Failure to load large Squeak .image files

I am using Squeak 3.8.18 on  Mac OSX with 2 Gbyte RAM to model large
numbers of complex neurons. With all the neuron instances the Squeak
image size can sometimes get well above 500 Mbytes. Up to image sizes
of 500 Mbyte everything works fine.

The problem is that for images of 550 Mbyte, the image continues to run
and I can save it, but if I try to start Squeak with a 550 Mbyte image,
the image desktop and windows appear for a few seconds, then Squeak
quits.

I have tried starting Squeak with the large image on a different Mac
and the same thing happens. The problem occurs even if it is
immediately after rebooting the Mac, with no other applications
running.

Is this a bug, or are there Squeak image size limits and if so what are
they and what generates them. Does anyone have any suggestions how to
get around the problem?

Thanks

Andrew


_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Hello Andrew,

AC> The problem is that for images of 550 Mbyte, the image continues to run
AC> and I can save it, but if I try to start Squeak with a 550 Mbyte image,
AC> the image desktop and windows appear for a few seconds, then Squeak
AC> quits.
there have been threads on Squeak dev about memeory sizes.

In general, you should use the latest VM and I guess this question
is not a beginners thing. Maybe even try the VM developers list.

AC> Is this a bug, or are there Squeak image size limits and if so what are
AC> they and what generates them. Does anyone have any suggestions how to
AC> get around the problem?

Maybe you can save your neurons to disk via ReferenceStream and load
them after image startup. But that's not fast with hundreds of Megs.

Or try to stick them into several files, so the memory consumption
rises gradually.



Cheers,

Herbert  

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners