RE: Memory Leak in "schedule"

I run though the simulation using the Windows version of breve 1.7 and
here are the results.  I'll test it on the Mac version later.  I left
the number of objects to 80 but increased scheduling to time+0.1.

Initial memory usage (t=0): 9,036K
Breve time index t=100.82: 9,412K
T=251.02: 9,716K
T=500.23: 10,216K
T=1000.23: 11,220K
T=2000.90: 13,220K
Etc.

When I increased object count to 800, it initially starts with 21,020K,
it increases a little over the next 20 iterations but seems stable.
However, after 20 iterations, it starts to increase again consistently:
T=20.46: 28,768K
T=100.32: 28,920K
T=250.58: 29,208K
Etc.

Marcin


> -----Original Message-----
> From: [email protected]
[mailto:[email protected]]
> On Behalf Of jon klein
> Sent: December 10, 2003 6:57 AM
> To: [email protected]
> Subject: Re: [breve] Memory Leak in "schedule"
> 
> 
> On Dec 9, 2003, at 8:39 PM, Marcin Pilat wrote:
> 
> > The leak is due to the “schedule” method.  Something in there is not
> > being freed after the scheduled method call.  I’m attaching a simple
> > test file to demonstrate this – in that simulation the memory
> > increases at a rate of about 4KB/sec.  This can be attained by
calling
> > the schedule method from within a group of mobile objects every few
> > seconds or so.  Increasing the number of calls increases the leak.
> 
> I'm unable to reproduce the problem with the file that you sent.  I
> tried increasing the number of objects to  800 (from 80) and the
> frequency of scheduled calls to every .1 seconds (from 1.0 seconds).
I
> then let the simulation run and saw that after the first few simulated
> seconds, memory usage levelled off and didn't increase anymore.
> 
> Can you retest this?
> 
> - jon klein
> 
> _______________________________________________
> breve mailing list
> [email protected]
> http://www.spiderland.org/mailman/listinfo/breve


_______________________________________________
breve mailing list
[email protected]
http://www.spiderland.org/mailman/listinfo/breve