles, particularly dissecting the internals of Ruby and how amazing it feels.

Have a look at Smalltalk blocks, or FP languages, to see where Ruby's inspiration comes from.

Ruby is basically what happens when Smalltalk and Perl (compare the command line and the "$"-variables for example) has a baby and it inexplicably looks much better than both parents.

An interesting thing about Smalltalk and Ruby blocks is that they aren't just anonymous functions/lambdas, right? i.e. if you 'return' / '^' in a block, it's the context around the block that you return from, not just the block itself? That's what struck me about both of them when I was used to thinking in basic Lisp terms.

As much as I love Ruby (and I love it a lot - I right now have Claude fix eigenclass compilation issues in another window in my long-languishing Ruby compiler hobby project), this is also the root of a lot of confusion.

Ruby has two types of closures: lambda's, where return returns to the caller, and proc's/blocks where return acts as a return in the defining scope.

Beginners often struggle with the distinction between lambda and block, and often will be wildly confused if they ever see "proc" in the wild.

Yet lambda and proc both return a Proc object, and if you take the value of a block, you also get a Proc object... Just with different return semantics...

It'd be nice if this was a bit clearer.

Yes, that is why that behaviour is known as closures.

Also why in languages like C++, you get to control what is captured from the calling context.

My next post, which is on loops, is about the common stuff with smalltalk as well!

Take a look at Kotlin, it perfected this idea

What Kotlin offers is already present in Scala or other languages from ML linage.

I believe the underlying behaviour of Ruby blocks is one of those mechanics that isn't talked about that much for newcomers, they just get used to how Ruby code look like when they see Rails, Cucumber or RSpec

Blocks and method_missing is one of those things in Ruby is what makes it so powerful! I remember watching a talk where someone was able to run JS snippets on pure Ruby just by recreating the syntax. That proves how powerful Ruby is for creating your own DSL

It's also a double edged sword and something you have to be careful with on collaborative codebases. Always prefer simplicity and readability over exotic and neat tricks, but I understand the difficulty when you have access to such a powerful tool

IMO blocks are not something to be careful about in ruby. If you don’t use blocks you are the weirdo in this language.

Method missing is a different beast altogether. I would probably avoid it nowadays.

There are legitimate uses of method_missing (though often you'd want to call define_method dynamically instead for performance), but they tend to be when you use an object as a proxy for something (say a remote API) where you genuinely can't know the set of messages the object might need to be able to receive.

Blocks yield a lot more flexibility to ruby. It was the primary reason why they are so well-appreciated.

Ruby is still so good to read and hack things with. It is a shame that it is not so popular and you know the reasons why. I still wish with such a good DSL friendly structure, it should have become an IaC de-facto standard language or used in some niche where I could use it freely without question.

The biggest thing holding Ruby back is lack of gradual typing imo. I honestly think javascript is a better fit for IaC. Not only is already the language of the web (everyone has to know some javsacript) but JSDOC is supported by most IDEs giving it gradual typing. Many people don't realize jsdoc is typescript and is a full replacement.

Nowadays I like to reach for Julia for quick one-off scripts and webscraping. It has a beautiful and accessible syntax with built-in gradual typing. I would love to see it more widely adopted in the IaC world

> The biggest thing holding Ruby back is lack of gradual typing imo.

There is: https://sorbet.org/

I've used Sorbet a lot but don't really count it. I understand why others would but I find the type system is extremely shallow and limited and the overhead it adds to development (and even performance) is substantial.

Also Ruby has RBS now which is not inline and... much maligned to say the least. I think the entire ecosystem is at a crossroads rn wrt typed Ruby

And also RBS-inline if you want Jsdoc style typing

Wow I hadn't seen this before. Pretty neat, thanks. Hope this project really takes off and gains more support

No worries, I also think it deserves a bit more highlight, especially to those who are against having rbs as separate file and to those who despise the Sorbet DSL in Ruby. The plan with Rbs-inline is to merge with rbs-gem so it will come included in Rbs!

It was revolutionary for readability and has influenced a lot of newer languages. That's a pretty good legacy IMO.

Ruby Blocks and how they're used literally everywhere is one of the hallmarks of why it's so nice to write the first time in that language; and it shaped how I use and select future languages.

Ruby Blocks are almost certainly the reason why I love Kotlin so much - it feels like a well-typed, curly-bracket-styled Ruby in those ways. The collection operation chains in both languages just. feel. good. And I blame Ruby for my first exposure to them, and possibly a lot of people's early exposure to them that helped languages that came after become better.

Ruby does take it to a-whole-nother level though, in particular with its 'space as separator' syntax, so you can make a *robust* DSL that's even more powerful than Kotlin's "if the last param is a function, you can just put a curly bracket and go" style.

I still don’t get how

  it "adds two numbers" do
    calc = Calculator.new
    expect(calc.add(2, 3)).to eq(5)
  end

  test "add two numbers":
    let calc = Calculator.new
    check calc.add(2, 3) == 5

Your example is Rspec, a testing framework for Ruby, not everyone enjoys it

Here's with Minitest (part of std)

    it "adds two numbers" do
      calc = Calculator.new
      assert_equal 5, calc.add(2, 3)
    end

    assert calc.add(2, 3) == 5

Ruby is really let down by the tooling around the language. The language itself would be so much more fun to write if the lsp would reliably jump to the definition of functions etc that seem to appear out of no where. It has been the biggest source of frustration for me while learning Ruby.

You're right! Although I get faily far by using Bust-a-gem VS Code extension. (The underlying ripper-tags gem can work with any IDE) https://github.com/gurgeous/bust-a-gem

I have an "on save" hook that runs ripper-tags on every file save. This keeps the definitions always up to date.

I'm hoping the new age tooling that's coming out is going to make things better. I want to contribute to rv personally.

I say the opposite, the lack of tooling highlights the weakness of the language. The drive to make it declarative/mimic "natural" language by reshuffling and overloading can be "delightful" to some, but beyond the paper covered surface is a mess. And for what ? "5.times do something unless" isn't cute to me. It's a dog "talking" by putting peanut butter in his mouth. But I think I'm the only one who feels this way

What I've noticed about language design is that people confuse their preferences and opinions with genuine objective observations. It's more of an art and aesthetics than a science. I agree "5.times" is an abomination but that's a subjective reaction - there's no principle or law that it violates (other than perhaps principle of least surprise). We can't quantify and measure how good or bad some invented syntax is.

As for lack of tooling, it apparently wasn't a priority when the language was designed. I'm guessing the emphasis was more on the ease of reading/writing, and having internal logical consistency. How such language design decisions make the development of tooling more difficult was a secondary concern.

And it's twice as bad on windows.

This level of cuteness and obsession with syntax is partly what drives me away from Ruby. A function should just be a function. We don't need to make programming languages look more like English. There are certainly issues with other languages and their verbosity (like with Java). But I don't want to have more ways to do the same thing and worry about how poetic my code reads as English - we should worry how poetic it reads as computer code.

That said, we don't need just one programming language. Perhaps Ruby is easier to learn for those new to programming and we should introduce it to students.

I think Ruby teaches a sense of style, but I'm not sure that style carries over to other languages. Python was my primary language for 12 years but I'm disappointed in the depth of Knowledge python Devs have. Most barely understand the language or try to. Ruby seems to coax people into coding into a ruby way. I like that.

> but I'm disappointed in the depth of Knowledge python Devs have

Not super relevant, but my favorite fun fact about Python (specifically CPython) is that hash(x) where x is an int is always the value of the int - except for -1, because the underlying C function uses -1 to represent errors. So hash(-1) == -2, which means you have a hash collision right around 0 where hash(-1) == hash(-2)

Is a block basically a lambda or is there more to it?

Assuming that by lambda you mean "an anonymous function"[1], for most intents and purposes they are, except their returns are non-local whereas functions usually have local returns.

However blocks are special forms of the language, unless reified to procs they can only be passed as parameter (not returned), and a method can only take one block. They also have some oddities in how they interact with parameters (unless reified to lambda procs).

[1] because Ruby has something called "lambda procs"

Basically. But Ruby has two types of lambdas, one called lambda, and one called proc. Blocks act like the second type.

The difference is that if return is called in the former, you return to the expression after where it is called, while I'm the latter return returns from the scope where it is defined.

This is usually going to feel reasonably intuitive, but can become weird if you e.g. reify a block into a named object and pass it around.

It can be very useful though. E.g. you can pass a block down into a recursive algorithm as a means to do a conditional early exit when a given criteria is met.

Notably you can do this even if a piece of code takes a named argument instead of a block and doesn't anticipate that use and expect you to pass a lambda instead of a process...

In most of the typical uses of a block, most people will just naturally expect the behaviour you see, because the block looks and feels like part of the scope it will return from.

My understanding is that the 'extra thing' is control flow - blocks can force a return in their calling scope. For example a loop that calls a block may be terminated/skipped by a break/continue statement in the block itself. However I'm not a Ruby programmer, so please check my working.

Blocks can force a return in their defining scope. If you pass a block down through multiple methods and then call it, the return will exit the lexical scope the block was defined in, not the method it was ultimately called from.

It usually looks pretty much the same, and so for most intents what you described is "close enough" to get most used of blocks.

They are closures. But Ruby can do interesting and slightly reckless things, like transplanting a closure into a different evaluation scope. It’s very powerful, and also very dangerous in the wrong hands.

Sounds a bit like a lisp macro? Or in JS using eval?

It's a closure. The lisp equivalent is also a closure, and the JS equivalent is a function (which is a closure as they can refer to variables in the scope they're defined in. To test this define a function inside a function and reference variables defined in the outer function, and return the inner function).

I mean, that is the point of a closure. The name refers to closing over its environment.

In terms of implementation, the closure needs to keep a reference to its environment, so it may be "transplanted" but it effectively runs in its own scope.

That's what I'm saying: In Ruby, you can transplant code from a closure out of its lexical scope and evaluate it as if it was defined in a different environment.

Should you ever do this? No. But you can. :-)

My point is that this isn't "reckless" - it's a core part of the value of a closure.

We do this regularly enough in Ruby - usually with lambda rather than proc/blocks because of the return semantics, but you can do it just as safely with proc/blocks as long as you don't call "return" within them (there's no reason to, on the other hand).

It's a way of e.g. building generator methods that returns distinct callable objects without constructing a whole class. In that case the closure environment effectively becomes the instance variables.

It is the opposite of what a closure typically means, because it is precisely not “closing” over its lexical scope.

It’s an interesting feature to be sure, and it enables some really beautiful things, but it also enables some incredibly difficult bugs.

I recommend reading Shopify CEO Tobi's try[0] for good example of how Ruby's block behavior and meta-programming makes it easy to create a single file, shell wrapper.

[0]: https://github.com/tobi/try/blob/main/try.rb

what the fuck...

This is the first time I ever got ruby. Reading it always felt like magic to me and I think it's still too much, if I really want to understand the abstractions, which unfortunately I do. But the understanding of blocks and especially instance_eval has helped a lot.

I'm glad I helped a bit. I'm going to write a few more posts since I just started a job where I need to do Ruby full time. I'm writing about Loops today.

I heard some rumor that ruby was a type 2 lisp. There were some guys who rate programming languages on how much lisp features they reassemble. What is the source?

It's only more readable if you already understand it. Otherwise it is not, it requires the same kind of hand waving that happened at the start.

Important to understand that readability doesn't mean it should be closer to natural language, in programming it means that a junior dev troubleshooting that code later down the line can easily understand what's happening.

The python examples are certainly more readable from a maintainability and comprehension standpoint. Verbosity is not a bad thing at all.

> It's only more readable if you already understand it. Otherwise it is not, it requires the same kind of hand waving that happened at the start.

Yes, but it’s a core language feature, so if you spend any time programming Ruby, you’ll come to understand it.

Ruby is beautiful.

It's weird, and different and therefore a bit repulsive (at least to me it was) at first. But, once you learn it, it's so easy to read it and to understand what's going on.*

* Side-note: Sometimes variables or methods look the same as parenthesis () are optional. So, yes, there's more things that can look like magic or be interpreted in multiple ways, but more times than not, it helps to understand the code faster, because `clients` and `clients()` (variable or method) doesn't matter if all it does is "get clients" and you just need to assume what's stored/returned from that expression. Also "get clients" can be easily memoized in the method implementation so it gets as close as possible to being an actual variable.

I can’t unsee it either. Will try it later

Welcome

blocks are just procs :)

It does look like it but procs encapsulates a block, but a block alone is not a proc

That’s just a performance optimization because blocks are typically only invoked, not passed or otherwise manipulated. If you want to do anything with the block, you pay a tiny tax and have a Proc instance.

Was this human-generated? Clearly it's self-submitted, and the useless "hero image" looks like lazy AI slop with Pixar characters, but did a human generate the text?

I wrote this entirely. I did use Gemini for the post image, I cannot draw and I'm not using my blog to monetize anything so I don't think there's anything wrong in AI generated images.

With autocomplete being so good today, do we really need languages with cryptic syntax that obscure what is going on?

To an outsider watching Ruby, it's cryptic, esoteric and maybe magical. But when you actually use it and learn mechanism underneath it, things start to make sense. However, Rubys dynamic architecture have also made it difficult for DX. Things like autocomplete barely works in the ecosystem, because it's so unpredictable until runtime

Market forces will ensure this system isn’t utilized regardless of how good it is. Ruby is simply to similar to Python to consider training people solely in that.

It’s a hard reality because I’m sure Ruby is better according to some criteria, but be realistic their share of the market is going to shrink until it’s not really an option of most large companies.

I know people will disagree with me, but I wish I knew earlier in my Carr how little this sort of thing matters.

Thankfully, programming languages are subject to strong network effects but not massive ones, Ruby will continue to be a niche language that has active users and projects for many more years than you expect.

That’s going to reassure nobody.

Let’s check back in 6 months after the AI bubble pops

Why? Python is already 10x the usage of Ruby outside of AI? Find a recent grad who even knows what Ruby is.