On Friday 15 September 2006 08:22, Kartik Vaddadi wrote:
Tom Shackell wrote:
I was trying to ask a slightly different question. Once Freja decides that it wants to collect trace information on a certain portion of the EDT, does the trace exist seperate from the heap memory used by the program, or is (part of the) heap itself used as the EDT like in my scheme? Would I be correct in saying that Freja (and Hat, Hood & Buddha) all duplicate information -- once in the normal heap of the program being debugged, and once again in the trace structures of the debugger?
Hat, Hood and Buddha all duplicate the information. Though 'duplicate' is somewhat inaccurate because of course all three allow garbage collection meaning the vast majority of the data in the heap is collected (and hence no longer duplicated). I can't remember whether Freja does or not, my suggestion is to read Nilsson's thesis if you haven't done so already: http://citeseer.ist.psu.edu/nilsson98declarative.html Cheers :-) Tom