Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!steinmetz!davidsen
From: davidsen@steinmetz.steinmetz.UUCP (William E. Davidsen Jr)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Arc vs compress
Message-ID: <7887@steinmetz.steinmetz.UUCP>
Date: Mon, 16-Nov-87 18:00:17 EST
Article-I.D.: steinmet.7887
Posted: Mon Nov 16 18:00:17 1987
Date-Received: Wed, 18-Nov-87 04:32:03 EST
References: <696@rlgvax.UUCP> <7406@eddie.MIT.EDU> <343@gethen.UUCP>
Reply-To: davidsen@crdos1.UUCP (bill davidsen)
Organization: General Electric CRD, Schenectady, NY
Lines: 40
Keywords: Why arc?

I suspect that the best compression available with file extraction still
allowed is from a combination of compress and cpio. Here is how it's
done. Note that this is available to any system which has cpio (or afio)
and compress.

  zoo:	zoo a testit *.c
  arc:	arc a testit *.c
  c+c:	ls *.c | cpio -oc | compress > testit.CC


Extract single file
  zoo:	zoo x testit file.c
  arc:	arc x testit *.c
  c+c:	zcat <testit.CC | cpio -icm file.c

List an archive:
  zoo:	zoo v testit (or l or lf)
  arc:	arc v testit (or l)
  c+c:	zcat <testit.cc | cpio -octv (or oct)

Performance:
  I tried this on a bunch of files which consisted of C source and
compiled files, I got these results (total input 34365 bytes):

	CPU (u+s)	file size
  arc:	80.34		25736
  zoo:	16.30		26282
  c+c:	 8.40		20511

The compression of the c+c method improves as the archive file gets
larger, while the others improve as the size of individual files gets
larger, until the 13 bit limit is reached.

I still use zoo for reasons I have mentioned before. It has a convenient
if baroque user interface, and is available for many systems including
VMS.  Although I have cpio and compress for MS-DOS, I prefer zoo.
-- 
	bill davidsen		(wedu@ge-crd.arpa)
  {uunet | philabs | seismo}!steinmetz!crdos1!davidsen
"Stupidity, like virtue, is its own reward" -me