Tuesday, May 13, 2008

JRuby Won & It Was Fun

JavaOne is all about interesting demos, latest buzz around Java and some painful marketing talks. I didn't attend JavaOne this year, but I've read a few stories online about the different keynotes, demos and followup discussions. I just finished a conversion with a group of friends/colleagues who attended JavaOne 2008 and most of them enjoyed the show.

One interesting session at JavaOne 2008 was hosted by Raghavan aka Rags from Sun Microsystems. The session was called the Script Bowl and was a war between the scripting gurus from Groovy, JRuby, Jython and Scala. All the four languages are dynamic scripting languages that run on the JVM. LaForge represented Groovy, Nutter represented JRuby, Wierzbicki represented Jython and Jorge Ortiz represented Scala. All four showcased excellent skills and I watched the video one of my friends had taken.

And like the Idol (American - Indian whatever...), there was SMS voting and it was awesome competition between the scripting languages and their crowd of supporters watching the session. Finally JRuby won the day and it was a close fight. JRuby won big points at Round 3 (Best Language Features). You can read the full competition details and look below for the result details:.

A = Groovy; B = JRuby; C = Jython and D = Scala



Anonymous said...

JRuby is pretty slow at the moment. Scala is probably the fastest scripting language at the moment. And on the JVM the scripting languages suck even more...
Look at the Language Benchmark at Debian for proof!ldzsr

Anonymous said...

Why is scala considered a scripting language? It's an FP/OO mix with the Erlang inspired actor model. It has better typing that Java, actually.

Anonymous said...


There are several Ruby implementations by now. The one that the Debian benchmark uses is probably the standard Ruby implementation in the released version (1.8.x). Both the dev version of the standard implementation (1.9.x) and the JRuby implementation in its current version should do better.

May I suggest that you look at a Ruby benchmark and re-evaluate your statement?

And if you are interested in the topic of scripting languages on the JVM this post should provide some interesting reading.