Ruby Interpreters: MRI Vs. JRuby

Have you ever wondered what the deal is with JRuby? What is it? Why would you use it in the first place? In this talk, I’ll explain the basics concepts behind a Ruby interpreter and outline a few of the reasons why you may or may not want to consider using JRuby.

Transcript

1. Ruby Interpreters By Katlyn Parvin 01001000 01100101 01101100 01101100 01101
2. MRI Matz’s Ruby Interpreter JRuby In particular... &
3. But before we get to tonight’s main event...
4. Interpreters Let’s learn a little about them! • Why do we need an interpreter? • How do they work? ?
5. Interpreter A computer program that interprets and translates source code into a format that the computer can understand.
6. To start at the beginning... we will have to travel back in time to the early 1940s • Before the first interpreter in 1958 programs were written in assembly • Assembly languages required a lot of knowledge about the hardware and were very error-prone • Programs needed to be written in a way that the processor could understand
7. Ex. MIPS Assembly Language Load the data at location 32($s3) in memory to the local register $t0 Perform the addition of $t0 and $s2 and store the result in register $t0 Store the value of register $t0 to location 48($s3) in memory Ruby Program my_array[12] = 10 + my_array[8] Equivalent MIPS Program
8. High-level Languages Strongly abstracted from the implementation details of the computer C Java C++ Ruby
9. • Scans the entire program and translates it as a whole • Fast execution time • Compile once -> run many times Compiler • Translates the program one statement at a time • Slower execution time • Faster analysis of the program • Translates the source code every time it is run Interpreter Converting from high-level language or
10. Cool but, how do interpreters work?
11. Let’s see how things work inside of MRI
12. Example Code # Ruby Example x = 10 y = 5 puts x + y
13. Step 1 Tokenize Breakdown the ruby code into structured data # Ruby Example x = 10 y = 5 puts x + y
14. Step 2 Lexical Analysis Tagging each token as an identifier, operator or integer # Ruby Example x = 10 y = 5 puts x + y
15. Step 3 Build a Abstract Syntax Tree Representation of your program in memory # Ruby Example x = 10 y = 5 puts x + y
16. Step 4 Compile to Bytecode Run in YARV the ruby VM (Ruby 1.9+) # Ruby Example x = 10 y = 5 puts x + y
17. If you want to learn more about interpreters... I would highly recommend this talk by Ryan Davis From the 2013 Golden Gate Ruby Conf https://youtu.be/RPxvx9OkNic
18. But back to MRI and JRuby...
19. MRI Matz’s Ruby Interpreter - Written in C by the core Ruby team - Sets the standard for the Ruby Language - Well maintained with more paid core team members than any other Ruby Implementation - Consistent release cycle - Most likely what you are using to interpret Ruby
20. JRuby - Written in primarily java - Implemented on top of the JVM - Which means you can run Ruby code anywhere Java can run. - Ruby 2.3 Compatible - Actively being maintained - Supports true parallel multi-threading - Not all gems work with JRuby
21. Cool things about JRuby • Use Java libraries ◦ Great for companies that maintain Java codebases ◦ Utilize 3rd party Java Libraries ▪ Ex. Android development (ruboto) • Use ruby in your Java files ◦ Ex. utilize a ruby codebase within an existing Java app • Allows for a lot more performance optimizations options ◦ Tuning the JIT compiler ◦ Threading
22. True Multi-Threading • Ruby 1.8.7 the GIL (Global Interpreter Lock) prevent more than 1 thread of Ruby code to run at any time. • Ruby 1.9+ can use multiple OS threads for I/O bound operations, but we cannot have parallel ruby processes • JRuby can use true OS threads allowing us to utilize multiple CPUs http://www.restlessprogrammer.com/2013/02/multi-threading-in-jruby.html
23. Other interpreters to keep an eye on... Designed by Matz to be an embeddable implementation of Ruby Ruby interpreter written in RPython Implementation of Ruby that implements a generic runtime for dynamic languages on top of a Low Level Virtual Machine (LLVM) mRuby Topaz Rubinius Many more...
24. MRI Matz’s Ruby Interpreter JRuby In conclusion... &
25. Thank you! Katlyn Parvin Director of Engineering @ Mavenlink We are hiring! www.mavenlink.com/careers @KatlynParvin katlyn333@gmail.com