[gclist] Destructor FAQ

[Daniel C. Wang]

Mario.Wolczko@Eng.Sun.COM (Mario Wolczko) writes:
> Nobody has mentioned post-mortem finalization in any detail, so I'll
> throw in a paragraph or two...
> 
{stuff deleted}
> So, the key ideas of this scheme are that objects referenced only by
> weak pointers are reclaimed anyway, the weak pointers are nilled by
> the collector, and the mutator can find out which weak pointers have
> just been nilled, in order to perform finalization.
> 
> The only published description of this technique that I'm aware of is
> in the ParcPlace Smalltalk manuals.


I think there was some work done by the developers of Chez Scheme system
that had a similar flavor but with a higher level interface referred to as
"guardians." The mutator registered an object with a guardian, and could
query the guardian to return a list of objects that were "dead" and recycle
the resources. I remember coming across a paper about this a while back,
but can't jog my memory enough to remember the exact reference. I personally
thought it had a pretty good survey of various finalization techniques in
the intro.. Anyone else able to remember the paper I'm talking about?

+----------------------+
|     Daniel Wang      |
|                      |
| dw3u+@andrew.cmu.edu |
+----------------------+