swarm-modeling
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Choice of model language


From: Jan Kreft
Subject: Re: Choice of model language
Date: Fri, 5 Oct 2001 10:09:44 +0100 (BST)

Hi,

On Tue, 4 Sep 2001, Methot, Chris wrote:

> Hi (2 questions below)
> 
> I am trying to decide on Objective-C or Java.  I have no predilections
> toward either and no C experience (which may imply a direction in itself).
> Perhaps there are those that feel the tighter restrictions of Java are more
> suited to beginners.  
> 
> I tried to do due diligence and search the ftp://ftp.swarm.org/ support and
> model archives but the ftp server has not been reliable from here lately.  
> 
> The only thing I subjectively noticed between them is that the interpreted
> Java example models seem to run slower than the Objective-C.  Both use every
> spare cycle of my CPU but I have the feeling that the C programs are doing
> more.  Now I have been doing performance work for 20 years so I know this
> subjective conclusion is fraught with errors but I do believe it is a safe
> conclusion based on what I have seen of interpreted code.  
> 
> Here are my 2 questions:
> 1. If compiled OOP-C is more efficient than interpreted Java can the Java
> models be turned into a native executable like C?  

Incidentally, I've just converted a snippet of Java code for solving
nonlinear PDE's into pure C by hand, in order to see how far off my Java
code is from C in terms of speed.

Running the Java version with IBM jdk1.3.0 on Linux, it was only 5% slower
that the C version compiled with gcc -O3 (-O[1-3] all same speed). So,
with a modern jdk, speed is no longer an issue.

Interpreted Java is dead slow, however. Factor 20 slower for my program.

In other words, choose Java.

HTH, Jan Kreft.

> 2. Is Java a better bet for a beginner who used Fortran and Basic many years
> ago? 
>       I am more likely to run into people who know C than Java (which is
> still primarily a web tool).  
> 
> Chris Methot
> 
> 
> 
> 
>                   ==================================
>    Swarm-Modelling is for discussion of Simulation and Modelling techniques
>    esp. using Swarm.  For list administration needs (esp. [un]subscribing),
>    please send a message to <address@hidden> with "help" in the
>    body of the message.
>                   ==================================
> 



                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================


reply via email to

[Prev in Thread] Current Thread [Next in Thread]