Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!emory!athena.cs.uga.edu!is
From: is@athena.cs.uga.edu ( Bob Stearns)
Newsgroups: comp.sys.ibm.pc.misc
Subject: Re: Norton QUerase and .ZIP files
Message-ID: <1990Dec6.145826.3800@athena.cs.uga.edu>
Date: 6 Dec 90 14:58:26 GMT
References: <4512@idunno.Princeton.EDU> <1990Dec6.120904.1214@urz.unibas.ch>
Organization: University of Georgia, Athens
Lines: 15

Your problem is the way that QU selects clusters to add to the unerased file.
As I understand it (and I have NOT gone deeply into to the code to verify
this) the clusters are chosen consecutively from the free list. If your disk
is badly fragmented, the original situation might have been:
     ZZZxZZZyZZZaZZZb
before the erase where ZZZZZZZZZZZZ is the large zip and x, y, a, b are small
files contributed to the zip. After the zip is recreated by QU, the FAT will
show the pattern
     ZZZZZZZZZZZZ****
where ZZZZZZZZZZZZ is the zip file and **** is unused space. Obviously not what
you wanted.

The correct solution to your problem is to use PKZIP more effectively:
    PKZIP -m all *.*
this will do what you want all in one command.