Two semi-related patches within the span of less than three days and
I overlooked the overlap. When the() deals with a capitalized string,
it calls CapitalMon() to check whether the string matches a monster
and if so whether usage should be "the Monster" (Oracle or Olog-hai
for example) or just "Monster" (Medusa). The first time that gets
called, it collects all capitalized non-the Monsters from both mons[]
and the bogusmon file. The latter just got changed to pad short
lines, and that works fine for selecting hallucinatory monsters at
random via get_rnd_text(), but non-the Monster collection processes
the bogusmon file directly and wasn't updated to strip the padding.
While testing 'the("Capitalized Hallucinatory Monster")' I noticed that
some hallucinatory monsters showed up more often than others. When
the random engravings, epitaphs, and bogus monsters were converted from
hard-coded arrays to data files accessed by random seek (3.6.0), they
became subject to the same distribution irregularites that rumors suffer
from. The chance that an entry will be chosen depends upon the chance
that a random seek will hit somewhere in the line which precedes it, so
entries that follow long lines are more likely to be chosen and entries
that follow short lines are less likely. We improved that for rumors
by having makedefs pad the shortest lines. Distribution still isn't
uniform but is much better than it was (and could be further improved
with a longer padding length at the cost of making data files bigger so
possibly slower to access; both overall size and access speed mattered
back when floppy discs were supported but are probably irrelevant now).
Start doing the same thing for the newer files: pad the shortest lines
to increase the chance that seek will find them. The tradeoff is that
the data files become bigger. Rumors, engravings, and epitaphs lines
are all at least 60 characters now; bogus monsters are at least 20.
These are the data file sizes I see (in bytes: old, new; padding for
rumors was already in use so its size hasn't changed):
bogusmon 4449 7211
engrave 1326 2894
epitaph 14159 24075
rumors 49173 49173
The only place that padding is noticeable in-game is #wizrumorcheck.
A shop-owned glob picked up by the hero was added to shop's bill
and if that shrank to nothing it moved from the unpaid portion to
used-up portion as intended. But once there it retained obj->owt
of 0 and if 'sanity_check' was enabled, that triggered a warning
every move until finally paid for. Both the 'Ix' list of used-up
items and itemized shop billing revealed a weight of 0 aum if
'wizweight' was enabled.
Keep track of the weight a glob had when it becomes unpaid, then
reset from 0 to that amount if it becomes used-up. This overloads
the obj->oextra->omid field which is an unsigned int previously
only used for corpses. Now for globs it is pre-bill obj->owt which
is also unsigned int. I didn't add new oextra access functions for
it; it is only used in two places and existing omid ones suffice.
Reported by entrez: when fruit name is given the name of an artifact
that doesn't use "the" at the start of its name, messages about the
artifact could be altered. Example was fruit=Excalibur causing usual
|You are blased by Excalibur's power!
to unintentionally change to
|You are blased by the Excalibur's power!
because of a false match during special handling for named fruit in
function the().
This fixes that, and also changes basic inventory formatting. Former
|f - an Excalibur
will now be
|f - Excalibur
for a fruit that has been assigned that name. When sort pack in On,
as it is by default, that will be listed under Comestibles rather than
under Weapons so really shouldn't fool anyone. And
|f - 2 Excaliburs
also breaks the illusion.
This formatting change only affects named fruits. User assigned names
for object types or for individual objects behave the same as before.
Noticed while testing revised "the" handling and trying to trigger
"you can't polymorph into [the] Oracle", if the character was human
it would turn into a new man or woman instead of having the unique
form be rejected.
Function the() wasn't supposed to be used for monsters because many
of the ones with capitalized names confuse it, but over time multiple
instances of the(mon_nam()) have crept into the code. Instead of
ripping those out, modify the() to handle that situation better.
Pull request #636 by entrez dealt with this with one extra line of
code, but could end up scanning all the names in mons[] repeatedly
if the("Capitalized string") gets called a lot. This uses a similar
one line fix but calls a whole new routine that scans through mons[]
once collecting all the relevant special case names. As a bonus,
it does the same for hallucinatory monster names which name_to_mon()
couldn't handle.
Fixes#626
The hero in an aquatic polyform loses HP for time spent out of the
water; allow magical breathing to prevent this, just as it allows the
hero in her non-aquatic natural form to breathe underwater.
Also add a similar rule for monster fish-out-of-water HP loss, even
though currently monsters can't use amulets of magical breathing and
there's no non-breathing fish/eel -- just in case this changes at some
point.
Reported by Vivit-R, the chance of freeing a monster from being
trapped in a web was very small and failure resulted in creation of
a new web at the hero's spot, eventually making it impossible to
attempt to #untrap the monster (if hero escaped the new web and moved
to another spot adjacent to trapped monster, eventually surrounding
it with webs). [That's actually misleading. Escaping the new web
but staying at the same spot lets the player try again; repeat as
needed.... But chance of failure still seemed way too high.]
This reduces the chance of failure from 29/30 to 6/7 and makes the
chance of spreading the web after failure be 1/3 instead of 100%. It
also supplies missing feedback about the monster still being trapped
if the attempt to make a new web fails.
Closes#608
Incorporate the functionality of the loadable DLL's (nhraykey.dll,
nhdefkey.dll, and nh340key.dll) into the consoletty.c code and
remove the dll building
For a glob in a container carried by the hero, shrinking away to
nothing would have indirectly updated the container's weight when
obj_extract_self() was called, then the 'old_top_owt' value would
never be different from current topcontnr->owt. That only matters
for the shrink-but-not-gone case and could only happen for the gone
case so didn't result in anything noticeably wrong. But fix it to
match the comment about weight not being adjusted yet.
If the hero left a level that had globs on the floor or in floor
containers or being carried by monsters and stayed away for a
while, returning to the level only shrunk them by one unit of
weight. Account for all the time away. The complexity of this
has steadily grown; I hope its peak has been reached.
When carrying a glob, possibly inside a container, give shrink
feedback more often (twice in the ~500 turn cycle to shrink from
20aum to 0aum rather than just once).
I had wizweight On when testing glob changes and noticed
|Slasher drops a gold piece (0 aum).
Coins are supposed to weigh 1/100 of a unit, and the calculation
rounds rather than simply truncates any fraction, but that still
yielded 0 for quantities of 1..49. Force any non-zero stack of
gold to weigh at least 1 unit.
Also, add a check for attempting to weigh a quantity 0 or less
(of anything, not only for gold) just in case.
Don't hardcode the weight (20) of an unaugmented glob, use
objects[].oc_weight (also 20) instead.
When a glob inside a container has decayed all the way to nothing
(weight 0), take it out before updating the container's weight.
Otherwise weight() would use objects[].oc_weight instead of 0 for
that glob.
Globs never rotted away but did become tainted after a relatively
short while, which seemed like a contradiction. Change them to never
be tainted but shrink by 1 unit of weight approximately every 25
turns. An ordinary glob (one that hasn't combined with any others)
starts out weighing 20 units, so it takes about 500 turns to vanish.
That's roughly twice as long as a corpse takes to rot away.
Shrinking globs give feedback when in hero's invent or in a container
in hero's inventory, but rarely (when going from an exact multiple
of 20 weight units; that is, from integral number of N globs to
N-1 + 19/20, or if weight reduction triggers an encumbrance change).
When a glob goes away completely, there is feedback for those two
circumstances and also for seeing the glob vanish from the floor.
I haven't touched how much nutrition eating a glob confers. I have
changed formatting of glob names to use "small", "medium", "large",
"very large" instead of "small", [no adjective], "large", &c. You
still need to have at least five globs coalesced together for the
adjective to become "medium", same amount as before.
I don't think EDITLEVEL needs to be modified but have incremented it
anyway to play things safe.
Giving new_light_source '(genericptr_t) mtmp' leads to the light
source's id.a_monst being set to 'mtmp->nmon' rather than 'mtmp',
since that's what is stored in the initial byte of the monst struct.
When mtmp->nmon == 0x0 this can cause a segfault in do_light_sources.
I thought there were more places that checked for "it" and substituted
"someone" or "something". Perhaps there are and I'm just not finding
them now. Anyway, this extends x_monnam() and adds some_mon_nam() and
Some_Monnam() to do that during monster name formatting instead of
having various bits of code try fix it up after the fact. The fixups
could be fooled by monsters given the name "it" or "It"; x_monnam()
won't be.
Guard against any other places besides minimal_xname() that set up a
fake object without being aware that for boulders the corpsenm field
should be 0 rather than NON_PM. If that field is unexpectedly -1,
xname() will format it as ordinary "boulder" rather than producing
special "next boulder". An explict value of 1 is now required for
the latter.
Reported by Vivit-R with comments by several others. The prize item
in one of the closets off the Sokoban treasure zoo is sometimes
missing, most likely picked up by an elf who won't be dissuaded by
the presence of engraved Elbereth or a scroll of scare monster.
This fix prevents monsters from targetting the mines' and sokoban's
prizes for pickup (or for eating). Once the hero picks either of the
prizes up, they stop being prizes and will be ordinary monster fodder
if dropped/stolen/stashed.
One of the comments by copperwater suggested this approach as a
possible way to fix things. I had already implemented it from scratch
before noticing that. It handles the usual monster behavior toward
items, but there could easily be some unusual cases still susceptible
to taking the prize before the hero gets to it. Those are the breaks.
Fixes#603
like the leg hackery, if the baalz level gets flipped during creation
the fixup that modifies the locations in front of the eyes to allow
digging needs to deal with that too.
Because some spots in the fly's 'legs' on the Baalzebub lair level are
specified as pools in the level file, then later converted to walls in a
post-generation fixup routine, monsters can be generated on those spots
and then left walled up and inaccessible, Cask of Amontillado style.
For the love of God, makemon-tresor!
Some code already existed to relocate these monsters after generating
the level, but it depends on misorientated 'leg segments' being fixed up
in a particular way. That wasn't being triggered because it didn't
account for the possible rotation of the level; as a result, the
monsters in the leg segments wouldn't be relocated, and the leg segments
themselves would continue to have the wrong orientation.
Account for possible level rotation so that the monsters are relocated
properly (and the leg segments are 'fixed').
Reported by eakaye. Selection of equipment when creating a soldier
or watchperson can pick a polearm, but random selection among those
had a chance to choose dwarvish mattock which doesn't use polearms
skill and isn't appropriate for a human soldier or watchperson.
Not mentioned, but lance was in the same boat.
Change the selection to only pick something which uses polearms
skill, then make that moot by moving lance and dwarvish mattock out
of the midst of the polearms so that they're no longer candidates
for special case rejection.
A couple of other things which might have had a similar issue were
already ok. Giving a polearm when creating a troll selects between
a few choices rather than among all the polearms. And wishing for
"polearm" only considers items which use polearms skill.
While changing objects.h to reorder the two non-polearms, I removed
a bunch of tabs that were present in the scroll definitions.
EDITLEVEL is incremented due to objects[] reordering, so existing
save and bones files will be invalidated.
Fixes#623
Another (latent) case of an artifact possibly being generated and
immediately deleted: part of the process of a mimic disguising itself as
an item involves generating a random object, then deleting it. If this
item is a box or sack, it will generate with random contents, which will
be deleted along with the container. If artifacts are allowed as random
box contents, this can silently remove an artifact from being available
in the game.
This is effectively blocked already, since none of the artifacts
eligible for random generation are items from classes marked as valid
box contents (see boxiprobs[] in mkobj.c). Nonetheless, formally
preventing artifacts from generating as box contents will guarantee this
issue won't crop up if a randomly generated artifact tool, ring, etc, is
added in the future.
Death drops will generate a random item, which may be an artifact.
After the object is created, some rules are applied to check whether its
size matches the monster that dropped it; if it fails these checks,
the item is deleted. aosdict recently pointed out that if the death
drop was an artifact, this would permanently remove it from the game.
Use artifact_exist(..., FALSE) in various cases where potential
artifact objects are created, then immediately removed. This will
prevent artifacts from being removed from play by marking the artifact
as available to create again.
A vampire in bat form was seen via infravision or possibly telepathy,
then when it changed into fog cloud form the feedback was
|You now detect it where the vampire bat was.
The message substitutes "detect" for "see" when the new form can't
be seen and the monster name formatting yields "it" for that case.
Give a vanish message instead since that is effectively what happens.
The pull request by argrath changes weight_cap() to never return a
value less than 1 because try_lift() divides by that return value
and a 0 would trigger a crash. The code involved is used when
attempting to pull a monster out of a pit via #untrap.
I'm fairly sure that weight_cap() can never produce a value that's
less than 1 already, but have put in a variation of the PR's fix.
I've also implemented a different fix that removes the division
from try_lift(). The original code seems to have gone out of its
way to avoid calculating inv_weight() twice, but doing the latter
(for the once in a hundred games where it might happen) greatly
simplifies things by removing details of carrying capacity.
Fixes#621
Reported by G7Nation; attacking walls and such with a polearm just
gave lackluster "you miss; nobody's there" feedback.
Make applying a polearm at some non-monster locations give feedback
similar to using 'F'orcefight with melee weapons. Was
|You miss; there is no one there to hit.
now
|You uselessly attack the closed door.
Also, extend the supported locations to include dungeon furniture.
Was
|You attack thin air. ('F')
now
|You harmlessly attack the throne. ('F')
|You uselessly attack the throne. ('a')
This doesn't address #613: attempting to hit non-visible locations
with an applied polearm.
Closes#614
Reported by entrez, wielding something fragile (potion of acid
perhaps), and using F to smash it against iron bars called breaktest()
directly, then a second time indirectly through hero_breaks() via
hit_bars(). There is a random chance to resist breaking (99% for
artifacts, 1% for other items) so breaktest() might say that something
will break on the first call and that it will not break on the second
call, or vice versa. That could remove uwep from inventory then leave
it in limbo without destroying it, or destroy uwep without removing it
from inventory first triggering impossible "obfree: deleting worn obj".
whitelist the valid cases showing up
If an earlier version of clang is showing more cases (particularly
if they don't make sense), the re-enabling of the warning in
sys/unix/hints/include/compiler.2020 can be made clang-version
specific instead. I had no way to test earlier versions.
Give feedback for "<Mon> jumps onto a polymorph trap" in more
circumstances. It used to give the message if the monster could
be seen, and it would map the trap if the monster's location could
be seen. That led to some confusion when an infra-visible monster
was at a spot in the dark so the trap was accurately described but
didn't get mapped.
Map the trap if either the monster's or the trap's location can be
seen. If it isn't already known and isn't becoming known (the
infravision case), say "hidden trap" instead of "polymorph trap"
in the message. Give the "jumps onto trap" message if either the
monster can be seen or the trap's spot can be seen. When monster
can't be seen, it would normally be described as "It"; switch to
"Someone" (when humanoid) or "Something" (non-humanoid) instead.
Minor unrelated bit: the use-container action was between the 2nd
and 3rd polymorph cases. Move it, after the third one.
Engraving with a weapon which has known enchantment or with a magic
marker which has known charge count that decremented the amount did
so without updating the persistent inventory display. Report was
for curses but applied to every interface that can show perm_invent.
This includes some reformatting (mostly
if {
}
else {
}
to
if {
} else {
}
multiple times; also add some new braces after 'if' or 'else' where
one half already had those and the other didn't). And removal of
some unnecessary 'register' designators.
Plus fix up one sanity check which complained if number of charges
being used was more than a marker had, but then let the engraving
continue and decrement the count by that large amount. If that
impossible situation ever occurred, it would result in a marker
with negative charges.
Noticed when testing the "wall_angle: unknown" fix, if there is a
boulder rather than a door in the breach in a vault's walls at the
spot where the guard arrives, the guard would walk onto it, tell the
hero to drop any gold and follow, then move back. The boulder would
remain in the hero's way and couldn't be pushed because the guard was
in the boulder's way. Have the guard smash any such boulder(s) into
rocks when arriving (with no explanation for how that is accomplished,
just a message about seeing or hearing boulder(s) shatter). Later
when repairing the walls, delete any rocks or boulders at all vault
wall locations (even when no gap-to-wall repair is needed).
simpleonames() calls minimal_xname() which was setting up a dummy
object and making its corpsenm field be NON_PM. For a boulder, that
yielded "next boulder" instead of the intended "boulder".
Reported directly to devteam by entrez via email:
>
> I noticed some potential issues with (melting) ice:
>
> * Digging down into ice, or setting a land mine on the ice and
> triggering it, doesn't remove the melt_ice timeout, so it can result
> in a sequence like dig down -> pit fills with water -> freeze water
> -> freezing water tries to set melt_ice timeout -> duplicate timeout
> impossible. Or if you don't freeze the water again, melt_ice will
> run on a non-ice surface, which might at least produce strange
> messages.
>
> * Setting a land mine on ice: melting ice doesn't do anything with
> the trap, so there is still a land mine which you can trigger by
> flying over the water (the land mine's trigger is also still
> described as being 'in a pile of soil', despite being underwater at
> this point). Similar thing happens with bear traps.
>
> * Not really related to _melting_ ice, but an exploding land mine
> doesn't reset the typ from ICE to FLOOR (like normal digging does),
> so it will result in a square with a pit that is also an ice square,
> where the ice can melt under the pit and produce a combination
> pit/moat. If you then freeze the moat, the pit reappears on top of
> the ice.
Fix the vault repair issue that could lead to "wall_angle: unknown"
warning. Unlike shop repair, the original wall info isn't available
so this recreates it. The extra 'flags' field added yesterday could
be eliminated but this leaves it in place.
Fixes#606
