[Biojava-l] [Biojava-dev] Port an Algorithm to Java

[Amr AL-Hossary]

Dear Dr. Andreas,
There exists optimization when converting C code into Java code.
But on the level of implementation, not on on the level of algorithm itself.

Because java deals with data types in slightly different way, (e.g. unsigned 
numbers, double numbers, etc.) it will need some posttranslational 
modification to be optimized to work on java.

Amr

-----Original Message----- 
From: Andreas Prlic
Sent: Wednesday, March 28, 2012 2:06 AM
To: Dragos-Bogdan Sima
Cc: Biojava ; Scooter Willis
Subject: Re: [Biojava-l] [Biojava-dev] Port an Algorithm to Java

Hi Dragos,

many years of research have been put into both Blast and Hmmer and it
will be highly unlikely that you will be able to come up with
something better in a summer. As such best to focus on getting a plain
port. If you get ideas how to make the algorithms perform better
during that time, probably best to feed back that information to the
original developers.

> Secondly, converting C/C++ source code into Java would be a very 
> interesting
> and challenging task for me. At least for the C++ part I am thinkig to
> approach the use oh JNI, but is it possible to occur problems with
> portability, building, or JVM's stability? I have also read about NestedVM
> which provides binary translation for Java Bycode, and this approach could
> be useful.

Scooter, do you want to answer this one?

Thanks,

Andreas
_______________________________________________
Biojava-l mailing list  -  Biojava-l at lists.open-bio.org
http://lists.open-bio.org/mailman/listinfo/biojava-l