Re: robustness vs. conservative GC

[Tom Lord <lord at regexps dot com>]

       You're talking about a denial of service attack on say, a Java
       implementation, by causing it to retain extra memory?  I don't
       think that's fundamentally different from retaining lots of
       memory with real pointers.

I'm not convinced that a detailed discussion of the possible exploits
that are uniquely attributable to a conservative GC is a wise topic
for disucssion on this list, though I'm glad to see the threat
mentioned more directly than I had put it.  

Also: my primary reason for debunking conservative GC myths is to
argue for support for my run-time system project.  In this context,
that's directed mostly at those interested in GCJ.  I'm not sure that
a detailed discussion of exploits is really necessary for that
argument.

Regards,
-t