[MLton-devel] Minimizing space usage in ML (fwd)

[Stephen Weeks]

> My  take  is  that  unless  you  have a reference counted GC, or at least the
> reference counts, it isn't going to do much because you won't know which  way
> to  change the reference.  I.e., if I have lots of pointers pointing to equal
> things, then I would think that the normal behavior would be to just  shuffle
> things  around  without  ever  causing any thing to be dead. 

You're right -- it's certainly a *lot* worse without reference counts.
But I don't agree that we'll just shuffle stuff around, especially if
there are lots of reference counts of one.  The tendency will be for
objects with fewer pointers to die, since they have a higher
probability of their count going to zero via random shuffles.  But I
don't think it's worth worrying about.

Joe, since MLton doesn't do reference counting this optimization
probably isn't a good idea.

> If  MLton  had something in the heap indicating immutability (which could, of
> course be generated from the stack and some code  tables  I  think)  then  we
> could  do  the  sharing hack I described that I did long ago for Mathematica.

Yes, I would like to see that someday.  I once again ask for
volunteers!


-------------------------------------------------------
This sf.net emial is sponsored by: Influence the future of 
Java(TM) technology. Join the Java Community Process(SM) (JCP(SM)) 
program now. http://ad.doubleclick.net/clk;4699841;7576301;v?
http://www.sun.com/javavote
_______________________________________________
MLton-devel mailing list
MLton-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mlton-devel