Combined patch for Frotz 2.32
-----------------------------

This is a combined patch for Frotz 2.32, incorporating the two patches:

The z-strict patch written by Torbjörn Anderson, from
ftp://ftp.gmd.de/if-archive/interpreters-infocom-zcode/frotz/frotz_zstrict_patch

This permits the interpreter to detect, and optionally report, invalid
attempts to access object 0.  (This often happens with old versions of the
Inform library, and usually has no ill effects but occasionally causes games
to crash).  See comments in the patched version of frotz.h.  If you only
play games, you may wish to define STRICTZ_DEFAULT_REPORT_MODE in frotz.h to 
STRICTZ_REPORT_NEVER so that the interpreter won't issue annoying error 
messages when playing games (but will handle the error safely.)  If you write
games, it may be a good idea to increase STRICTZ_DEFAULT_REPORT_MODE to
STRICTZ_REPORT_ALWAYS for the purposes of testing.  

The QUETZAL patch written by Martin Frost, from
http://www.geocities.com/SiliconValley/Vista/6631/frotz-quetzal.tgz

This enables the use of the portable QUETZAL save file format, which is now
becoming the standard.  Note that if you compile with this option
you won't be able to read old non-QUETZAL save files.

This patch is for the Unix version but can probably be applied to any
version of Frotz 2.32.

Applying the patch under a Unix system
--------------------------------------

Solaris users are advised to install the GNU version of patch
(available from several sources e.g. sunsite.doc.ic.ac.uk).  (However, you may
get the Solaris version to work if you remove all the 'diff' statements from
the patch file.)

Install the quetzal.c, strictz.c and frotznew.patch files in the Frotz source 
directory, then type

patch -p1 < frotznew.patch

Check the output for any failed 'hunks' and apply those patches by hand.
For example, you may need to patch the Makefile by hand if you have made any
changes to it.