[plt-scheme] DrScheme3m Under V350.3

Thanks very much for the example! I fixed a bug in GCing near the
boundary of shared stack segments in nested continuations, and now your
example program consistently runs for me.

(I especially like that your program prints "Problem solved!" at the
end --- very nice to see after chasing down such a subtle bug. I hope
it's accurate. :)

Matthew

At Mon, 3 Jul 2006 17:44:19 -0700 , "Williams, M. Douglas" wrote:
> I have tested my inference engine code on both Windows XP and Linux and on
> both systems it crashes DrScheme3m under V350.3.  It runs fine under V350
> (both DrScheme and DrScheme3m) and also runs fine under DrScheme for V350.3.
> Under Windows XP, it generally dies with the GC indicator icon on and gives
> the following generic error.
> 
>  
> 
> MrEd-based application has encountered a problem and needs to close.  We are
> sorry for the inconvenience.
> 
>  
> 
> For Linux (Ubuntu 6.06) the window and application just go away.  It is
> almost immediate on starting the computation and the crash is 100%
> repeatable.  I'm not sure what other information I can give you.  I can, of
> course, send the code to anyone who wants to investigate the problem.
> 
>  
> 
> The code is strictly Scheme - there are no FFI packages and it is entirely
> self contained (there are no requires at all).  The only unusual thing I can
> think of is that it does do quite a bit of destructive list manipulations.
> It is also prototype code and likely leaks some, but I can't see that that
> is the problem since it does run outside of the V350.3 3m environment.
> 
>  
> 
> [Note that in earlier posts today I was saying in at least some places 305.3
> when I meant 350.3.]
> 
>  
> 
> Doug
> 
>  
> 
> M. Douglas Williams, Ph.D.
> 
> Sr. Scientist
> 
> Science Applications International Corporation
> 
>  
> 
> _________________________________________________
>   For list-related administrative tasks:
>   http://list.cs.brown.edu/mailman/listinfo/plt-scheme