Re: Ever use electric fence with Swarm?

[Marcus G. Daniels]

>>>>> "D4" == D D Donalson <address@hidden> writes:

D4> Before you put a huge amount of effort into chasing this "bug".
D4> Check it on a W2000, and another Win 98 machine.  I had this
D4> problem a month or so ago and discovered that it was machine
D4> specific.  I was able to increase the amount of time until the
D4> fault, on average, by changing code around and making some
D4> dynamically allocated parameters hard coded, but was never able to
D4> get rid of the problem all together. 

That sounds like an array overrun problem to me. 

It isn't that unusual to find a machine with bad RAM, but it is rare
to find a malloc implementation that is broken.  Almost everything
relies on malloc working. 

Also, Cygwin is a pretty conservative environment.  The malloc
implementation is the same one that most Linux-based systems uses
(Doug Lea's), and the compiler is a heavily-tested version of GCC for Intel.
If a Swarm crashes in batch mode on two more-or-less identical Windows
machines, the responsible thing to do is search for a bug in your model.

Btw, these extremely weird memory overrun things don't happen with
Java Swarm models.



                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.