preece%ccvaxa@GSWD-VMS.ARPA (Scott E. Preece) (10/06/86)
> From: Dan Hoey <hoey@nrl-aic.ARPA> > Let me first deplore the abuse of language by which it is claimed that > Xerox has a garbage collector at all. In the language of computer > science, Xerox reclaims storage using a ``reference counter'' > technique, rather than a ``garbage collector.'' This terminology > appears in Knuth's 1973 *Art of Computer Programming* and originated in > papers published in 1960. I remain undecided as to whether Xerox's > misuse of the term stems from an attempt at conciseness, ignorance of > standard terminology, or a conscious act of deceit. ---------- Hoey's pedantic insistence on a precision which does not exist in the "standard terminology" is apparently also an incorrect characterization of the Xerox approach, which [from the descriptions I have read] combines some aspects of the "pure" reference counting approach described by Knuth and some aspects of "pure" garbage collection. The Deutsch paper [CACM, 9/76] explicitly separates the two kinds of storage reclamation techniques and then proposes a combined method with features of both. In fact, however, the distinction on which Hoey places so much importance seems to have mostly vanished from the literature in the years since Knuth's description (why he places it in 1973 I don't know, my copy dates to 1968). Many more recent sources consider reference counting simply one form of garbage identification. The survey by Cohen (Computing Surveys, 9/81), for instance, discusses reference counting and marking as just two alternative ways of identifying garbage. Gabriel (Performance and Evaluation of Lisp Systems) says of the Xerox scheme, "Garbage collection is patterned after that described by [Deutsch, 1976]. A reference count is maintained..." Moon ("Garbage Collection in a Large Lisp System") discusses reference counting alternatives under the name garbage collection. Reference counting seems to have been accepted as a method of preforming one sub-task of garbage collection; Hoey's nit-picking is neither productive nor, since the Xerox approach is not pure reference counting, accurate. -- scott preece gould/csd - urbana uucp: ihnp4!uiucdcs!ccvaxa!preece arpa: preece@gswd-vms