Path: utzoo!news-server.csri.toronto.edu!rutgers!att!emory!samsung!uakari.primate.wisc.edu!sdd.hp.com!cs.utexas.edu!uunet!ccicpg!mendip!ndjc
From: ndjc@mendip.UUCP (Nick Crossley)
Newsgroups: comp.lang.perl
Subject: Re: find2perl
Summary: Please don't include -cpio and -tar
Message-ID: <15719@mendip.UUCP>
Date: 10 Mar 91 22:01:40 GMT
References: <11674@jpl-devvax.JPL.NASA.GOV> <40887@genrad.UUCP> <11709@jpl-devvax.JPL.NASA.GOV>
Organization: ICL North America, Irvine, CA 92718
Lines: 28

In article <11709@jpl-devvax.JPL.NASA.GOV> lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) writes:
>I've fixed that, and added -depth, -ls, -cpio, -ncpio and -tar.  So you
>can call this a beta.
>
>I didn't realize till now just how much random crap tar and cpio leave sitting
>around in their output files...

There is also the problem that tar and cpio are not as standard as
one would like.  There are a surprising number of differences in
the exact formats (mainly caused by design and implementation bugs
in cpio, various extensions of tar, POSIX changes, etc.).  V.4 has
a new cpio (in fact, a merge of cpio and tar, with support for most
of the old formats plus at least two new ones).

With all this confusion, I'm not sure it's a good idea to build in
yet another cpio/tar encoding.  It was a mistake for find to include
a copy of the cpio code in the first place, and in V.4 those options
are marked as obsolete and due to be withdrawn.

Of course, you could argue that the perl version is an attempt to
impose a portable cpio, but then the old cpio format with its header
fields too small for 32 bit numbers is a poor design to begin with,
and not something that one would want to perpetuate.
-- 

<<< standard disclaimers >>>
Nick Crossley, ICL NA, 9801 Muirlands, Irvine, CA 92718-2521, USA 714-458-7282
uunet!ccicpg!ndjc  /  ndjc@irv.icl.com