Reduce the implied reliance of a specific version of lua.
Instead of copying liblua.a to src/, copy it to lib/. Instead of
telling the compiler to look for headers in lib/lua-5.3.5/src/ as
well as in include/, copy the relevant ones to lib/ and tell the
compiler to look for them there. 'make spotless' in src/ will
remove both the object library and the header files from lib/ but
there really should be a new Makefile.lib to take care of that
directory.
Update Makefile.src to be able to build lua in case someone starts
with 'make all' there instead of in the top Makefile. It doesn't
duplicate the option to fetch the lua source package though.
NHinternal/../genFiles.c has been updated to mention lib/liblua.a
and lib/lua*.h as 'generated at compile time' in Files and to skip
lib/lua-* entirely if it comes across that (so not operating on a
completely clean tree). But it won't be accurate unless/until
other ports stage their lua files there instead of in src/ and
lib/lua-$(VERSION)/src/.
I haven't tried 'make depend' to see what it makes of the numerous
changes....
Developed for 3.6 but deferred to 3.7. Most of the testing was with
the earlier incarnation.
Report was that pronouns were accurate for the underlying monsters
when hallucination was describing something random, and also that the
gender prefix flag from bogusmon.txt wasn't being used. The latter
is still the case, but pronouns are now chosen at random while under
the influence of hallucination. One of the choices is plural and an
attempt is made to make the monster name and verb fit that usage.
|The homunculus picks up a wand of speed monster.
|The large cats zap themselves with a wand of speed monster!
|The blue dragon is suddenly moving faster.
There is no attempt to match gender for the singular cases; you might
get
|The succubus zaps himself [...]
or
|The incubus zaps herself [...]
Unix Makefile.utl wasn't aware of the dependency of makedefs.o on
src/mdlib.c so didn't rebuild makedefs when it should have.
Eliminate several warnings:
mdlib.c - #if inside the arguments to macro Sprintf();
nhlua.c - nhl_error() ends with a call to lua_error() which doesn't
return, but neither of them were declared that way;
nhlsel.c - because of the previous, the 'else error' case of
l_selection_ellipse() led to complaints about uninitialized
variables;
sp_lev.c - missing 'const'.
I did minimal testing which went ok, but revisiting a couple of levels
gave me un-freed memory allocated by restore.c line 1337. (I haven't
looked at that at all.)
This is similar to the helm of opposite alignment case fixed some
time ago. Deferring the setting of foo->known until an item is fully
worn (because it used to get set earlier but gave away information if
the wear operation was interrupted) didn't take into account that foo
might end up Null in various circumstances. So Boots_on() needs to
validate uarmf before setting uarmf->known in case putting on boots
of levitation while on a sink caused them to come right back off.
I put similar validation into all foo_on() just in case (as far as
I'm aware, only Boots_on() and Helmet_on() actually need that).
I think the previous expression would allow moving (via swapping
places) from a pool to solid rock or closed door which was not what
was intended (but moot since there aren't any pools on the Astral
level). This revised expression does what is intended: can only
swap to a pool location if already located in/over (the Riders fly?
they should probably be non-breathing) another pool.
Don't let Riders swap places with something (fog or ooze, perhaps)
located at a closed door spot because if it gets killed there, there
won't be any corpse and it will stop auto-reviving.
Just avoid moving to spots where mondied() won't place a corpse
instead of worrying about whether a bargethrough creature (if there
ever are any besides the Riders) might be able to survive at the
destination (so ignore pass-walls, door-opening, swimming, &c).
Noticed while testing something: hero drank a potion of see invisible
and nearby invisible monster could now be seen--in theory--but I was
asked what to call the potion while the updated map was buffered. So
I didn't see the invisible monster until after naming the potion.
pline() flushes buffered map updates, but getlin() doesn't. I didn't
change that, but I've made docall() do so since the updated map may
make a difference in what the player can tell about whatever is being
'called'.
Fix the issue where a hallucinatory monster name which begins with
a slash is having that stripped off as if it was a gendor and/or
personal-name flag.
The main issue was pronouns ignoring hallucination and this doesn't
attempt to address that.
Also, add new hallucinatory name "leathery-winged avian" which has
been lurking for a while.
Provide a little more information when dumplog is unavailable.
While testing various permutations, I encountered a couple of
problems with conditionally declared variables.
Have the --showpaths feedback mention whether dlb is in use or not,
and show the container file name(s) when it is. Users of prebuilt
binaries or who build with a hints file instead of picking and
choosing things in config.h might not know, and vms (if it ever
catches up with --showpaths) uses a different container name from
everybody else ("nh-data.dlb" instead of "nhdat").
The mapglyph() change made a variable obsolete but it got left in
(idx = SYM_PET_OVERRIDE...). Take it out and fix up the formatting
for the block of code that had it.
remove_region() calls newsym() when removing gas clouds, but when
newsym() checked whether it was updating a gas cloud location it
always got a false 'yes' because the region hadn't been removed yet.
Fixing this didn't seem to make any observable difference so it must
be followed fairly rapidly by a full vision recalc.
Don't show the gas cloud glyph at locations where monsters can be
sensed (telepathy, warning, extended detection). It will work
better when/if vision of gas cloud locations gets fixed. (Such
clouds behave as the hero can see into them, so warning doesn't
have any unseen monsters to show unless they're unseen for some
reason other than the gas cloud.)
Region processing does a lot of looping--when there are actually
regions present--and calls functions in those loops which do more
looping of their own. This moves some of the simpler tests so that
they get done sooner and can avoid some of those function calls.
I was hoping that it would speed up the turn cycle on the Plane of
Fire where the spontaneous irregularly shaped fumaroles are composed
of a lot of small regions but I don't think there's any noticeable
difference.
In process of doing that, I discovered a bug (no doubt copy+paste
which escaped an intended update) with monster handling. The check
for whether a monster is entering a region depends upon whether the
hero is in that same region rather than whether the monster is
already inside. So a monster can enter a region--or have a moving
one enclose it--with impunity if the hero is already in that region.
Once the hero moves out of it, the monster will finally enter it.
'orient' is the name of an enum defined in wincurs.h so don't use it
as a variable name in cursstat.c. My compiler didn't complain using
'-Wshadow' but apparently some other one does.
Make the same change in the dead code located in the second half of
that file, plus a couple of formatting tweaks.
From hardfought; latest gcc complains that /* fall through other stuff */
doesn't match its pattern for /* fall through */ comment indicating
that omitted 'break' statement is intentional and one switch case is
deliberately continuing into the code for another.
While not a path exactly, the dumplog file isn't placed somewhere
fixed so being able to see where it is placed could be useful.
This cascaded a bit during testing. Fix one of the warnings from
hardfought (fqn_prefix_names[]). And a few more that came up with
SYSCF disabled (panictrace_gdb, two unused variables if files.c).
A comment in macosx10.14 describing how to create a signed package
had the last sample command wrong. Fix it now in case that part
carries over to hypothetical--but likely--new macosx10.15 hints.
Also, shorten a couple of wide comment lines in the same area.
The report mentioned whistles but I had forgotten all about them by
the time I tried to deal with musical instruments. Plain whistles
had deaf handling but magic whistles didn't.
Fixes#240
Monster versus monster (melee and throwing) didn't handle shades
(need silver or blessed weapon to take damage) or silver feedback
(extra info when silver-haters are hit).
I did a lot of test, revise, re-test but didn't always re-test
everything that had previously been tested, so bugs that I thought
were quashed might have crept in.
Now if a missile weapon "passes harmlessly through the shade" it
will continue on and maybe hit something else. (Regular misses
still stop at the missed target.)
A couple of minor ball&chain changes accidentally got included.
Menus with wide header or separator lines were rendered wide enough
to avoid wrapping those lines, but ones with narrow header/separators
and wide selectable entries were limited to half the display even
though lots of lines that would fit with full width were being wrapped.
Change the latter behavior.
Menus are right justified with the edge of the map when narrower than
it, left justified otherwise, and if the display is wider than the map,
they'll extend beyond its right edge. (That hasn't actually changed;
it's just that left-justification is more likely now that menus will
be wide enough to show wide inventory lines without wrapping.)
Get rid of my ridiculous hack to force wider menu for the 'symset'
and 'roguesymset' sub-menus of 'O' since it's no longer useful.
There's still room for improvement. If any lines need to be wrapped
despite using the full width, or perhaps are just a lot wider than
most of the entries, menu width could be narrowed to just enough for
'normal' lines to fit so that one or two really long entries don't
distort the menu. That's a bit more complicated than I want to deal
with right now. [If implemented, it would be relevant for tty too.]
Fix a couple of warnings about unused variables but mainly handle
OSX's multiple-choice for configuration file name. I tried moving
my ~/.nethackrc to '~/Library/Prefences/NetHack Defaults.txt' and
both regular play and --showpaths are finding it successfully.
Also an untested fix for VMS. "nethackini" is not supposed to have
any path information attached.
Move makeplural(body_part(FINGER)) into its own routine, with option
to substitute gloves when wearing such.
Wearing slippery gloves (ie, wearing gloves while having slippery
fingers) wouldn't let you put on a ring because you can't take the
gloves off, but removing a worn ring lacked the same restriction.
After changing that, teach prayer that slippery gloves is another
reason why a ring of levitation can't be removed.
Slippery fingers would transfer from bare hands to gloved hands if
you put gloves on. The reverse, transfering from gloves to bare
hands when taking gloves off, was already being prevented for
directly taking them off, but still allowed the slipperiness to
transfer when gloves were lost. This prevents putting on gloves
when fingers are slippery and attempts to handle cases where gloves
get unworn by ways other than 'T' (or 'R') or 'A'.
There's no slippery attribute for objects (way too much work for too
little value); slippery gloves is just the combination of wearing
gloves and having slippery fingers (which now has to have happened
while already wearing those gloves). This changes inventory to use
"(being worn; slippery)" when applicable and much of the patch deals
with funnelling Glib changes through new make_glib() to try to make
sure that persistent inventory adds or removes "; slippery" right
away when changes happen.
If gloves are taken off involuntarily (shapechange to a form that
can't wear them, destruction via scroll of destroy armor or monster
spell of same or via overenchantment, theft), slippery fingers ends
right away instead of the usual few turns later.
When I expanded the Guidebook's sample configuration file I added
several status_hilite options. I decided that I'd better test what
was written and discovered that if Xp had an up or changed rule as
well as one or more percentage rules, it was showing bogus changes
whenever the integer value of the percentage changed. The fix
turned out to be simple but it took a while to figure out.
I ultimately left the status_hilite settings out of the sample
options, because they tended to be too wide for Guidebook.txt's
formatting rather than because they weren't working as expected.
Expand the sample configuration file a little and prevent it from
going past the right margin in Guidebook.txt.
Replace all instances of "config file" with "configuration file".
Reformat the "notes" at the end of the table of map symbols.
Unfortunately Guidebook.pdf from Guidebook.ps from Guidebook.mn
puts a page break between the header line "notes" and the two
actual notes.
Value 1 for 'mouse_support' was not just exceeding the margin of
Guidebook.txt but wrapping to the next line. Shorten it.
Guidebook.tex had a typo "in the foler" (where 'folder' was meant)
and Guidebook.mn didn't have that phrase at all.
Remove a few trailing spaces.
Noticed that "multiple" was misspelled for entry about menu action
':' misbehaving under curses, but the phrasing for whole entry was
difficult to comprehend so try to word it better.
[This one is in the highlights list with the old wording.]
Reported directly to devteam, the POSIX_TYPES subset (most? all
these days?) of Unix that defines USE_FCNTL was unlocking lock file
'perm' when done with it but wasn't explicitly closing it unless
the unlocking failed. Triggered a valgrind complaint and could have
posed a problem if restart gets implemented for this configuraiton.
Feedback when playing music while hallucinating misspelled
"butterflies".
Other bits in the same code (not part of #H9407):
All feedback messages while impaired gave "You produce <something>"
which was immediately followed by many of the instruments giving
their own "You produce <some other thing>." Change the verb for the
playing-while-impaired messages to avoid having two consecutive
"you produce" ones.
Also, multiple impairments (two or more of stunned, confused, and
hallucinating) always gave the generic "what you produce is far
from music" message. Have them sometimes ignore excess impairments
to give the message for one of those.
Wizard mode shows the number of points needed to reach the next level
(unless already maxxed out at 30) for ^X and end of game disclosure.
Do it in normal play for the latter too. (I think it would ok to do
that for ^X too but haven't gone that far.)
Even when it was wizard mode only, the phrasing for past tense had a
minor grammar bug, and it could make the line a little too long for
tty and curses (not sure about others) when level was high, resulting
in wrapped text. That looked bad for tty, which first tries removing
indentation (just 1 space in this case), making that line outdented
as well as wrapped. So change the phrasing slightly when experience
level is 'too high'. I had a version which formatted, measured, and
re-formatted if necessary but that was overkill; simple hardcoded
rephrasing suffices particularly when measuring was against assumed
display width (80) rather than actual width.
Turns the "fix" in commit 319dcf4746
handled removing the default answer for single-line-prompt plus
multi-line-answer but not for multi-line-prompt plus long-enough-
answer-to-reach-another-line. The logic wasn't quite right and I
misunderstood what is stored in linestarts[] so even correct logic
wouldn't have solved things.
Symbol set changes...
curses:
S_tree (use plus-or-minus sign instead of accidental horizontal line);
S_bars (switch from default '#' to not-equals sign);
DECgraphics:
S_altar (switch from default '_' to 'pi');
S_bars (switch from 'pi' to not-equals sign);
DECgraphics_2: get rid of this commented out set since its contents are
now folded into DECgraphics.
(The IBMgraphics one is just whitespace; delete a <space> that precedes
a <tab>.)
Something I realized while following up a newsgroup post. If you
knew an item's bless/curse state and dipped it into water while
blind, the bknown flag stayed set and you learned the item's new
bless/curse state without seeing any "glows blue/black/&c" feedback.
Clear the flag unless you know that the potion being dipped into is
water (or is clear if not water has not been discovered) and also
know the water potion's own bless/curse state.
Music wasn't using You_hear() so needs to handle Deaf itself. Have
it give alternate messages for sounds being emitted from instruments.
This doesn't implement the suggestion that a Deaf hero shouldn't be
able to produce the same music as a non-deaf one.
More symbols: I left out a word on the continutation line for S_stone
and S_tool has been forcing narrow space where ordinary <comma><space>
was intended.
Guidebook.tex not tested. The \, in the continuation of S_stone is
a narrow space in LaTeX. It should be \hspace{ width of '(' } but
I don't remember how to measure something in LaTeX or TeX.
Guidebook.mn is getting two small changes: I forgot about the
guideline that each sentence be on its own line. Also, this adds the
fix for wide lines in the dungeoneers list although the reordering
accompanying a recent change might have made that moot (unless some
future reordering puts wide names on the same line again, assuming
that was the issue).
The table listing symset entries was ok for ps/pdf output but for
plain text output it was wrapping to a second line for S_xan,
S_pet_override, and S_player_override and also going past the right
margin for some others. I've shortened the text for those three
and narrowed the first column to give more room for the third column,
bringing all the lines within their margin.
The entry for S_stone was rather surprising: "dark part of a room".
It now says "solid rock or unexplored terrain or dark part of a room"
and needed to span two lines.
The first column has been labeled "Default" but that was why that
column was wider than necessary. It now has a label of two spaces
and is left justified instead of being centered. It is also now
forcing constant-width Roman as the font and every character in that
column should be the same size so one reason for centering is removed.
The title used to be centered but always looked a bit off due to the
fact that the three columns have differing widths and/or because of
the over-wide lines changing the location of "center". It is now
left justified and looks a little bit like it's labeling the first
column. Perhaps it needs a blank line separating it from the rest.
Guidebook.tex is lagging again.
Creation of guardian angel bypasses tamedog() and marks it tame
directly but it wasn't updating the map after changing the monster.
So if 'hilite_pet' was On when entering the Astral Plane level, the
angel appeared to be ordinary monster rather than a tame one until
it moved or the screen was redrawn. Also, the message about it
appearing was issued before marking it tame, so a tiny bit of code
reordering has been done to get the sequence correct.
This sample comment attending the sample SYMBOLS statement fit ok in
the pdf output but wrapped to another line in the plain text output,
so shorten it.
S_xan, S_pet_override, and S_player_override still wrap, but by a bit
less than before. Their width causes the line of dashes at the top
of the table to wrap too.
Guidebook.tex is still lagging behind.
I think it was issued by tbl but it wasn't about any of the tables.
warning: numeric expression expected (got 'R')
was complaining about
.lp SYMBOLS, ROGUESYMBOLS
Exclosing the paragraph label within quotes to make it one argument
to the lp macro or command would have fixed this, but I've split
them into two paragraphs. And reordered several of the paragraphs
which were alphabetized just enough to make the out of order ones
seem like mistakes.
Guidebook.tex needs to catch up.
Make paranoid_query() (yn question requiring explicit "yes" answer)
protect itself from overly long prompt strings. I'm not aware of
any specific overflowing queries so I temporary reduced QBUFSZ within
paranoid_query() in order to test.
For EDIT_GETLIN, don't use previous response as default if we loop
after neither "yes" nor "no" was given for paranoid confirm.
Fixes#238
For the three fields in the 'scores' option's argument: top, around,
and own, if any was separated from preceding one by space(s) rather
than slash and lacked a count prefix, the argument parsing skipped
over it. So "10t/3a/o" and "10t 3a 1o" worked but "10t 3a o" ignored
the 'o'. The issue report was about 'own' but there's nothing special
about 'own' itself; just that it doesn't warrant a count prefix and is
usually last (in other words, normally preceded by one or both of the
other two) so more likely to trip over this.
[I thought there was another report about 'scores' misbehaving (from
quite a while ago) but couldn't find one. If it exists, it might have
been about the same thing.]
Mimic-as-slime_mold needs to keep track of the fruit index the same
way that mimic-as-corpse keeps track of corpse's monster type. The
mimic description was changing (for '/' and ';' feedback) whenever
the player assiged a new fruit name.
That wasn't noticeable when applying a stethoscope because
mimic-as-slime_mold always yielded "that fruit is really a mimic".
Change it to report the fruit's type instead of generic "fruit".
For 'O's menu, make the current symbol set be pre-selected so that
the set in use is clearly marked while contemplating changing it.
Using Return or Enter will pick it again; Escape is now needed to
deliberately not make any selection.
Also, change several symbol set initializations to use the new method
of deciding whether the default symbols are still in place.
Fix a couple of bugs I stumbled across while testing something else.
The sell prompt for a container dropped in a shop had phrasing issues.
This fixes a couple but there are more. The message composition
assumes that contents fall into two categories: those already owned
by the shop and those the shopkeeper is offering to buy from the hero.
But there is a third: stuff the shopkeeper doesn't care about so
won't buy. The count_contents() routine can supply total contents or
shop-owned contents. Subtracting one from the other yields combined
hero-owned without any way to separate out shk-cares and don't-care.
Dying in a shop while carrying partly eaten food would place that food
on the floor without marking it no_charge. But marking it that way
wouldn't have helped because as bones data gets saved, every object on
the level has its no_charge flag cleared. So 'no_charge' needs to be
explicitly set for partly eaten food in tended shops as a bones level
gets loaded.
Most of the shk.c diff is reformatting, but it does change the
get_pricing_units() routine to lie that the quantity is zero for
partly eaten food so that when multiplying with price it won't matter
whether the price has been forced to zero or been left non-zero.
One of the comments for pull request #234 mentions that the colors
for sleep and poison are backwards. Yellow dragons had green breath.
Green dragons would have had yellow breath if they used the normal
zap symbols, but they use the poison cloud one which is green so
they had green breath. I rarely have color enabled and had never
noticed. (At the moment I can't find where the switch from zap to
cloud is being done.)
Pull request #234 is based on #231, about control of the shield
effect animation when resisting. But it changes struct flags so
would break 3.6.x save file compatibility.
Fixes#235
For initial options under curses, specifying 'DECgraphics' as a
boolean rather than as 'symset:DECgraphics' wasn't overriding the
new default 'symset:curses'. Since previously DECgraphics was
rejected for curses, it's possible that no one noticed.
Primary and rogue symbols were being set to default if primary hadn't
been given a value, possibly clobbering rogue symbols if those had
been given a value. Initialize them independenly.
Return early from curses_convert_glyph() if the value doesn't have
the 8th bit set since it now deals exclusively with DECgraphics
handling. Force a sane value for returning early on rogue level.
Change the way symbol sets are loaded to make them have the same order
as they appear in the symbols file rather than being reversed.
Revise dat/symbols so that the new ordering yields a result similar
to the old ordering, more or less. I've added a few set descriptions.
The only substantive change is marking DECgraphics as primary-only
(not available on rogue level) and adding new set DECgraphics_2 which
is commented out near the end.
Define symbol handling H_MAC since one of the sets specifies
'handling: MAC'. All H_MAC is used for now is to avoid showing
MACgraphics as a symset when compiled without MAC_GRAPHICS_ENV (which
was used for pre-OSX Mac by the old code in sys/mac/), so it will be
hidden for everyone.
I left handling H_CURS even though curses doesn't implement anything
for it. It could do something when rendering the map or assign a
function to 'cursesgraphics_mode_callback' for special init or both
but hasn't needed to. Since curses is now supporting DECgraphics,
define 'decgraphics_mode_callback' for it. No value is being
assigned so that doesn't do anything; curses seems to be setting up
the primary character set as text and secondary one as line-drawing
without the need for that hook.
With the added set descriptions, 'O's symset menu looked horrible for
curses due to the way curses decides to set the width of menus and
the resulting line wrapping which took place because of a too-narrow
menu. I've added a chunk of code to the options handling code which
shouldn't really be there but makes the menu much easier to read.
Lastly, do some formatting cleanup in files.c.
Move a declaration that became mid-block when a preceding 'if () {'
got removed to top of block suppress warning about C99 feature.
Add new entry for the curses symset change to fixes36.3.
This time I'm putting things in as-is before making a few tweaks.
The pull request was three or four separate changes. I used the
patch instead so they've been collected into one commit.
Highlighting for monsters shown due to extended monster detection and
for lava shown in black and white didn't work because that keys off
of 'iflags.use_inverse' (actually a macro for 'iflags.wc_inverse') and
curses wasn't enabling that window-capability option. To be fair, it
was probably unconditional at the time the curses interface was first
developed. It checked for whether a monster was supposed to be drawn
with inverse highlighting but wouldn't draw it that way because the
flag was always false. Inverse b&w lava is relatively new and curses
hadn't been taught about it.
Various other things such as pets (if hilite_pet is on) and object
piles (if hilite_pile is on) get highlighted with inverse video when
use_color is off, regardless of whether use_inverse is on or off.
That's probably a bug.
Fixes#230
Incorporate github pull request #230, support for DECgraphics-style
line drawing in the curses interface. I've rewritten the
curses_convert_glyph() part so that it doesn't require C99 and
doesn't reinitialize its pair of arrays for every character written
to the map. The DECgraphics conversion is now a straight char for
char one, DEC line drawing code to ACS, without regard to what map
symbol is intended or what 'cursesgraphics' uses for that symbol.
For the DECgraphics symbol set, down ladder is the greater-than-or-
equal-to character as intended and up ladder is less-than-or-equal-to,
but comments in dat/symbols had their descriptions transposed.
If you chat with a peaceful gnome while hallucinating, you might
get a silly message from the TV show South Park.
To make it work for non-gnome characters, I changed the speech of
monsters who normally just grunt (gnomes, orcs, ogres, a couple
of other groups) to full humanoid when the hero is hallucinating.
(It already does that for orcs if the hero--hallucinating or not--
is an orc and for gnomes when the hero is a gnome.)
If mksobj() was told to initialize the object it's creating and the
object class was something it didn't understand, it would issue a
warning and return Null. But an unknown object class is a severe
internal error and very few callers were prepared to deal with a
Null result, so change mksobj() to panic instead. Also eliminate the
few attempts to deal with Null result that are present in mkobj.c;
I didn't go looking elsewhere.
Noticed while investigating the message loop. If I had level files
from an interrupted game and was asked "Destroy old game?" when
starting a new one, answering 'n' left the terminal in an unusable
state. Executing 'stty sane' (invisibly since input echo was off)
repaired things but the user shouldn't have to do that.
Change unixtty.c's error() to shut down windowing if that has been
initialized. This might need some tweaking for tty, which will now
clear the screen before showing the startup error message. Other
systems besides unix use unixtty.c so are affected, but I think the
change doesn't introduce anything that should cause trouble (aside
from the potential screen erasure).
Sword given to angels used obj->spe = max(obj->spe, rn2(4)) [except
using a temporary to sanely work with max() macro]. But the obj was
explicitly created as no-init, so obj->spe was always 0 and the max()
was pointless. Shield given to angels was manipulating bless/curse
state directly instead of using the functions intended for that, a
no-no and also pointless to be clearing 'cursed' for a no-init item.
Mace for priests had useless handling for object creation failure.
Object creation failure could only happen if the mksobj() call had a
valid entry in objects[] (or out of bounds access that didn't crash)
for an object class that it doesn't know how to handle. That can't
happen unless somebody screws up big time. If it ever did happen,
it would have produced a memory leak.
Report #H9243 misinterpreted W_WEAPON as W_WEP and attributed a
hypothetical ball and chain sanity checking problem to that.
Rename the former to W_WEAPONS to emphasize that it includes
alternate/secondary weapon and quivered stack as well as wielded
weapon.
I had this in place at one point but must have accidentally undone it
before deciding that yesterday's patch was finished. Defer fetching
'pw' until it's needed.
Fixes#26
Report stated that the attempt to look up the player's username
(on Unix) failed (reason unknown) and nethack refused to allow the
player to execute the #explore command even though sysconf was set
to use character names (CHECK_PLNAME=1) instead of user names.
Setting EXPLORERS to "*" overcomes this glitch, but the fix moves
a bit of code around to honor CHECK_PLNAME before fetching username
so that that isn't necessary.
I ended up doing some formattng clean up (replace tabs with spaces,
whitespace cleanup in 'port_insert_pastebuf()'). The actual change
to fix#26 is only a few lines.
Another part of github issue 227. Casting a function pointer when
passing it to another function is iffy when lying about the return
type. tputs() expects a routine which returns int, so give it one.
Other xputc() usage is equivalent to putchar(), so define xputc()
with the same function signature as that has.
The tputs() declarations in system.h should probably be changed
(third argument is a function which takes an int rather than
unspecified parameters) but I've left them alone. I made that change
to tputs() in sys/share/tclib.c though.
NT and MSDOS changes are untested. tclib.c compiles ok with clang-
as-gcc on OSX but hasn't been tested with the port that uses it (VMS).
Fixes#227
Travel, <ctrl><direction>, <g|G><direction> all stop on engravings,
but <shift><direction> told the player what the engraving said and
kept going. The message output is buffered until map update or
another message, so player couldn't tell where hero was at the time
the engraving got shown. Make <shift> running stop on engravings.
More from github issue 229: mark routines in util/dlb_main.c which
don't return as such and add some 'break' statements for compilers
that don't have support for that.
Another part of github issue 229, mixtype() didn't have either 'break'
or '/*FALLTHRU*/' separating healing from extra healing, extra healing
from full healing, and full healing from unicorn horn. So dipping
"bad" potions (sickness, confusion, blindness, hallucination) into
healing/extra healing/full healing or vice versa operated the same as
dipping a unicorn horn into the bad potion (producing fruit juice for
sickness and water for the others). It wasn't clear from the code
whether or not that was intentional. It actually seems reasonable
(albeit suboptimal use of {plain, extra, full} healing), so continue
to allow it and make the code clear that it's intentional.
Fix the first part of github issue 229. sortloot_classify() tries to
group musical instruments separately from other tools, but missing
'break' in a 'switch' prevented that from happening--they were mixed
together.
Since that grouping isn't documented anywhere, only source divers
would ever notice that it wasn't working as intended.
when polymorph causes loss of levitation boots or water walking boots
while over water. If discarding stuff while trying to crawl out got
rid of the taken-off boots, they wouldn't be in inventory any more
when break_armor() tried to drop them after taking them off.
when attempting to crawl out of water fails and hero is life-saved
or wizard-/explore-mode player declines to die. If the hero is saved
by positioning him/her one step further away from ball and chain,
teleds() tries to drag them, but if 'emergency disrobe' left the hero
stressed or worse, dragging fails. Ball and chain were being left
where they were, with chain no longer being adjacent to hero.
If drag_ball() fails, have teleds() teleport ball and chain to hero's
new spot, same as when that isn't within range of a one step drag.
A couple of early returns could result in temporary windows getting
left around instead of being released for re-use, which in turn might
lead to a panic due to lack of available window slots. The first
one is accompanied by an 'impossible' warning which no one has ever
reported and the second one could only happen if data file 'keyhelp'
was missing, so panic due to either of these is hypothetical as far
as released versions go. Somebody making modifications could run
afoul of either of them though.
query_category() - switch from early return to 'goto' so that the
temporary window used for a menu will always be destroyed;
whatdoes_help() - defer creating the display window until after the
data file has been successfully opened so that early return won't
need any window cleanup.
Reported directly to devteam rather than via the web contact form:
throwing wielded aklys while swallowed would hit the engulfer and
return to the hero's hand but leave a stale 'thrownobj' pointer if
the monster survived. Under usual circumstances, throwing anything
else or throwing the aklys again when not engulfed would clear that
pointer, putting things back to normal. However, killing any engulfer
with the same weapon would try to add it to engulfer's inventory to
be dropped as it died. If the killing blow was via melee rather than
another throw, the object in question would still be in hero's
inventory instead of free, hence panic.
The initial returning-aklys implementation shared Mjollnir's code
which doesn't have this issue. This reverts from having attached
aklys always returning successfully when thrown while swallowed to
Mjollnir's 99% chance of return and 99% to be caught when it does
come back. (That was already the case if the engulfer was killed by
the throw, where hero wasn't swallowed anymore after the damage was
inflicted.)
Zapping wand of opening or spell of knock at self while trapped:
"You are released from pit."
ought to be
"You are released from the pit."
Likewise for most of the other held-in-place situations.
Also, when released like that vision wasn't being recalculated right
away to update line of sight to reach beyond the edge of the pit.
Fixes#223
The touchstone code treated all rings as if they had gemstones, but
quite a few don't and feedback could be unexpected. Cited case was
an iron ring yielding a cyan (hi_metal) streak instead of the normal
iron result ("scritch, scritch"). A gold ring yielded a yellow
streak rather than a golden scratch. I didn't test silver ring but
suspect it yielded a silver streak rather than a silvery scratch.
This changes touchstones to treat non-gemstone rings like other
classes of objects instead of like gems. I made mineral rings keep
acting like gemstone rings--I'm not sure whether that's right.
Fixes#221
Routine unfixeable_trouble_count() is used for both applying a unicorn
horn (possibly internally via #monster if poly'd into a unicorn) and
drinking a blessed potion of restore ability. For the latter case, it
always gave the wrong answer (unless the hero happened to be all of
Sick and Stunned and Confused and Hallucinating and Vomiting and Deaf).
Since the actual count wasn't used to decide whether hero felt "great"
or just "good", having any of those conditions would hide the problem.
Fixes#216
A github pull request changed one of the fake mail messages so that
our web site's URL is added at compile time instead of being hard-
coded. However, it wouldn't compile with a pre-ANSI compiler since
it relied on concatenating adjacent string literals. This is more
complex but achieves the same result, and also makes the existing
run-time subsitution be a bit clearer.
Testing was a hassle but eventually successful.
PatR