Allow a way to configure NetHack to run entirely from a USB stick
or other removable device in a way that allows everything to
reside entirely on the USB stick, and nothing on the computer's
hard drive. That could be done in versions prior to 3.6.3.
Sample:
i: is a USB stick
i:\nhdist contains the NetHack Windows distribution and a sysconf
file dropped into that distribution with the following entry in it:
portable_device_top = nethack
No device is included in the portable_device_top entry, the device
is always the device that the nethack exe resides on. If you try
to specify a device in the portable_device_top path, the device
portion will be ignored.
portable_device_top specifies the folder on the device that is writable
by NetHack and as such it cannot be the same folder that the executable
resides in.
i:\nhdist\nethack --showpaths
Variable playground locations:
[hackdir ]="i:\nethack\"
[leveldir ]="i:\nethack\"
[savedir ]="i:\nethack\"
[bonesdir ]="i:\nethack\"
[datadir ]="i:\nhdist\"
[scoredir ]="i:\nethack\"
[lockdir ]="i:\nethack\"
[sysconfdir]="i:\nhdist\"
[configdir ]="i:\nethack\"
[troubledir]="i:\nethack\"
NetHack's system configuration file (in sysconfdir):
"i:\nhdist\sysconf"
The loadable symbols file (in sysconfdir):
"i:\nhdist\symbols"
Basic data files (in datadir) are collected inside:
"i:\nhdist\nhdat363"
No end-of-game disclosure file (disabled).
Writable folder for portable device config (sysconf portable_device_top):
"i:\nethack\"
Your personal configuration file (in configdir):
"i:\nethack\.nethackrc"
Without that sysconf file in the NetHack distribution folder on the
USB stick with the 'portable_device_top = ' entry, the paths
return to the default locations for 3.6.3 on Windows:
i:\nhdist\nethack --showpaths
Variable playground locations:
[hackdir ]="C:\Users\JaneDoe\NetHack\3.6\"
[leveldir ]="C:\Users\JaneDoe\AppData\Local\NetHack\3.6\"
[savedir ]="C:\Users\JaneDoe\AppData\Local\NetHack\3.6\"
[bonesdir ]="C:\ProgramData\NetHack\3.6\"
[datadir ]="i:\nhdist\"
[scoredir ]="C:\ProgramData\NetHack\3.6\"
[lockdir ]="C:\ProgramData\NetHack\3.6\"
[sysconfdir]="C:\ProgramData\NetHack\3.6\"
[configdir ]="C:\Users\JaneDoe\NetHack\"
[troubledir]="C:\Users\JaneDoe\NetHack\3.6\"
NetHack's system configuration file (in sysconfdir):
"C:\ProgramData\NetHack\3.6\sysconf"
The loadable symbols file (in sysconfdir):
"C:\ProgramData\NetHack\3.6\symbols"
Basic data files (in datadir) are collected inside:
"i:\nhdist\nhdat363"
No end-of-game disclosure file (disabled).
Your personal configuration file (in configdir):
"C:\Users\JaneDoe\NetHack\.nethackrc"
Some platforms compile ncurses in such a way that the tinfo library needs to
be separately linked in (e.g. Gentoo).
The tinfo library is a low-level library being used by and bundled with
ncurses.
Adding '-ltinfo' to WINTTYLIB in the hint files was considered the most robust
solution to this problem. This still works on systems with ncurses 5.9.
Using pkg-config was ruled out as the default build of ncurses doesn't install
the necessary files.
Cherry-pick 3.7.0's 565e020573. Again,
conflict because the fixes entry goes into a different file.
When picking up from floor or removing from container fails because
there aren't any inventory slots available, pickup/take-out stops.
But the message
|Your knapsack can't accomodate any more items.
is inaccurate if there is gold beyond the stopping point. Actually
continuing in order to pickup/take-out gold would require substantial
changes, but varying the message to be
|Your knapsack can't accomodate any more items (except gold).
when stopping is a one line fix. The parenthesized remark is only
added if there is actually some gold after the current object and is
given regardless of whether autopickup happens to be targetting it.
Fixes#246
Cherry-pick 3.7.0's 4a3d5f95d9
(github pull request #252). Slightly tricky because the fix entry
nowgoes into a different file. fixes37.0 will need fixing up.
When you ride your steed into a polymorph trap and it changes into a
creature that can still wear the saddle, the message is
|You have to adjust youself in the saddle on <foo>.
which sounds as if the game is telling the player that he or she needs
to do something. Simplify it to
|You adjust yourself in the saddle on <foo>.
Submitted for 3.7.0; all but one also apply to 3.6.3.
I rewrote the curses terminal-too-small message instead of just
fixing the spelling of "minumum".
Changing from BETA to RELEASED resulted in turning off PANICTRACE
and that exposed a minor memory leak. Only applies to program exit
so doesn't impact play.
the prompting on Windows wasn't working correctly if a prior game had crashed
and the self-recover feature was trying to kick in. This impacts tty, curses,
and mswin (GUI).
Taking off no-delay helmets, gloves, and boots were unintentionally
taking off suit instead and stayed worn themselves. As far as I
saw, only helmet types "fedora" and "dented pot" were applicable;
all gloves and boots have a small multi-turn delay. This was an
unintended side-effect of the first "slippery gloves" commit so
happened about three weeks ago.
Performance profiling showed that multiple strcmpi() calls were
occurring each and every time a character was going to the map.
This update:
- honors the WC_COLOR capability
- It allows a window-port to control individual color availability should the window-port wish to do so.
- Makes checking on the individual colors for the active window-port is a straightforward table lookup at the CLR_ offset.
iflags.use_color remains a master on/off switch for use of color, regardless of the capability
compiled into the game (default TRUE).
The has_color() routine, which is now a shared routine in src/windows.c, could likely be made
into a simple macro to eliminate the function call, but this update does not go that far.
This hits a lot of port files due to the window-port interface change, mostly cookie-cutter.
do.c:1005:54: warning: address of function 'uescaped_shaft' will always
evaluate to 'true' [-Wpointer-bool-conversion]
if (trap && (uteetering_at_seen_pit(trap) || uescaped_shaft)) {
~~ ^~~~~~~~~~~~~~
Add 'eating' (synonym 'continue') to the list of things that can be
set via paranoid_confirmation to require "yes" instead of "y" when
the user is prompted about something, in this case "Continue eating?".
dat/opthelp was missing a few of the paranoid_confirmation choices.
