Re: swarm support (was compare swarm with repast)

[Darren Schreiber]

Since I haven't had the technical knowledge to contribute anything on 
the discussions about Objective-C and Java, I have just been sitting 
back and enjoying the wide range of thoughts.  Although I know a number 
of programming languages, I have only learned one of them in a formal 
setting and that was back in high school, so my knowledge of the 
computer science-type issues is low.  Discussions like this help me to 
understand what the consequences of my implementation language choices 
are without having to research issues that are not at the top of my 
list.  So, thanks to the contributors.

One of my concerns as it relates to this discussion regards teaching.  
As one of Lars Erik's students four years ago, I know how hard it was 
for me to grasp Objective-C and Swarm.  Now, I feel very comfortable 
with both and can turn ideas into code pretty quickly (when I get a 
chance to code.)

But, I haven't had much time to code.  Part of that is because I am a 
strictly Mac person.  And, while I eventually did get Swarm running on 
my G3 under a Linux PPC setup, it took a lot of work and I never really 
understood what I was doing (thanks again for Alex, Marcus, and Paul for 
helping me through that period.)  Unfortunately, about a year ago, the 
hard drive with Linux PPC on it died and I just haven't had the energy 
to go through another installation.

Further, I have a great new Mac Titanium Powerbook that I am loving and 
wouldn't want to mess up with partitioning for Linux PPC (which would 
seem particularly silly given the Unix/BSD/Darwin implementation.)  All 
of the brain imaging analysis software that was originally developed for 
multiprocessor Unix systems, Silicon Graphics, Alphas, and Sun 
workstations is now on my laptop, so I have no need to trek to the 
opposite side of campus and use their equipment for fMRI analysis.  But, 
whenever I have a good modeling idea.  My only current option is to go 
to my old office at the law school and use it on the Swarm install I 
have on their PC.  Far from ideal.

So, I have been hoping Swarm on Mac OSX would come out, but remain in a 
holding pattern on that.

And, when I think about my future role as an instructor, I fear the 
challenges Lars Erik had in getting everyone installed with Swarm (your 
modal poli sci grad student (or undergrad) is no where near the computer 
geek that I am and I am low on the computer geek totem pole for this 
mail list.)  I have to work hard to help grad students get synched up 
with using Excel or SPSS, and I have nightmares about what Swarm 
installation and programming would be like.

The installation of the last official version of Swarm was a cinch on 
Windows.  If something like that would be available for Mac OS X, it 
would be _very_ useful to me and allow me start sketching my 
computational model for my dissertation (a computational model of 
political cognition.)

My swarm options as I see them now are:

1) buy a PC
2) load Virtual PC on my powerbook and use Swarm on top of Linux or 
Windows (this is how I wrote my political parties model)
3) pray for Swarm on OS X and experience mystical ecstasy when I finally 
run Swam on my own computer again

So, this is my wandering plea for the Swarm port.  As a grad student, 
that is the best incentive I can offer (plus my next meager donation to 
the Swarm Development Group).  It isn't just a good idea because Macs 
are nice computers, it's a good idea because Swarm users like me are 
committed to Agent-Based Modeling and to Macs and yet lack the skill set 
to port software.

        Darren




On Monday, August 12, 2002, at 11:57 AM, Marcus G. Daniels wrote:

> > There exists the idea under the swarm developers that OSX is cool, but 
> > not to be taken seriously.  If we could change the believes of the 
> > agents, we would get another behavior from them.
>
>
> OS X is a based on Darwin, a BSD-like system based on the Mach 
> microkernel.
> The microkernel has had numerous extensions (loadable kernel drivers) 
> and optimizations (RPC calls optimized into function calls) and the 
> system as a whole works pretty well now.  At the application level,  OS 
> X is a very usable system.  The Microsoft Office suite is elegant, and 
> the Chimera Netscape browser renders beautiful text via Quartz.
> On the toolchain side, Apple has an active compiler team and their 
> extensions like precompiled headers will probably be incorporated into 
> the mainline GCC sources over time, or at least inspire a new 
> implementation.  Altivec SIMD support is in the latest release of stock 
> GCC.  Likewise, device support is very good.  Firewire devices work 
> better on Mac OS X than on Linux-based systems or Microsoft Windows, 
> for example.
>
> So I'd say I'd have as good as grasp as anyone here where Mac OS X is 
> at these days.  The reason this "swarm developer" is not working on 
> this project is simply because no one has provided me with an incentive 
> to do that.   (Other than that it is a "good idea" or whatever.)
>
>
>
>
>                  ==================================
>   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.
                 ==================================