kolibrios/programs/fs/unzip60/beos
..
beocfg.h
beos.c
beos.h
beosmain.cpp
Contents
Makefile
README
unzip.rsc
unzipsfx.rsc

UnZip 5.4 for BeOS

NOTE:

If you want to build UnZip 5.4 or later from the source, you'll need to
have the "xres" tool installed (unless you remove the "xres" lines in the
beos/Makefile).  This will cease to be a problem when BeOS R4 ships this
fall.  Until then, you can get xres from
ftp://ftp.be.com/pub/experimental/tools/xres-102.zip.

HISTORY

UnZip 5.30 was the first official release of Info-ZIP's UnZip to support
the filesystem in BeOS.

UnZip 5.31 added support for the new filesystem that appeared in the
Advanced Access Preview (aka DR9) Release of BeOS.

UnZip 5.32 added several important bug fixes.

UnZip 5.4:

- supports BeOS on x86 hardware (and cross-compiling, if a compiler is
  present)

- ask the Registrar to assign a file type to files that don't have one

- adds a new -J option on BeOS; this lets you extract the data for a file
  without restoring its file attributes (handy if you stumble on really
  old BeOS ZIP archives... from before BeOS Preview Release)

- will restore attributes properly on symbolic links (you'll need
  zip 2.21 or later to create ZIP files that store attributes for
  symbolic links)

*** WARNING ***
You may find some extremely old BeOS zip archives that store their
file attributes differently; these will be from DR8 and earlier (when
BeOS copied the MacOS type/creator fields instead of using the current
extremely flexible scheme).

You can still unpack the _data_ in older zip files, but you won't be
able to recover the file attributes in those archives.  Use the -J option
with these files or you'll get "compressed EA data missing" and "zipfile
probably corrupt" errors, even though the data is intact!

The new scheme makes handling BeOS file attributes much more robust, and
allows for possible future expansion without another round of
incompatibilities.

That's life on the edge!
*** WARNING ***

The new filesystem allows for huge files (up to several terabytes!) with
huge amounts of meta-data (up to several terabytes!).  The existing ZIP
format was designed when this much data on a personal computer was
science fiction; as a result, it's quite possible that large amounts of file
attributes (more than maybe 100+K bytes) could be truncated.  Zip and UnZip
try to deal with this in a fairly sensible way, working on the assumption
that the data in the file is more important than the data in the file
attributes.

One way to run into this problem is to mount an HFS volume and zip
some Mac files that have large resources attached to them.  This
happens more often than you'd expect; I've seen several 0-byte files that
had over four megabytes of resources.  Even more stupid, these resources
were _data_ (sound for a game), and could have been easily stored as
data...

KNOWN BUGS

None! Yahoo!

Please report any bugs to Zip-Bugs@lists.wku.edu.

- Chris Herborth (chrish@qnx.com)
  November 2/1998