Wield a polearm and use 'f'ire to automatically hit with it,
if there's a single valid target.
With fireassist-option, will swapweapon to a polearm.
This only applies if quiver is empty and autoquiver is off.
If the program is build without USER_SOUNDS and encounters any SOUND
or SOUNDDIR directive when parsing the run-time config file it only
issues an error message about the first one. When finished parsing,
have it include suppressed ones in the count of errors encountered.
Allowing optional arguments can be risky. The recent glyph changes
added a new argument to the routine that constructs a mirror image
of a tile but not to the call to that routine. Because an existing
argument was optional, the compiler didn't complain about the new
one being missing.
Obsolete optional 'fem' argument ought to removed but this doesn't
tackle that.
The Qt paper doll highlights known blessed/uncursed/cursed items with
a color border. It was trying to force obj->bknown for non-blinded
priest[ess] but passed the old role letter argument to Role_if()
instead of the monster number that's used these days. It was also
potentially modifying an invent item in a way that's observable to
the player but not updating persistent inventory to show that.
Probably didn't matter though; I don't think the situation it checks
for can occur anymore. On the off chance that it could, move the
check-and-set out of #if ENHANCED_PAPERDOLL so that same inventory
update would occur for ordinary paper doll even though that doesn't
care about displayed items' bless/curse state.
Since wearing red dragon scales/mail confers infravision, give that
ability to red dragons. Matters when the hero is polymorphed into one.
Also give it to the Chromatic Dragon, where I don't think it matters.
Make sure g.hero_seq has a sane value during restore before moveloop()
has a chance to update it.
Have curses use g.hero_seq for messages delivered via putmsghistory().
A giant that is carrying a boulder and standing on ice who drowns when
the ice gets melted could die a second time if the resulting pool gets
plugged by the boulder. It results in an impossible from dmonsfree()
about bookkeeping inconsistency when dead monsters are removed at the
end of the turn. The fuzzer escalates that to a panic.
Switch from 'moves' to 'hero_seq' for tracking whether consecutive
messages were issued on the same move and for whether current move
is still same one after played has responded to --More-- with ESC.
'moves' is actually turns and there hasn't been any straightforward
way to track actual hero moves. Add hero_seq for that. It isn't a
counter but is distinct each time the hero makes a move. I wanted
it for curses ^P support but so far use it for checking stethoscope
usage and for shopkeeper behavior when items in a shop are broken by
the hero.
Increment EDITLEVEL due to change in save file contents.
Fix a segfault when polymorphed into a dragon and using ^X.
One inconsistency I've spotted that I hadn't noticed earlier: if
you wear red dragon scales/mail you obtain infravision ability, but
if polymorph into a red dragon, you don't.
Special abilities conferred by wearing dragon armor was implemented in
a somewhat half-assed fashion; extend it to 3/4-assed. Abilities came
from wearing dragon armor but not from being poly'd into a dragon or
for monsters that were wearing dragon armor or actually were dragons.
This covers much of that.
There are umpteen calls of 'resists_foo(mon)' and some are now
'resists_foo(mon) || defended(mon, AD_FOO)' but the second part ought
to be incorporated into update_mon_intrinics() so that the extra
'|| defended()' doesn't have to be spread all over the place and the
ones being put in now could/should be removed.
While testing, I noticed that a monster wielding Fire Brand did not
resist being hit by a wand of fire. This fixes that and should also
fix various comparable situations for other artifacts. But so far it
has only been done for zapping (and any other actions which use the
zapping code). Folding defended() checks into update_mon_intrinsics()
matters more than that probably sounds.
Give messages when sickness countdown has nearly expired to warn
player that hero is dying and also so that death at the end doesn't
seem so abrupt after an arbitrary period with just "TermIll" and/or
"FoodPois" on the status line.
Also, abuse constitution each turn when Sick (either variation, but
not a double amount if both).
This is comparable to the recent fix for tty. When messages aren't
currently being suppressed by use of ESC at --More-- (">>" for
curses), if an urgent message itself triggers --More--, don't start
suppressing messages if player dismisses it with ESC.
This one has me baffled, first how/when it happened and then why no
one reported it. The line
current_mesg->turn = g.moves;
vanished from mesg_add_line() at some point. It is visible in diff
context of commit 99ed00012e from March,
2019 but I can't find any commit since that time which removed it.
[I've been using
git log --no-min-parents --no-max-parents --patch win/curses/cursmesg.c
and then searching within the pager. Maybe that's flaky, but if so,
things wouldn't be any less strange.]
The missing line resulted in mesg->turn being uninitialized, so when
^P compared consecutive messages to decide whether they were issued on
the same turn, arbitrary junk made them all seem to be from different
turns so "---" got inserted before every message. I suppose that if
someone uses a malloc that zeroes the memory it hands out, mesg->turn
field would always be 0 and ^P would behave as if all messages were
from the same turn, so not show any "---" separators. Then players
might not be aware that "---" between groups of messages was intended.
[The messages ought to be grouped by move rather than by turn, but
that's something the core would have to provide.]
Avoids any need for MONITOR_HEAP hackery. Link src/alloc.o and
util/panic.o into util/makedefs.
When replacing the realloc() call in fgetline(), I noticed that
fgetline() would miss the last line of an input file if it lacked a
terminating newline. This was hard to test because OSX seems to be
supplying one when it is missing (VMS would do that too). I had to
modify epitaph (my test bed) to take off the final character, run
'makedefs -s' under debugger control and strip away final newline
that stdio added back, build new nhdat and move it into place, then
run nethack and execute #wizrumorcheck all multiple times before
the fix and once more after it. Much effort for little gain...
Plus some of the recent reformatting: indent labels one space,
replace tabs with spaces, shorten or split wide lines.
When testing the urgent message for having weapon be snagged by a
bullwhip, in between the occasional weapon grabs (which mention
flicking the bullwhip) I saw lots of regular attacks that said
"<mon> swings his bullwhip." That is accurate but seems odd, so
change it to "<mon> lashes his bullwhip." Do same for the hero.
While working on that, I discovered that monsters using a polearm
for a ranged attack always showed "<mon> thrusts <a polearm>" even
for ones that aren't defined as piercing so should be swung rather
that thrust. And they're allowed to do that when adjacent where
there isn't enough room to thrust or swing a long polearm. Now it's
"<mon> bashes with <a polearm>" in that situation.
When you genocided a monster a vampire was shapeshifted into,
the vampire might've stayed in the genocided form.
Always revert back to vampire form if we tried to pick a genocided form.
Dragon scales and dragon scale mails will provide some extra effects
when worn:
- blue: very fast speed
- black: level-drain resistance
- green: sickness resistance
- gold: hallucination resistance
- orange: free action
- red: infravision
- white: slow digestion
- yellow: stoning resistance
gray and silver don't have extra effects - those two are already the
best ones, so don't need any.
Now I've remembered why I didn't follow through with these back when
I originally laid the groundwork. New urgent messages:
having an item of armor be destroyed
having weapon be grabbed by a monster's bullwhip
becoming engulfed
being grabbed by an eel
subsequently being drowned by an eel
dying of petrification
turning into slime then dying due to genocide
dying due to fatal illness
There are lots more candidates.
If messages aren't currently being skipped due to --More--ESC when a
message flagged as urgent is issued and that urgent message itself
triggers --More-- to have the user acknowledge the previous message,
and the user types ESC at this new --More--, message suppression
starts. With the overly simplistic existing code from a few days
ago, it was too late for the current message to override that. Since
the urgent message gets buffered like any other (until another message
needs the top line or until input is needed), it wouldn't be shown
when the next message came along and discovered suppression in effect.
I'm not sure that all the changes here are necessary; there was some
flailing about involved. But it seems to behave as intended now.
When shortening/splitting wide lines I noticed that the save and
restore code for regions had a bunch of those and they could be
shortened by using an intermediate variable. Easier to read too.
Also, change several 'unsigned int' to just 'unsigned' as is used in
most of the rest of the code.
At one point I omitted a (genericptr_t) cast (which should no longer
be necessary...) and discovered that bwrite() wasn't declaring the
input buffer it never modifies as 'const'.
If persistent inventory is displayed and contains an entry for a leash
attached to a pet and the pet's type or name changes, the perm_invent
window didn't get updated to reflect the new leash information:
x - leash (attached to <mon>)
Report was for polymorph but applied to growing into bigger form and
to being (re-/un-)christened as well.
Redo the warning suppression in create_door(). The unreachable 'goto'
was unnecessary and without it there's no need to toggle unreachable
code warnings off and back on. Taking a step back, the whole 'default'
case is unnecessary since rn2(4) will always hit one of the 0..3 cases.
So instead of just getting rid of the unreachable 'goto', get rid of
the panic() too.
This fixes the broken paper doll display. Changes for new glyphs
added code to an 'if' but not to the corresponding 'else'.
Showing map column 0 as something other than blank is still a problem.
The splash window shows a red dragon and says "loading". I'm pretty
sure that that used to be a crisp image but it is now a fuzzy one.
It doesn't use glyphs or their tiles so shouldn't have been affected
by changes to them.
Follow up on some old groundwork. For tty, if the core has designated
a message as 'urgent', override any message suppression taking place
because of ESC typed at the --More-- prompt. Right now, "You die"
messages, feedback about having something stolen, feedback for
"amorous demon" interaction (mainly in case of armor removal), and
exploding a bag of holding are treated as urgent.
The "You die" case is already handled by a hack in top-line handling;
I left that in place so the conversion of 3 or 4 pline("You die.*")
to custompline(URGENT_MESSAGE, "You die.*") was redundant. There
are probably various non-You_die messages which precede done() which
should be marked urgent too.
Other interfaces might want to do something similar. And we ought to
implement MSGTYPE=force or MSGTYPE=urgent to allow players to indicate
other messages that they want have to override suppression. But I'm
not intending to work on either of those. I mainly wanted to force
the magic bag explosion message to be shown since a sequence of "You
put <foo> into <bag>." messages is a likely candidate for --More--ESC.
monstone() was using a hardcoded value when flagging the statue of
a unique monster as historic and didn't get updated when the corpse
and statue flags kept in obj->spe were changed last June by commit
04a8ddcce1.
Closes#643
The way statues of stoned unique monsters were marked as "historic" was
not changed when mkcorpstat flags were updated in 04a8ddc, resulting in
unique statues which were no longer marked as historic and were instead
marked as female.
While fixing that, also use the gender-related mkcorpstat flags to
assign the gender of a stoned monster to the resulting statue.
Indent all labels one space. Having uniform placement makes spotting
them much easier. (Having no indent at all would impact the change
bars of 'git diff'. Those display the last unindented line--which
doesn't start with punctuation--occuring before each band of changes,
so usually the name of the function being changed now that we no
longer have unindented K&R-style function argument declarations.)
While in there, shorten or split various wide lines and replace a few
tabs with spaces.
I tried building with MONITOR_HEAP defined for the first time in a
while. It wasn't pretty.
date.c was calling libc's strdup() instead of our dupstr() so alloc.c
wasn't tracking those allocations when/if MONITOR_HEAP is enabled.
Then it called free() which is actually a call to nhfree() in that
situation. If the file of allocations and releases was subsequently
fed to heaputil, it would complain about freeing pointers that hadn't
been allocated.
Worse, makedefs and tilemap wouldn't link. For MONITOR_HEAP,
makedefs was undefining free() in order to avoid nhfree() but it now
links with date.o so got nhfree() calls anyway. And it wouldn't link
because that routine isn't available without alloc.o. tilemap
doesn't link with date.o but it does call malloc() and free() and it
wasn't undefining free(), so looked for nhfree() when linking and got
the same no-such-routine failure.
If #wizrumorscheck isn't able to open or read the rumors file,
continue on to check engravings, epitaphs, and bogusmons instead of
immediately returning.
Compiling with WANT_LIBNH=1 and TARGETPFX=(...) failed because ioctl.o
could not be found.
This is because the targets are specified with the $(TARGETPFX) but the
libnh targets consumed these targets without the $(TARGETPFX).
