fa6098bfd4
[Long writeup committed with flag.h and options.c only.]
This is a reworking of a user contributed patch available in Pasi
Kallinen's NetHack Patch Database at http://bilious.homelinux.org under
the name "Paranoid_Quit". It was created by David Damerell and extended
by several others, and I haven't attempted to preserve attribution.
Their patch added three new boolean run-time options (this one
doesn't; details below):
paranoid_quit: if true, change the "really quit?" prompt to require an
answer of yes rather than just y to quit, presumeably for players who
type faster than they read and think (been there, done that...). It
also applies to the "do you want to enter explore mode?" prompt. The
changed prompt shows yes and no rather than yn as possible answers.
After having used it a few times, I find it easily noticeable that
"yes"<return> is expected instead of just single keystroke 'y', and
if you mess up somehow you can just reissue the #quit or X command
with no harm done. (And the default setting is off; the game still
issues the original yn prompt.)
paranoid_hit: if true, make a similar change for the "really attack <the
peaceful monster>?" prompt. Definitely helpful if you bump into a
monster while in the midst of using 'y' to move diagonally upper left.
Note that this just changes the expected/required answer to an existing
prompt; it doesn't change interaction between the hero and monsters.
paranoid_remove: if true, the 'R' and 'T' commands will prompt the player
to select an appropriate item from inventory even when there's only one
applicable item (instead of simply removing or taking off that only
item). Helpful if you think you've got more than one thing on and
intend to take off something other than the last one (which might be a
ring of levitation keeping you from dropping into lava or a blindfold
and you're trying to play the whole game blinded).
Their patch also made two other changes which weren't controllable via
options: when dipping, after picking what to dip, mention it in the
second prompt for what to dip into; and require yes instead of y at the
wizard mode "save bones?" prompt. We've had the enhanced dipping prompt
for a while, and "unknown" installed a fix-up (which wasn't needed with
their version) for it recently. I've left our bones prompt alone, the
original yn query. Anyone who saves bones by accident can remove them,
if not externally they by using wizard mode to revisit the same dungeon
depth, load the bones, and unlink them.
#####
That's a summary of the contributed patch. Now for the implemented
one.... Instead of separate booleans, this adds a single compound option
called "paranoid_confirmation" that takes a string argument of space
separated words: "quit die attack pray Remove". And it puts the actual
yes vs y querying into a new routine instead of duplicating it at each
affected prompting location.
paranoid_confirm:quit - as above, if true then require yes instead of y
to answer the "really quit?" and "do you want to enter explore mode?"
prompts. Can also be supplied as paranoid_confirm:explore or even
"quit explore" but that's just redundant; it's a single flag which
controls prompting for both game-ending or game-altering commands.
paranoid_confirm:die - applicable only for explore and wizard modes but
visible/settable during option viewing/changing in normal play. If
true then require yes instead of y at the "die?" prompt. This wasn't
part of their original patch, but should have been since the effect
of accidental y is just as drastic as unintentionally quitting.
paranoid_confirm:attack - as above, yes vs y for "really attack <the
peaceful monster?". Can also be supplied as paranoid_confirm:hit.
paranoid_confirm:pray - supersedes the existing prayconfirm boolean.
That option is still accepted and honored duing config file processing,
but option viewing/changing with 'O' only handles the new variant.
This does not control "yes" vs 'y', but rather whether there's a prompt
first or prayer simply starts. When used, the prompt itself is the
same yn one already being asked with prayconfirm. Presumably config
file support for prayconfirm will go away in some future version.
Unlike the other paranoid settings, this one defaults to 'on' in order
to match the 3.3.0 through 3.4.3 behavior controlled by prayconfirm,
whose default was on (but maybe should have been off...).
paranoid_confirm:Remove - as above, causes the 'R' and 'T' commands to
use a "what do you want to remove?" or "what do you want to takeoff?"
inventory selection prompt even when only one accessory or piece of
armor is worn. Player can pick the inventory letter and remove/takeoff
the item, use ? or * to see what the candidate item is, or cancel with
ESC. Can be supplied as paranoid_confirm:takeoff or "remove takeoff",
but like with "quit explore", a single flag controls the behavior of
both 'R' and 'T'.
Option file processing accepts two other settings, paranoid_confirm:none
and paranoid_confirm:all, but those are not available (nor needed) when
using the 'O' command. "none" is useful because it's the value shown to
the player by 'O' when none of the paranoia flags are set, and it's a
way to turn off paranoid_confirm:pray without turning on any of the other
choices. "all" probably isn't very useful but was trivial to tack on.
This is an example of the menu that 'O' puts up after picking option
paranoid_confirmation from the main list. I've shifted everything left
to reduce whitespace here; it appears on the right side of the screen for
tty menuing.
Actions requiring extra confirmation:
q - yes vs y to quit or to enter explore mode
d - yes vs y to die (explore mode or debug mode)
a - yes vs y to attack a peaceful monster
p + y to pray (supersedes old "prayconfirm" option)
R - always pick from inventory for Remove and Takeoff
(end)
Currently set paranoia features are marked as preselected and can be
toggled off along with toggling any others on as desired. I've just
realized that this menu relies on showing entries marked via preselection
rather than explicitly annotating each one as [on] or [off]; that seemed
perfectly natural during testing so I think I'll leave it this way, at
least for the time being.
NetHack 3.5.0 -- General information
NetHack 3.5 is an enhancement to the dungeon exploration game NetHack.
It is a distant descendent of Rogue and Hack, and a direct descendent of
NetHack 3.4.
NetHack 3.5.0 has many new features.
* List new features here
A fuller list of changes for this release can be found in the file
doc/fixes35.0 in the source distribution. The text in there was written
for the development team's own use and is provided "as is", so please do
not ask us to further explain the entries in that file.
- - - - - - - - - - -
Please read items (1), (2) and (3) BEFORE doing anything with your new code.
1. Unpack the code in a dedicated new directory. We will refer to that
directory as the 'Top' directory. It makes no difference what you
call it.
2. If there is no flaw in the packaging, many sub-directories will be
automatically created, and files will be deposited in them:
a. A 'dat' directory, which contains a variety of data files.
b. A 'doc' directory, which contains various documentation.
c. An 'include' directory, which contains *.h files.
d. A 'src' directory, which contains game *.c files used by all versions.
e. A 'util' directory, which contains files for utility programs.
f. A 'sys' directory, which contains subdirectories for files that
are operating-system specific.
g. A 'sys/share' subdirectory, which contains files shared by some OSs.
h. A 'sys/share/sounds' subsubdirectory, which contains sound files
shared by some OSs.
i. A 'sys/amiga' subdirectory, which contains files specific to AmigaDOS.
j. A 'sys/atari' subdirectory, which contains files specific to TOS.
k. A 'sys/be' subdirectory, which contains files specific to Be OS.
l. A 'sys/mac' subdirectory, which contains files specific to MacOS.
m. A 'sys/msdos' subdirectory, which contains files specific to MS-DOS.
n. A 'sys/os2' subdirectory, which contains files specific to OS/2.
o. A 'sys/unix' subdirectory, which contains files specific to UNIX.
p. A 'sys/vms' subdirectory, which contains files specific to VMS.
q. A 'sys/wince' subdirectory, which contains files specific to Windows CE.
r. A 'sys/wince/ceinc' subdirectory; header files for Windows CE
s. A 'sys/wince/ceinc/sys' subdirectory; ditto
t. A 'sys/winnt' subdirectory, which contains files specific to Windows NT.
u. A 'win' directory, which contains subdirectories for files that
are windowing-system specific (but not operating-system specific).
v. A 'win/share' subdirectory, which contains files shared by some
windowing systems.
w. A 'win/Qt' subdirectory, which contains files specific to Qt.
x. A 'win/X11' subdirectory, which contains files specific to X11.
y. A 'win/gem' subdirectory, which contains files specific to GEM.
z. A 'win/gnome' subdirectory, which contains files specific to GNOME.
A. A 'win/tty' subdirectory, which contains files specific to ttys.
B. A 'win/win32' subdirectory, which contains files specific to the
Windows Win32 API.
The names of these directories should not be changed unless you are
ready to go through the makefiles and the makedefs program and change
all the directory references in them.
3. Having unpacked, you should have a file called 'Files' in your Top
directory. This file contains the list of all the files you now SHOULD
have in each directory. Please check the files in each directory
against this list to make sure that you have a complete set.
4. Before you do anything else, please read carefully the file called
"license" in the 'dat' subdirectory. It is expected that you comply
with the terms of that license, and we are very serious about it.
5. If everything is in order, you can now turn to trying to get the program
to compile and run on your particular system. It is worth mentioning
that the default configuration is SysV/Sun/Solaris2.x (simply because
the code was housed on such a system). It is also worth mentioning
here that NetHack 3.5 is a huge program. If you intend to run it on a
small machine, you'll have to make hard choices among the options
available in config.h.
The files sys/*/Install.* were written to guide you in configuring the
program for your operating system. The files win/*/Install.* are
available, where necessary, to help you in configuring the program
for particular windowing environments. Reading them, and the man pages,
should answer most of your questions.
At the time of this release, NetHack 3.5 is known to run/compile on:
Apple Macintosh running MacOS 7.5 or higher, LinuxPPC, BeOS 4.0
Atari ST/TT/Falcon running TOS (or MultiTOS) with GCC
Commodore Amiga running AmigaDOS 3.0 or higher with SAS/C 6.x
(but see Makefile.ami about DICE and Manx)
DEC Alpha/VMS (aka OpenVMS AXP), running V1.x through V7.1
DEC VAX/VMS, running V4.6 through V7.1
HP 9000s700 running HP-UX 10.x, 11.x
IBM PS/2 and AT compatibles running OS/2 - 2.0 and up with GCC emx
Intel 80386 or greater (or clone) boxes running MS-DOS with DPMI.
Intel 80386 or greater (or clone) boxes running Linux, BSDI, or
Windows 95/98/NT/2000/XP
Intel Pentium or better (or clone) running BeOS 4.5
Sun SPARC based machine running SunOS 4.x, Solaris 2.x, or Solaris 7
NetHack 3.5 will also run on the following, but a cross-compiler hosted
on another platform, such as win32, is required to build from source.
Pocket PC devices running Windows CE 3.0 and higher
H/PC Pro devices running Windows CE 2.11 and higher.
Palm Size PC 1.1 devices running Windows CE 2.11
Previous versions of NetHack were tested on the following systems,
and we expect that NetHack 3.5 will work on them as well:
AT&T 3B1 running System V (3.51)
AT&T 3B2/600 & 3B2/622 running System V R3.2.1
AT&T 3B2/1000 Model 80 running System V R3.2.2
AT&T 3B4000 running System V
AT&T 6386 running System V R3.2
Data General AViiON systems running DG/UX
DEC vaxen running BSD, Ultrix
Decstations running Ultrix 3.1, 4.x
Encore Multimax running UMAX 4.2
Gould NP1 running UTX 3/2
HP 9000s300 running HP-UX
HP 9000s700 running HP-UX 9.x
IBM PC/RT and RS/6000 running AIX 3.x
IBM PS/2 and AT compatibles running OS/2 1.1 - 2.0 (and probably
Warp) with Microsoft 6.0, and OS/2 2.0 and up with IBM CSet++ 2.0.
Intel 80386 or greater (or clone) running 386BSD
Mips M2000 running RiscOS 4.1
NeXT running Mach (using BSD configuration)
Pyramid 9820x running OSx 4.4c
SGI Iris running IRIX
Stardent Vistra 800 running SysV R4.0
Stride 460 running UniStride 2.1
Sun-3s, -4s, and -386is running SunOS 3.x
Sun-3s and -386is running SunOS 4.x
Valid Logic Systems SCALD-System
Unless otherwise mentioned, the compiler used was the OS-vendor's
C compiler.
With the demise of Windows NT on the DEC Alpha, no attempt has been
made to build NetHack 3.5 on that platform.
No attempt has been made to build or run NetHack 3.5 on Windows Me
at this point. It may work, but then again it may not.
A build for Intel 80286 machines and DOS "real mode" overlaid versions
has not been produced for NetHack 3.5. Nobody on the porting team has
the time or the software to attempt the necessary tuning that will allow
it to achieve the balance of having just the right amount of available
memory, and still have acceptable performance. The sources necessary
to do so are still included in the source distribution, so if someone
has access to a real-mode compiler and lots of spare time on their hands,
you may be able to get things working. Of course you do so at your own risk.
- - - - - - - - - - -
If you have problems building the game, or you find bugs in it, we recommend
filing a bug report from our "Contact Us" web page at:
http://www.nethack.org/
When sending correspondence, please observe the following:
o Please be sure to include your machine type, OS, and patchlevel.
o Never send us binary files (e.g. save files or bones files). Whichever
platform you are using, only a small minority of the development team has
access to it, and you will rapidly annoy the others. If you have found
a bug and think that your save file would aid in solving the problem,
send us a description in words of the problem, your machine type, your
operating system, and the version of NetHack. Tell us that you have a
save file, but do not actually send it.
In the rare case that we think your save file would be helpful, you will
be contacted by a member of the development team with the address of a
specific person to send the save file to.
o Though we make an effort to reply to each bug report, it may take some
time before you receive feedback. This is especially true during the
period immediately after a new release, when we get the most bug reports.
o We don't give hints for playing the game.
o Don't bother to ask when the next version will be out. You will not get
a reply.
If you don't have access to the world wide web, or if you want to submit
a patch for the NetHack source code via email directly, you can direct it
to this address:
nethack-bugs (at) nethack.org
If you've changed something to get NetHack to run on your system, it's likely
that others have done it by making slightly different modifications. By routing
your patches through the development team, we should be able to avoid making
everyone else choose among variant patches claiming to do the same thing, to keep
most of the copies of 3.5 synchronized by means of official patches, and to
maintain the painfully-created file organization. (This process has been working
since the time when everyone just posted their own patches to 2.3. At that time,
there were no archived bug-fixes to give to people who got 2.3 after its initial
release, so the same bugs kept being discovered by new batches of people.)
We have been successful in preventing this from happening since the 3.0
release. Please cooperate to keep this from happening to 3.5.
It is inevitable that we will reject some proposed additions of new features
either because they do not fit our conception of the game, or because they
require more code than we consider they're worth. If we reject your feature,
you are free, of course, to post the patches to the net yourself and let the
marketplace decide their worth.
All of this amounts to the following: If you decide to apply a free-lanced
patch to your 3.5 code, you are on your own. In our own patches, we will
assume that your code is synchronized with ours.
-- Good luck, and happy Hacking --