Why I care to find out more about Rubinius
The purpose of today’s blog is to explore Rubinius. What is Rubinius?
What is Rubinius?
The following is what Rubinius says about itself: An implementation of Ruby
designed for concurrency using native threads to run Ruby code on all the CPU cores.
Rubinius runs Ruby code fast with a low-pause generational garbage collector
and LLVM-based just-in-time (JIT) native machine code compiler. The Rubinius
core library and tools are built using Ruby, making a smaller, consistent
system that’s easier for Ruby programmers to understand and modify. Rubinius
created and maintains RubySpec, containing more than 20,000 specifications,
to monitor compatibility with Matz’s Ruby Implementation (MRI). Rubinius
provides a C-API interface for running C-extensions to ease migrating from MRI.
( huh?!? ) Let’s unpack this description a bit:
My understanding of concurrency is that it basically does some task while another task might still be working. Normally, Ruby is not so good at concurrency. Concurrency is best for a functional language that uses recursion. Ruby sucks at recursion and I have an example to prove it:
def factorial(number)
if number <= 1
return 1
else
return number * factorial(number - 1)
end
endI used IRB in my terminal to run this Ruby method and got a “stack-level-too-deep” error at number = 8721.
Now, the same method using a functional language like Elixir:
defmodule Factorial do
def fact(0), do: 1
def fact(number), do: number * fact( number - 1 )
endUsing IEX, which is the Elixir version of IRB, I was able to quickly find the factorial of 20,000. I waited two minutes, but was able to go up to the factorial of 150,000.
So, if Rubinius is concurrent, and it is written in Ruby, then there must be some kick-ass stuff going on under the hood.
…ok, well I’m going to turn this into a multi-part because I’m getting bored and need to work on other stuff…