Ruby 3.0.0 has been officially released. Ruby 3, whose goal is performance, concurrency, and Typing. Especially about performance, Matz stated “Ruby3 will be 3 times faster than Ruby2” a.k.a. Ruby 3×3.

Optcarrot 3000 frames

With Optcarrot benchmark, which measures single thread performance based on NES’s game emulation workload, it achieved 3x faster performance than Ruby 2.0!

These were measured at the environment written in https://benchmark-driver.github.io/hardware.html. 8c510e4095 was used as Ruby 3.0. It may not be 3x faster depending on your environment or benchmark.

Ruby 3.0.0 covers those goals by

  • Performance
    • MJIT
  • Concurrency
    • Ractor
    • Fiber Scheduler
  • Typing (Static Analysis)
    • RBS
    • TypeProf

With above performance improvement Ruby 3.0 introduces a number of new features described below.

Performance

When I first declared “Ruby3x3” in the conference keynote, many including members of the core team felt “Matz is a boaster”. In fact, I felt so too. But we did. I am honored to see the core team actually accomplished to make Ruby3.0 three times faster than Ruby2.0 (in some benchmarks). – Matz

MJIT

Many improvements were implemented in MJIT. See NEWS for details.

As of Ruby 3.0, JIT is supposed to give performance improvements in limited workloads, such as games (Optcarrot), AI (Rubykon), or whatever application that spends majority of time in calling a few methods many times.

Although Ruby 3.0 significantly decreased a size of JIT-ed code, it is still not ready for optimizing workloads like Rails, which often spend time on so many methods and therefore suffer from i-cache misses exacerbated by JIT. Stay tuned for Ruby 3.1 for further improvements on this issue.

Concurrency / Parallel

It’s multi-core age today. Concurrency is very important. With Ractor, along with Async Fiber, Ruby will be a real concurrent language. — Matz

Ractor (experimental)

Ractor is an Actor-model like concurrent abstraction designed to provide a parallel execution feature without thread-safety concerns.

You can make multiple ractors and you can run them in parallel. Ractor enables you to make thread-safe parallel programs because ractors can not share normal objects. Communication between ractors are supported by exchaning messages.

To limit sharing of objects, Ractor introduces several restrictions to the Ruby’s syntax (without multiple Ractors, there is no restriction).

The specification and implementation are not matured and may be changed in the future, so this feature is marked as experimental and show the “experimental feature” warning when the first Ractor.new.

The following small program measures the execution time of famous benchmark tak function (Tak (function) – Wikipedia), by executing it 4 times sequentially or 4 times in parallel with ractors.

def tarai(x, y, z) =
  x <= y ? y : tarai(tarai(x-1, y, z),
                     tarai(y-1, z, x),
                     tarai(z-1, x, y))
require 'benchmark'
Benchmark.bm do |x|
  # sequential version
  x.report('seq'){ 4.times{ tarai(14, 7, 0) } }

  # parallel version
  x.report('par'){
    4.times.map do
      Ractor.new { tarai(14, 7, 0) }
    end.each(&:take)
  }
end
Benchmark result:
          user     system      total        real
seq  64.560736   0.001101  64.561837 ( 64.562194)
par  66.422010   0.015999  66.438009 ( 16.685797)

The result was measured on Ubuntu 20.04, Intel(R) Core(TM) i7-6700 (4 cores, 8 hardware threads). It shows that the parallel version is 3.87 times faster than the sequential version.

See doc/ractor.md for more details.

Fiber Scheduler

Fiber#scheduler is introduced for intercepting blocking operations. This allows for light-weight concurrency without changing existing code. Watch “Don’t Wait For Me, Scalable Concurrency for Ruby 3” for an overview of how it works.

Currently supported classes/methods:

  • Mutex#lockMutex#unlockMutex#sleep
  • ConditionVariable#wait
  • Queue#popSizedQueue#push
  • Thread#join
  • Kernel#sleep
  • Process.wait
  • IO#waitIO#readIO#write and related methods (e.g. #wait_readable#gets#puts and so on).
  • IO#select is not supported.

This example program will perform several HTTP requests concurrently:

require 'async'
require 'net/http'
require 'uri'

Async do
  ["ruby", "rails", "async"].each do |topic|
    Async do
      Net::HTTP.get(URI "https://www.google.com/search?q=#{topic}")
    end
  end
end

It uses async which provides the event loop. This event loop uses the Fiber#scheduler hooks to make Net::HTTP non-blocking. Other gems can use this interface to provide non-blocking execution for Ruby, and those gems can be compatible with other implementations of Ruby (e.g. JRuby, TruffleRuby) which can support the same non-blocking hooks.

Static Analysis

2010s were an age of statically type programming languages. Ruby seeks the future with static type checking, without type declaration, using abstract interpretation. RBS & TypeProf are the first step to the future. More steps to come. — Matz

RBS

RBS is a language to describe the types of Ruby programs.

Type checkers including TypeProf and other tools supporting RBS will understand Ruby programs much better with RBS definitions.

You can write down the definition of classes and modules: methods defined in the class, instance variables and their types, and inheritance/mix-in relations.

The goal of RBS is to support commonly seen patterns in Ruby programs and it allows writing advanced types including union types, method overloading, and generics. It also supports duck typing with interface types.

Ruby 3.0 ships with rbs gem, which allows parsing and processing type definitions written in RBS. The following is a small example of RBS with class, module, and constant definitions.

module ChatApp
  VERSION: String
  class Channel
    attr_reader name: String
    attr_reader messages: Array[Message]
    attr_reader users: Array[User | Bot]              # `|` means union types, `User` or `Bot`.
    def initialize: (String) -> void
    def post: (String, from: User | Bot) -> Message   # Method overloading is supported.
            | (File, from: User | Bot) -> Message
  end
end

See README of rbs gem for more detail.

TypeProf

TypeProf is a type analysis tool bundled in the Ruby package.

Currently, TypeProf serves as a kind of type inference.

It reads plain (non-type-annotated) Ruby code, analyzes what methods are defined and how they are used, and generates a prototype of type signature in RBS format.

Here is a simple demo of TypeProf.

An example input:

# test.rb
class User
  def initialize(name:, age:)
    @name, @age = name, age
  end
  attr_reader :name, :age
end
User.new(name: "John", age: 20)

An example output:

$ typeprof test.rb
# Classes
class User
  attr_reader name : String
  attr_reader age : Integer
  def initialize : (name: String, age: Integer) -> [String, Integer]
end

You can run TypeProf by saving the input as “test.rb” and invoke a command called “typeprof test.rb”.

You can also try TypeProf online. (It runs TypeProf on the server side, so sorry if it is out!)

See the documentation and demos for details.

TypeProf is experimental and not so mature yet; only a subset of the Ruby language is supported, and the detection of type errors is limited. But it is still growing rapidly to improve the coverage of language features, the analysis performance, and usability. Any feedback is very welcome.

Other Notable New Features

  • One-line pattern matching is redesigned. (experimental)
    • => is added. It can be used as like rightward assignment.
      0 => a
      p a #=> 0
      
      {b: 0, c: 1} => {b:}
      p b #=> 0
      
    • in is changed to return true or false.
      # version 3.0
      0 in 1 #=> false
      
      # version 2.7
      0 in 1 #=> raise NoMatchingPatternError
      
  • Find pattern is added. (experimental)
    case ["a", 1, "b", "c", 2, "d", "e", "f", 3]
    in [*pre, String => x, String => y, *post]
      p pre  #=> ["a", 1]
      p x    #=> "b"
      p y    #=> "c"
      p post #=> [2, "d", "e", "f", 3]
    end
    
  • Endless method definition is added.
    def square(x) = x * x
    
  • Hash#except is now built-in.
    h = { a: 1, b: 2, c: 3 }
    p h.except(:a) #=> {:b=>2, :c=>3}
    
  • Memory view is added as an experimental feature
    • This is a new C-API set to exchange a raw memory area, such as a numeric array and a bitmap image, between extension libraries. The extension libraries can share also the metadata of the memory area that consists of the shape, the element format, and so on. Using these kinds of metadata, the extension libraries can share even a multidimensional array appropriately. This feature is designed by referring to Python’s buffer protocol.

Performance improvements

  • Pasting long code to IRB is 53 times faster than bundled with Ruby 2.7.0. For example, the time required to paste this sample code goes from 11.7 seconds to 0.22 seconds.

  • The measure command has been added to IRB. It allows simple execution time measurement.
    irb(main):001:0> 3
    => 3
    irb(main):002:0> measure
    TIME is added.
    => nil
    irb(main):003:0> 3
    processing time: 0.000058s
    => 3
    irb(main):004:0> measure :off
    => nil
    irb(main):005:0> 3
    => 3
    

Other notable changes since 2.7

  • Keyword arguments are separated from other arguments.
    • In principle, code that prints a warning on Ruby 2.7 won’t work. See the document in detail.
    • By the way, arguments forwarding now supports leading arguments.
      def method_missing(meth, ...)
        send(:"do_#{ meth }", ...)
      end
      
  • Pattern matching (case/in) is no longer experimental.
  • The $SAFE feature was completely removed; now it is a normal global variable.
  • The order of backtrace had been reversed at Ruby 2.5, and is reverted. Now it behaves like Ruby 2.4; an error message and the line number where the exception occurs are printed first, and its callers are printed later.
  • Some standard libraries are updated.
    • RubyGems 3.2.3
    • Bundler 2.2.3
    • IRB 1.3.0
    • Reline 0.2.0
    • Psych 3.3.0
    • JSON 2.5.1
    • BigDecimal 3.0.0
    • CSV 3.1.9
    • Date 3.1.0
    • Digest 3.0.0
    • Fiddle 1.0.6
    • StringIO 3.0.0
    • StringScanner 3.0.0
    • etc.
  • The following libraries are no longer bundled gems or standard libraries. Install the corresponding gems to use these features.
    • sdbm
    • webrick
    • net-telnet
    • xmlrpc
  • The following default gems are now bundled gems.
    • rexml
    • rss
  • The following stdlib files are now default gems and are published on rubygems.org.
    • English
    • abbrev
    • base64
    • drb
    • debug
    • erb
    • find
    • net-ftp
    • net-http
    • net-imap
    • net-protocol
    • open-uri
    • optparse
    • pp
    • prettyprint
    • resolv-replace
    • resolv
    • rinda
    • set
    • securerandom
    • shellwords
    • tempfile
    • tmpdir
    • time
    • tsort
    • un
    • weakref
    • digest
    • io-nonblock
    • io-wait
    • nkf
    • pathname
    • syslog
    • win32ole

See NEWS or commit logs for more details.

With those changes, 4028 files changed, 200058 insertions(+), 154063 deletions(-) since Ruby 2.7.0!

Ruby3.0 is a milestone. The language is evolved, keeping compatibility. But it’s not the end. Ruby will keep progressing, and become even greater. Stay tuned! — Matz

Merry Christmas, Happy Holidays, and enjoy programming with Ruby 3.0!

Leave a Reply

Your email address will not be published.

You may use these <abbr title="HyperText Markup Language">HTML</abbr> tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

*

%d bloggers like this: