Recent change to the stairs structure now lets each stair keep
the destination level number and dungeon where the stairs go to.
When a level that can be on different depth (such as the Oracle)
became a bones level, and it was loaded in another game at different
depth, the stairs were still pointing to the old level number.
Save it as relative to the current level instead of absolute.
If attempting to checkpoint when changing levels discovered that
the alock.0 or 123wizard.0 file was missing and the game was
running in wizard mode, play continued after reporting trickery
but screen updating was left disabled. An early return in
savegamestateinlock() wasn't resetting the program_state.saving
flag to revert to normal screen updates.
I added a few return statements at the ends of void routines,
where they're optional, because it makes searching for early
returns easier. (Without these then when no early return is
present between current point and end of routine, the search
would move past the routine looking for 'return' later in the
file.)
save_stairs() was placed in between saveobj() and saveobjchn()
so I've moved it. (Has no effect on the recently reported stair
anomalies.) It was also accumulating the total stairway data
size in 'len' and never using that for anything, so I got rid
of it. (Ditto about anomalies.)
I started activating new program_state.saving and discovered that
saving of ball and chain could access freed memory. The change
for the former and fix for the latter are mixed together here (but
easily distinguishable).
The saving flag inhibits status updating and perm_invent updating,
also map updating that goes through flush_screen(). That should
fix the exception triggered after an impossible warning was issued
during a save operation. impossible() goes through pline() which
tries to bring the screen up to date before issuing a message.
During save, data for that update can be in an inconsistent state.
The code to save ball and/or chain when not on floor or in invent
(I think swallowed is the only expected case) was examining the
memory pointed to by uball and uchain even if saving the level had
just freed floor objects and saving invent had just freed carried
objects. So for the usual cases, stale pointer values for uball
and uchain would be present and checking their obj->where field
was not reliable.
Noticed while working on Qt's extended command handling, there
are an awful lot of "seeXYZ" commands. Keep the inventory display
ones (named versions of ')' to show wielded weapon(s), '[' for
worn armor, '"' for worn amulet, &c) and rename the others:
| #seenv -> #wizseenv debugging command
| #seespells -> #showspells '+' command
| #seetrap -> #showtrap '^' command
Also, expand the descriptions of #shell and #suspend a bit in
the Guidebook. LaTeX version is untested.
Move the core's global restoring flag (not the same as main()'s
local resuming flag) to a more logical place. Add a saving flag
in the process, but it isn't being set or cleared anywhere yet.
(Once in use it will probably fix the exception during save that
was just reported, but before that it would be useful to figure
out what specifically caused the event.)
The program_state struct really ought to be standalone rather
than part of struct g but I haven't made that change.
Removing an unused variable for wishing and some reformatting
that whent along with it got mixed in. Removes some trailing
whitespace in sfstruct.c too.
Only lightly tested...
Not caused by a hidden macro this time...
|mon.c:1549:10: warning: declaration shadows a variable in the
| global scope [-Wshadow]
| long flags = 0L;
| ^
|../include/flag.h:392:29: note: previous declaration is here
|extern NEARDATA struct flag flags;
Be prepared for life-saving to contradict "<mon> falls to pieces".
Purely hypothetically at present (with no plans to change) since
golems don't benefit from amulets of life-saving.
I was baffled about why moving the cursor across a fire elemental
kept putting up --More-- until I remembered that I once used
MSGTYPE=stop "[Ff]ire"
to test Qt's handling for that. Turns out that I left it in my
config file. autodescribe feedback should not be honoring that;
honoring MSGTYPE=norepeat is not as clear-cut but this disables
it too.
User sounds were also kept enabled during autodescribe but I have
no way to test them. Like norepeat, disabling just falls into
place.
The pline.c change is unrelated. It just eliminates a wide line
(from adding 'g.') in the source by using a shorter variable name.
This tries to fix the problem of the extra message when a tame
golem is completely destroyed (paper or straw golem burned, iron
golem rusted, wood or leather golem rotted) being issued at odd
times. I basically punted on the visibility aspect since the
original logic was strange: you had to be able to see both the
attacker's and defender's spots and at least one of those two
monsters. Now mon-attacks-mon visibility requires that you be
able to see one of the two and if you don't see both, the unseen
one will be referred to as "it". The "may the iron golem rust
in peace" message is independent of that and may be displayed
after "you have a sad feeling", but now that's intentional and
will refer to an unseen pet by name or monster type, not "it".
This needs a lot of testing and hasn't attempted to address
issue #402: only some attacks that should compeletely destroy
a golem actually do so. (So a hit by fire elemental against a
paper golem does, but passive fire counterattack when a paper
golem hits a fire elemental doesn't, nor does a wand of fire
or being hit by Firebrand.)
Fixes#401
"Name of your starting pet when it is a kitten" could be
construed as meaning that it will no longer apply once the
kitten grows into a housecat. Use "if" instead of "when".
The 'other settings' were in alphabetical order except for
"status condition fields" which presumably started out as
"condition fields". Move it into proper place for current
description.
The revamped options handling was't doing dynamic help properly.
After listing the booleans, it listed them again amongest the
compound options. Since their description field is Null, that
could be a big problem. sprintf(buf,"%s",NULL) on OSX produces
"(null)" but most sprintf()'s would probably crash instead.
The 'other' options (autopickup exceptions, menucolors, &c) were
not listed at all. (I don't remember whether that was also the
case before the revamp.) Now they're listed but not explained.
The 'msg_window' description was unhelpful; this replaces it.
A couple of others were longer than necessary so they've been
shortened. The rest of optlist.h is reformatting wide lines.
Recently added 'safe_wait' option was included in the Guidebook
but not in dat/opthelp; add it.
Kicking a container that had gold in it took the gold amount
away from hero's credit or added to hero's debt, then didn't
give a refund if the container and its gold landed within the
shop. Throwing behaved likewise, just less verbosely.
The problem is caused by addtobill() treating gold specially
and then subfrombill() not being able to perform a reverse
operation. Actually, it may be possible for subfrombill() to
do that, but verifying all its uses is too much work. This
moves the gold handling for drop+selling into its own routine
and adds calls to that for the throwing and kicking refunds.
The other calls to subfrombill() outside of shk.c appear to be
ok as-is. (The calls inside that file are the ones that still
need evaluation if the gold handling is to move to there.)
bill_dummy_object() now uses the same o_id assignment for its
dummy object as split_object() does for its new partial stack.
I don't know whether the old code led to any price glitches.
Noticed while working on Qt status highlighting: if levitation
and flying timed out at the same time, first Lev timeout called
float_down() which reported
You have stopped levitating and are now flying.
and then Fly timeout left stale "Fly" on the status line due to
an optimization which got subverted. ('was_flying' flag was
False due to Fly being blocked by Lev; that's correct behavior,
but the flag is effectively a cached value that becomes stale
when the Lev timeout code executes.)
The bug was wizard mode only because #wizintrinsic is the only
way to get timed flying.
Adds sanity checks for mtrapped and mundetected states.
Fixes cases where those were left in wrong state.
1. Trapped monster (eg. a nymph) teleported out of a trap
2. Monster was hiding under ball or chain, which then got removed
3. While restoring a level, a zombie corpse revived while monster
was hiding under it
4. A general case where the only object was deleted off floor and
a monster was hiding under it
Monsters hiding under ball or chain will now get revealed when
the b or c are moved.
