[BioRuby] Bringing the fun back to programming! (The first BioRuby IRC conference on Dec 19th)

Raoul Bonnal bonnalraoul at ingm.it
Mon Dec 13 12:29:19 UTC 2010


+1 for screen cast which demostrates how jruby + scala or clojure and parallel programming can outperform a regular ruby script.

I have to admit that parallel programming in Ruby is not very different from years ago however there are interesting methods for implementing it -I can't fine the title of a Ruby's book on this topic-.

Unfortunately Ruby is a terrific language which suffers from the association Ruby => Rails

Other languages have a lot of libraries which are very solid, mathematics, statistics and charting/plotting which I think are the main libs that we need in our everyday work.

I was fascinated from JRuby and the ability to use java libraries is a +1 from my computer scientist/programmer point of view. I could use R whit a very solid tcp protocol for example. The same thing is not possible with Ruby 1.9 or 1.8 because the actual implementation lacks complex handling of dataframe.

Btw... I think that we should poll here who wants to actively contribute to BioRuby ?
Then I think we should address what we need to use bioruby/ruby in a more productive way or easiest way.

@Francesco: which are the motivation or real case studies which drive you to shift from Ruby to Python ?

Jan did a very interesting post on Clojure about how easy it's to plot data for NGS, are there alternative in ruby or is there someone here that faced the same problem and tried to find a solution with BioRuby ?


A library that I started to use some time ago is http://ruby-statsample.rubyforge.org/  it's not R but I got the same results whitout the need to use R inside my script.

Why is there a lot of interest of Jruby? My opinion is: Containers and Rails. I you want to use a rails application in an enterprise company you need to put it into a container: glassfish, jboss, Oracle.
The only problem with Java for me is Oracle itself but... I'm quite scared about that.

Ciao.



Il giorno 13/dic/2010, alle ore 12.42, Pjotr Prins ha scritto:

> On Sun, Dec 12, 2010 at 08:43:15PM +0100, Francesco Strozzi wrote:
>> Ruby isn't so powerful for bioinformatics compared to other
>> languages. So, like you, I feel the need to add another language for
>> "real" and "though" bioinformatics work and I have started learning
>> Python in the last months which it's really impressive. The core
>> library is very poweful and BioPython is updated every 3 months (so
>> it is under constant development). 
> 
> Main comment would be that Python is not much different from Ruby. It
> is interpreted, dymamic, and does not have good support for parallel
> programming, nor the speed required for big data (BD). Perl, Python
> and Ruby are really in the same boat.
> 
>> The performance of many common tasks (e.g. parsing a large Blast
>> report) are superior than with BioRuby libraries, for what I have
>> seen. 
> 
> Here you have a good point. BioRuby is lagging behind BioPython,
> BioPerl and BioJava.
> 
> My proposal is to use BioJava together with BioRuby. You can do
> something similar with Jython (Python on the JVM), though I have the
> impression Jython is not as active as JRuby, anymore. JRuby is
> getting strong industry support, at this point.
> 
>> But Ruby syntax and logic is simply amazing (and to me even
>> better than Python) and it's a shame we don't have powerful
>> solutions to bring Ruby out of the "just for the web" corner where
>> it was put since Rails outbreak. 
> 
> I know people who would disagree with that last statement...
> 
> To me, Ruby is a way of thinking, which is reflected in Rails.
> 
>> So, from my point of view, every effort
>> that could make Ruby shine for bioinformatics...well, is really really
>> appreciated and I want to hear and read more!
>> From the technical point of view I have a couple of (newbie) questions:
>> - I have no idea what Clojure and Scala are, so a sort of "Functional
>> Programming for Dummies" will be appreciated.
> 
> There are plenty online resources. The main Scala book is great.
> 
>> - How much Java one need to know in order to be productive with
>> Ruby/BioRuby and Java/BioJava using JRuby?
> 
> Not Java per se. Mostly it is getting used to the JVM world and docs.
> 
>>   Because from my point of view, a Ruby programmer I think doesn't really
>> want to learn Java (which is a big language but with it's own issues and is
>> not so flexible as Ruby or Python).
> 
> I agree. I personally dislike the Java language.
> 
> After a project around 2001 I vowed not to touch it again. It is not a
> hard language, but merely a pain. I liked any of the other Bio*
> languages better.
> 
>> So in the end, I will read and hear with interest what you can tell us about
>> your experience and future directions for Ruby/BioRuby development.
> 
> My suggestion would be to study a language more removed from Python,
> Ruby and Perl.
> 
> I will try to make a screencast.
> 
> Pj.
> 
> _______________________________________________
> BioRuby Project - http://www.bioruby.org/
> BioRuby mailing list
> BioRuby at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioruby






More information about the BioRuby mailing list