[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: GC order
- From: Basile STARYNKEVITCH <basile@...>
- Date: Wed, 7 May 2003 21:16:08 +0200
>>>>> "Bilyk," == Bilyk, Alex <ABilyk@maxis.com> writes:
Bilyk,> Lua5 manual states that user data objects are collected in
Bilyk,> the order reverse to that of their creation. It, however,
Bilyk,> doesn't say anything about the order of GC in
Bilyk,> general. Hence, given two arbitrary Lua objects O1 and O2,
Bilyk,> can assumption be made that O1 is destroyed after O2 if it
Bilyk,> was created before O2 and vice versa?
Even if the current implementation enforces this, I think that you
should not expect this in all cases. Imagine another Lua
implementation (and such beast exist, e.g. the lua interpeter coded
in Ocaml inside QuickC-- see www.cminusminus.org for more). Then it
probably won't have the same destruction order than the original Lua.
In GC-ed implementations, destructions can occur at any time (or even
not at all: consider an execution so quick that no GC is needed at
all!)
--
Basile STARYNKEVITCH http://starynkevitch.net/Basile/
email: basile<at>starynkevitch<dot>net
aliases: basile<at>tunes<dot>org = bstarynk<at>nerim<dot>net
8, rue de la Faïencerie, 92340 Bourg La Reine, France