Article View: comp.lang.clos
Article #337How to neatly destroy instances?
From: Jon L White
Date: Wed, 14 Oct 1992 20:09
Date: Wed, 14 Oct 1992 20:09
24 lines
1502 bytes
1502 bytes
re: So I still claim having an explicit free (or at least
some amount of low-level control on memory allocation, WHEN NEEDED AND
APPROPRIATE) would be 'a good thing'.
Quite true. For just this reason, many vendors -- Symbolics and Lucid
included -- supply an explicit RESOURCE management facility (but this
isn't part of the ANSI proposal.)
Ususally, however, the issue for explicit RESOURCE's isn't efficiency of
GC or whatnot, but rather the pressing need to avoid invoking a potentially
heavyweight(**) GC at an inopportune time -- "Cons-free coding".
I think someone else earlier brought this up too.
-- JonL --
(**) Statistically, speaking, the "ephemeral" scavenges will do the work
"most" of the time; but ocasionally they may trigger an ordinary
dynamic GC, which is not constant-bounded the way an "ephemeral"
scavenge is; the time required for this kind of regular GC typically
depends linearly upon the total size of the live data.
Technical Details
Message-ID:
<9210142009.AA10239@kuwait.lucid>Path:
arc1-ca.pugleaf.net!sparky!uunet!ukma!darwin.sura.net!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!lucid.com!jonl%kuwaitReferences:
jonl%kuwait@lucid.com (Jon L White)Show Full Headers
Complete Article Headers
Newsgroups: comp.lang.clos Path: arc1-ca.pugleaf.net!sparky!uunet!ukma!darwin.sura.net!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!lucid.com!jonl%kuwait From: jonl%kuwait@lucid.com (Jon L White) Subject: How to neatly destroy instances? Message-ID: <9210142009.AA10239@kuwait.lucid> Sender: daemon@cis.ohio-state.edu Organization: The Ohio State University Department of Computer and Information Science References: jonl%kuwait@lucid.com (Jon L White) Date: Wed, 14 Oct 1992 20:09:44 GMT Lines: 25