RE: [Swarm-Support] OSX

[Darold Higa]

I got back from Swarmfest.  It was lots of fun.

Back to compiling my simulation.

As I discovered later, apparently when I hastily copied over some of my
archived source files over to the OSX machine, somehow the file transfer
must have hung, because several of the files had the proper file name, but
were 0k.

I copied over my current source code form a valid archive and everything
seems to compile algright (with the same list of errors complaining about
libraries and defaulting to 10.1).

So far so good.

I ran the simulation, and it was humming along very nicely.  After it had
gone through a few iterations I assumed that it was up and running.

In order to speed up my simulations, I do not render any of the rasters
until a predetermined iteration.  The magic number appeared and it crashed
again, this time with the following message:

"../../../../../src/objectcase/MessageProbe.m:401 - [MessageProbe
doubleDynamicCallOn:]
 Abort"

so I track down the offending code and find that
-populationGraphScreenUpdate {
        [populationGrapher step];
        return self;
}

is the offending bit of code.
I commented out the message to populationGrapher, and it runs fine.

The odd thing is that I have 3 other active graphs, none of them share this
problem.  All of the graphs probe modelSwarm by calling a "get[blah]"
function.

Any help would be appreciated!

Darold Higa