Issue reported by AndrioCelos: the corpse of an early-level fake
explorer supposedly killed by a trap would leave land mines intact.
Change land mines with a fake explorer's corpse into discovered pits.
Fixes#1449
Isssue reported by Tomsod: hero-as-target section of mhitm_ad_phys()
was not handling hero's Half_physical_damage attribute.
The issue was about cloning a pet from hero who is poly'd into a
pudding but it was more general than that. Half_physical_damage was
ignored for any hit by a monster-wielded weapon against poly'd hero.
It took a while to convince myself that Half_physical_damage wasn't
aleady being applied elsewhere but it doesn't seem to be.
Fixes#1447
Substitute $HOME/File if command line specifies --nethackrc=~/File
to avoid "Access to ~/File denied (2)".
Only implemented for opening run-time config file on Unix. Works for
NETHACKOPTIONS=@~/File too; the normally optional at-sign is required
since the tilde won't match a slash to distinguish file versus options.
Only supports "~/" file path prefix, not "~user/".
If using hints file sys/unix/hints/linux.370 or sys/unix/hints/macOS.370
allow the majority of the boilerplate compile switches to reside in a
compiler response file, instead of on the command line.
Include one of the following on your make command line:
make response=1
or
make resp=1
It can be combined with other make command line options. See
sys/unix/README-hints for further information about those.
The response files that it uses are:
CC (clang or gcc) src/nethack_cc.rsp
CXx (clang++ or g++) src/nethack_cxx.rsp
Note: I think the reduced clutter should actually become the default,
and the override should be noresponse=1 to NOT use it, but I'm
not sure how others feel, so for now, it requires
make resp=1
Feedback on whether that should become the default or not
is welcome.
Tested on Linux with gcc-15 and on Linux with clang-20.
I haven't had a chance to test it on macOS yet.
> Perhaps related: when a wand of sleep hits a disguised already-sleeping
> mimic (about which it is a separate question if they should go into
> disguise when sleeping, is it supposed to be automagic or conscious
> effort for them? but I digress), the mimic is not revealed (should it?)
> but the message says "hits a mimic".
Adjust restrap() so that a revealed mimic won't disguise itself while
sleeping. This seems to be in keeping with mimic lore.
Also, normal shop sounds (chime of register etc.) will wake a
mimic up from indeterminate sleep.
Closes#1441
Address the second part of the GitHub issue:
"Finally, when a wand of striking (or a shift+F melee attack for that
matter, but the latter gives no tactical advantage compared to just
bumping them) misses a disguised mimic, it is revealed. I don't know
what to make of it"
Issue reported by NullCGT: if the spot in front of a drawbridge held
water and got frozen, sanity checking for ice-melt timer would issue
complaints about melt timer for non-ice whenever the bridge was open.
Ice in front of closed bridge was handled ok, but ice beneath an open
bridge issued a spurious warning each turn if the sanity_check option
(wizard mode-only) was on.
Fixes#1440
Pull request from Feyorsh: in the sample run-time configuration
ile, doc/config.nh, the regular expression for "cursed" has bare
arentheses that should be quoted.
Fixes#1439
Part of the reasoning behind thrown objects not hitting mimics is
that a character who doesn't know there's a mimic there wouldn't
aim at it. But if you know there's a monster there (e.g. via
telepathy or monster detection), you would aim at it. Pushing a
boulder at it and hearing a monster works too (which is important
in cases where a mimic is trapped behind a boulder in the Sokoban
corridors).
Issue reported by vultur-cadens: 3.7's revised handling for initial
characteristic allocation included an unintended change from 3.6's.
I don't pretend to understand how characteristic allocation really
works. This should restore handling for values which are too low.
Fixes#1431
Issue reported by janne-hmp: examining an object on the map while
halluicinating might operate on an object whose name is Null since the
random object could be one that holds an extra description for item
shuffling at game start. Attempting to format the object led to a
crash.
I wasn't able to reproduce the crash, possibly because MacOS produces
the string "(null)" for sprintf("%s",NULL) instead of dereferencing
the Null pointer.
Perhaps random object selection for display should reject the extra
description objects in some classes. This susbstitutes a different
object if examining the map encounters one of those.
Fixes#1436
Data typo affecting '?' command's "longer explanation of game options".
Change the indication of menu_objsym's default value from [5] to [4]
since 4 is the actual default value in the code and the Guidebook.
Pull request from Umbire: starting gear for angelic beings should
use recently added silver mace instead of old ordinary mace.
It's simpler to just type in the change than to merge the commit.
Fixes#1433