This is enough to prevent abuse by denying access to functions and
denial of service (RAM and instruction step limits), but not enough
to allow restricted use of things that require finer control (e.g.
filesystem access).
If something goes wrong, the whole thing can be turned off, for
now, in config.h (see NHL_SANDBOX).
None of the current functionality requires changes to build systems;
some of the possible future functionality may require some #defines
- TBD. There is lots of dead code (#ifdef notyet) for bits of that
additional functionality; we can rip it out if we don't want those
additions or we can complete (parts of) it depending on our needs.
All current uses of Lua are connected to sandboxes and guarded with
nhl_pcall (sandbox and lua_pcall wrapper); options and limits can
be set at the callsites in the passed nhl_sandbox_info. Some of
the error handling may be wrong - panic() vs. impossible() vs
silence.
Memory and instruction step limits should be tuned prior to release;
there's no point tuning them now.
The air bubbles on the Plane of Water and the clouds on the Plane of
Air were being saved and restored as part of the current level's state
(which is the 'u' struct and invent and such) rather than with the
current level itself. That was ok for normal play, but for wizard
mode's ^V allowing you to return to a previously visited endgame level
after moving to a different one it meant a new set of bubbles for
Water and new set of clouds for Air. Even that was ok since it only
applied to wizard mode, but using #wizmakemap to recreate Water or Air
while you were on it added a new set of bubbles or clouds to the
existing ones. If repeated, eventually there wouldn't be much water
or air left.
Instead of just adding a hack to #wizmakemap, change save/restore to
keep the bubbles/clouds with the level rather than with the state.
That wasn't trivial and now I know why the old odd arrangement was
chosen. Saving hides u.uz by zeroing it out for levels that the hero
isn't on and it is zero during restore so simple checks for whether a
given level is water or air won't work.
This also adds another non-file/non-debugpline() use of DEBUGFILES:
DEBUGFILES=seethru nethack -D
will make water and clouds be transparent instead of opaque. It also
makes fumaroles and other light-blocking gas clouds be transparent
which wasn't really intended, but avoiding it would be extra work that
doesn't accomplish much.
Increments EDITLEVEL for the third time this week....
Noticed when testing the forcefight against obscured furniture fix:
if you attempted forcefight against the edge of the map, you got
feedback about having already moved as far in <direction> as possible
rather than about forcefight failing to attack anything. Have the
can't move out bounds test check for forcefight before deciding to
give to its normal feedback.
against "unknown obstacle"
Pull request #746 by entrez: give better feedback when 'F' prefix
is used toward furniture covered by objects and towards solid rock
terrain on arboreal levels.
Fixes#746
This is intended to address a couple quirks with force-fighting an
unoccupied spot that I noticed:
* Now that furniture is considered 'solid', an object is much more
likely to be sitting on the square, obscuring the terrain glyph. As
a result, the current glyph is no longer sufficient to accurately
describe the contents of the spot -- e.g., an altar with a corpse on
top of it was being described as "an unknown obstacle", even when the
hero knew exactly what furniture was there.
* When blind and attacking an unexplored 'solid' square, the attacked
position would always be described as 'the stone', even something
like a fountain or sink which didn't seem likely to be confused with
a stone wall.
* The feedback for attacking stone was previously changed from 'solid
rock' to 'stone' in order to be consistent with the feedback for
attacking an unseen wall, but they still weren't quite the same
("stone" vs "the stone").
* The 'stone' feedback for all STONE/SCORR spots was incorrect on
levels flagged as arboreal, where stone is rendered and described as
trees.
This relies on back_to_glyph for positions where the hero is aware of
the terrain and certain other spots (like stone, walls, etc) for which
back_to_glyph produces good results even if they're unseen, and falls
back to the generic "unknown terrain" in other cases.
Pretty long commit message for such a small commit, but oh well...
instead of as fake bear traps
Use the new traps and their tiles when confused gold detection finds
trapped doors and trapped chests. (Large boxes can be trapped too;
they use the trapped chest trap and corresponding tile rather than
have their own.)
Usually these pseudo-traps go away when as soon as they are within
line of sight. (While testing, I noticed that seeing a trapped door
from outside its room rather than inside didn't behave that way.
The door was created by wizard mode wishing; I don't know whether
that was a factor.)
I also discovered that secret doors weren't being handled correctly.
They can't be trapped because of their use of both the doormask and
wall_info overlays of levl[][].flags, but I had a secret door be
falsely displayed as a trap. This fixes that.
We should have obj->tknown and rm->D_TRAPKNOWN so that the hero won't
forget about these traps after declining to attempt to untrap them.
But that's more work than I care to tackle.
"add glyphs+tiles for door+chest traps",
commit d1217b9f25,
missed a couple of S_vibrating_square references, resulting in
screwed up rendering of zaps and explosions.
When trap detection finds trapped doors and trapped chests, it shows
those as bear traps. When the hero comes within view, they revert to
normal and the detected trap is forgotten. This doesn't change that,
it is just groundwork to be able to show them distinctly. Like the
TT_BEARTRAP patch, it increments EDITLEVEL so this seemed like a good
time to put the groudwork in place.
There shouldn't be any visible changes even though internal glyph and
tile values have been renumbered after inserting two new entries.
Adding traps after S_vibrating_square was quite a hassle and suffered
though a couple of off-by-one errors that weren't trivial to find and
fix.
Issue #745 by k2: when using two-weapon combat, the second attack
would still take place even if the first attack caused the hero to
become paralyzed (hitting a floating eye or g.cube).
Cleaver's up-to-three attacks had the same problem but did stop if
the hero underwent life-saving after some fatally damaging counter-
attack. Two-weapons didn't. Make them both stop early if either
paralysis or life-save occurs.
Multiple attacking monster against monster and against poly'd hero
already deal with paralysis; life-saving doesn't apply.
Fixes#745
- Microsoft Visual Studio 2017 Community Edition v 15.9.47
- Microsoft Visual Studio 2019 Community Edition v 16.11.13
- Microsoft Visual Studio 2022 Community Edition v 17.1.5
When not wielding anything, ^X reports "you are empty handed" if
wearing gloves or "you are bare handed" if no gloves. The ')',
'w-', and 'A)' commands were using "empty handed" unconditionally.
Make them be consisitent with ^X.
After this, body part HANDED is no longer used anywhere except in
body_part().
Reported directly to devteam: when a Rider revived, its corpse
didn't get used up.
The change to have delobj() never destroy Rider corpses, like it
won't destroy the Amulet or invocation items, didn't take into
account that they should be destroyed when Riders revive from them.
Add delobj_core() to be able to do that without changing existing
delobj() usage.
I'm surprised hardfought players haven't been all over this one.
Reported directly to devteam: changing levels while riding gave an
impossible warning, "no monster to remove".
mon_leaving_level() was trying to take hero's steed off the map but
that isn't on the map in the first place. Only noticable if built
with EXTRA_SANITY_CHECKS enabled. Normally remove_monster(x,y) just
sets level.monsters[x,y] to Null but with extra checks enabled it
first checks whether that is already Null.
Commit 593c3532fc took the 'catchup'
argument away from repair_damage() and calculated it in the routine.
Commit 699a25c00b put the argument
back but neglected to remove calculating and assigning it, making
the passed value be ignored. Take that away, finishing 699a25c00b.
That affected shop repair messages but wasn't enough to prevent a
shop wall repair display glitch. This seems to make things work
properly but is little iffy.
Add some basic functions to iterate through the monster list,
ignoring dead monsters. Mainly just to allow splitting up code
into discrete functions.
Not quite happy with the get_iter_mons_xy - should probably have
a pointer to iterator data struct, which gets passed through instead,
but this works for now.
Like the commit message for #743, the followup to it was misIDed
as #744. This one is for #744. There's no need to check a name
against all the "<class> of typename" unless it contains " of ".
Pull request #744 from vultur-cadens: the object specification
supported by des.object() was ambiguous for items whose name in
objects[] doesn't include the object class name. For instance ring
and spellbook of protection are just named "protection" (unlike
"cloak of protection" which has its full name).
Fixes#744
The commit message for pull request #743 misidentified it as #744,
so didn't get marked closed properly. Do that now...
Closes#743
and actually do so in the lua files.
Before this, it was not possible to specify (for example) "scroll of
teleportation" in des.object() because there is actually no object
defined in objects.h named "scroll of teleportation", so
find_objtype() failed to find it. Instead, one had to request
"teleportation", but that is ambiguous, and find_objtype() would find
the first defined item with that name instead (ring of teleportation).
In cases of ambiguity, I referred to the des files from 3.6.6 (before
the lua conversion).
mungspaces() returns its argument itself, so `newop` is assigned to `buf`, and always non-null.
`tfg` and `tbg` is assigned to (some addition of) `newop`, so these are also always non-null.
If you have 'fixinv' set to Off and and an inventory of three items,
they'll always be a and b and c. #adjust had you pick 'from' slot
among [abc] and the prompt for 'to' was supposed to show the letter
you picked plus 'd' for 'move to last'. But it was only showing
the 'from' letter itself as likely candidate, omitting the last+1
choice. (Anything after the last letter in use could be picked and
yield the right result, only the list of likely candidate slots in
the prompt wrong.)
Fixed more by trial and error than by understanding why the old code
didn't do what was intended.
There's no 'w-' or 'Q-' for alternate weapon, but context-sensitive
inventory is starting from the object rather than the command so can
finesse that. 'A' does allow alternate weapon to be directly unset
(aka reset to bare/gloved hands) but is not friendly to being passed
queued input.
This adds an extra internal command which only handles unset uswapwep,
even though that is something which is awfully specialized to get it's
own command. Users don't see this command so that shouldn't matter.
For context-sensitve inventory, if player picks the item occupying
the quiver slot, offer '-' as a choice. Like for wielded weapon,
picking that will clear the slot. Also, don't include the 'Q' for
quivering it choice since it is already quivered.
Unlike nethack4, we don't currently have a way to explicitly clear
uswapwep. The sequence #swap, #wield '-', #swap won't work if
primary weapon is cursed and isn't safe to use if alternate one is.
It would be simple enough to just directly call setuswapwep(NULL)
and untwoweapon() in the item-action routine instead of queuing up
some command to do it but that feels a bit unclean. Adding yet
another internal command for it would work too but this one wouldn't
be for variant user-interaction of an existing command.
The fire command could claim that time passed when it hadn't (fill
quiver with ammo, which takes no time, then queue commands to switch
to matching launcher, which should also take no time while queueing,
only during subsequent execution).
If quiver is empty or has ammo in it, give wielded thrown-and-return
weapon (aklys) priority over filling quiver or switching to ammo's
launcher. Don't do that if quiver has non-ammo in it, otherwise
players running Valks who wield Mjollnir with super strength but
want to throw quivered daggers would complain.
When player is being asked what to fill the quiver with, use the
\#quiver command to do that. Using it honors a count to split a
stack, handles switching uwep or uswapwep to uquiver, and gives
feedback. This is actually a fairly substantial change.
For 'fireassist', when switching to a launcher that isn't already
uswapwep pick one known to be blessed or uncursed over one having
unknown BUC status. But use the latter as last resort.
Reported by vultur-cadens: slow digestion from wearing white dragon
scales/mail blocked per-turn hunger and didn't cause any hunger,
unlike ring of slow digestion which blocks per-turn hunger but still
causes some hunger as a worn ring itself. If no rings or amulet
were worn, wearing the suit prevented the hero from ever burning any
nutrition.
Change to treat wearing the suit to be quite a bit like wearing a
ring, unless hero is also wearing an actual ring of slow digestion
(then the hunger cost of the suit is 'free').
Wearing a ring of slow digestion and another ring consumes two units
of nutrition every 20 turns, no matter what suit. Wearing white
dragon scales/mail and two non-slow digestion rings now consumes three
units of nutrition every 20 turns. Using the suit to effectively get
an extra ring finger isn't free.
Fixes#742
Issue #729 by argrath points out that one of the checks in
edibility_prompts() couldn't work.
For the next-meal effect after reading a blessed scroll of food
detection, the check whether a corpse was tainted but not dangerous
because the hero had sickness resistance could never be evaluated.
An earlier condition would cause the routine to return before
reaching that check.
Move it sooner, even though doing so violates the "order by most
to least dangerous" guideline. It was either that or eliminate it
altogether.
edibility_prompts() had a lot of repetitive code that has now been
condensed.
Simplify acid blob corpse handling--for all of eating, not just for
edibility_prompts()--by treating that as "never rots" so that the
'rotted' variable always stays 0. Now checks for that variable
being greater than some threshold don't need to include "and not an
acid blob corpse" as an exception. A side-effect of this change is
that not only do they never become tainted, they'll no longer yield
the "you feel sick" outcome when they're old but not old enough to
exceed the tainted threshold.
Bug fix: edibility testing stopped warning about green slimes.
That worked in 3.6.x, but 3.7 changed the 'cadaver' variable to
exclude them so the check for eating a glob of green slime could no
longer be reached.
Fixes#729
Reported by entrez. Don't make 50% of neuter monsters be flagged as
female. It doesn't matter for live monsters but gets inherited by
their corpses, where female and non-female corpses stack separately.
Noticed earlier when testing the status_hilite_menu_fld() changes:
if you enter a duplicate rule (with different highlighting), it didn't
replace the previous one and the old rule continued to be used.
This still doesn't do replacement, but by adding the new rule at the
end of the list of rules for the specified field instead of inserting
it at the beginning, the new rule gets used.
A change made 5 or 6 weeks ago that was meant to enhance tracking of
artifact creation had an unintended side-effect of making every object
obtained via wishing have its dknown flag be set. That made them
behave differenly from items picked up off the floor, so revert to the
old behavior.
