Path: utzoo!yunexus!geac!syntron!jtsv16!uunet!lll-winken!lll-tis!ames!pasteur!ucbvax!lti.com!reg
From: reg@lti.com (Rick Genter x18)
Newsgroups: comp.lang.c
Subject: Re:  info-c digest v2n230
Message-ID: <8810262138.AA06757@lti.com>
Date: 26 Oct 88 21:38:28 GMT
Article-I.D.: lti.8810262138.AA06757
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 14

> And don't forget that at some level the record-access has to be doing
> byte-stream I/O *anyway*. The only way it can win is by buffering. And wouldn't
> you rather tune your own buffer sizes?

This is not correct.  Record-access in the IBM mainframe world is also
referred to or built on top of an `access method'.  These provide direct
access I/O to the disk at the record level, and the method of address
is of varying complexity.   There may be hash tables, B-trees, etc.,
involved that most certainly do *N*O*T* do byte-stream I/O to the disk.

					- reg
--
Rick Genter					...!{buita,bbn}!lti!reg
Language Technology, Inc.			reg%lti.uucp@bu-it.bu.edu
27 Congress St., Salem, MA 01970		(508) 741-1507