1) add graves to the dungeon features being tracked;
2) report on known bones (determined by seeing map spot(s) where previous
hero(es) died since there's no guarantee of graves or ghosts);
3) add automatic annotations for oracle, sokoban, bigroom, rogue level,
Ft.Ludios, castle, valley, and Moloch's sanctum. For bigroom and rogue
level you just need to visit that level, for the others you need to get
far enough along to learn something specific (oracle: her room, sokoban:
annotation is either "solved" or "unsolved" depending upon whether all
the holes and pits have been filled, fort and castle: see the drawbridge,
valley and sanctum: see inside the tended temple). Discovering the
relevant locations via magic mapping counts as "far enough along".
There should probably also be automatic annotations for Medusa and the
vibrating square but I'm not sure what criteria should be used for the
former or what phrasing to use for the latter. Demon lord/prince lairs fall
into similar category as Medusa.
TODO: add final #overview as an end of game disclosure option. (I was
planning this even before I saw that nitrohack has implemented it....)
Noticed while working on #overview/#annotate revisions, one of which
will key off the oracle's welcome message. On a bones level, the oracle
could be outside her room, or the room's one time welcome message could be
used up, or both. During bones creation, discard her if she's on the wrong
level (probably not possible, aside from wizard mode ^G), try to put her
back into her room if she's outside it on the right level, and restore the
room's type (if she's still in it, or has been successfully moved back into
it) so that next hero who loads the bones will get her welcome message the
first time that room is entered.
Other special rooms could be fixed up too, provided that they're
sufficiently intact (stealthy hero might enter and get the one-time message
then run away and die elsewhere on the level; at present, next hero won't
get the room's entry message) but this doesn't attempt to deal with that.
I killed a mimic in a shop, then left. My dog or cat entered that shop
where I could no longer see it, and I got
You see a tripe ration appear where it was!
when it evidently ate the corpse, without me being able to see any such thing.
This fix is only approximate but I didn't want to figure out all the
permutations of esp or prolonged monster detection or infravision. (There
probably aren't as many permutations as I first thought since only "ordinary"
pets will take on alternate monster form, so won't ever switch from detectable
via esp to not detectable or vice versa. Maybe this is good enough.)
No fixes entry; this is post-3.4.3 code. There was an early return with
a comment stating that idx==0 was impossible. I took that out since 0 is not
only possible, it's a valid index into the array of transformations.
From a bug report, sitting while
swallowed gave "you are sitting on air," and the intended message "there are
no seats in here" was never reached. Move the latter so that it works, and
add a new one when you try to sit while a grabber is holding on to you.
Also add theft vs donning fixes entry which was left out two days ago.
The message "you stop taking off <that armor>" when interrupted by a
nymph's or monkey's theft attack would only be given if you were using 'A'
to take off the armor. If you used 'T', you'd get "you stop putting on
<that armor>" instead. The fix for that also makes it easy to vary the
nymph message "<the nymph persuades> you to start taking off" to be "<the
nymph persuades you to continue taking off" when taking that same piece
of armor off was interrupted by the theft.
From a bug report, having some
armor stolen while in the midst of putting on armor--when both items have
a multiple turn completion delay--could result in side-effects for the
latter item being reversed even though they hadn't been applied yet. So
you'd lose points of Int and Wis when attempting to put on a positively
enchanted helm of brilliance, or gain such with a negatively enchanted one.
steal() was assigning to afternmv before it had been used to finish the
action of putting on or taking off armor. Fix by interrupting the attempt
to put on or take off armor when being victimized by theft (or being hit by
succubus or incubus seduction). The existing stop_occupation() call wasn't
sufficient because afternmv is different from occupation.
When testing armor theft by nymph I got a message "you dream that you
hear <something or other>" even though I was awake. steal() was leaving
nomovemsg null in order to get the default of "you can move again", but
unconscious() was treating null value as 'yes, hero is unconscious'. I'm
pretty sure its intent was just to guard against passing null to strncmpi()
and didn't really mean that null indicates unconsciousness.
Reformat the DUNGEON_OVERVIEW code in dungeon.c. It's clear from the
way lines were wrapping that the original author used an editor that let
him set tab expansion to columns of four, and when they're treated as the
conventional eight then some longish lines won't fit. Switch to using a
mix of tabs and four spaces instead of all tab characters.
I've separated out my more interesting changes (which will come later).
However, there are a bunch of minor ones included:
1) the lastseentyp array is reused for each level visited, but it wasn't
being reinitialized when creating a new level, so remembered fountains,
altars, and so forth could be erroneously propogated across levels
(the original contributed patch may not have suffered from this because
it handled last-seen data differently than the code which is in place);
2) add 3.5.0 health food store to the list of recognized shop types;
3) make an #annotate value of a single space delete any old annotation
without adding a new one, the way monster and object naming works;
4) the code to discard overview data for a branch of the dungeon which
can no longer be reached (quest expulsion) wasn't capable of doing so
for the very first level (a hypothetical problem since level 1 isn't in
the quest...) and didn't free memory used for user-supplied annotations;
5) reorganize dooverview() where Michael's compiler reported that a
variable might be used before being initialized (it wasn't, but it also
wasn't even needed to achieve the intended result);
6) redo the #overview formatting macros so that they'll work with pre-ANSI
compilers that don't support concatenation of adjacent string literals;
7) function-like macro ADDNTOBUF() was used without terminating semi-colon,
which confused emacs when indenting, so this rewrites it such that it
expects ordinary termination and will work correctly if ever used in the
form 'if (some_condition) ADDNTOBUF(args); else ...';
8) comment out water/ice/lava with #if 0 ... #endif rather than /* ... */.
A priest quest message that was supposed to say "brotherhood" or
"sisterhood" said "itood" instead. Text "%shood" used the post-3.4.3 'h'
modifier to substitute a pronoun in place of the %s value. That's only
a valid modifier when it follows %d (deity), %l (leader), %n (nemesis),
or %o (artifact). Change the substitution routine to leave it as an 'h'
when it follows anything else. [No fixes entry needed.]
> On 01/30/2012 08:20 PM, <Someone> wrote:
> The boulder from a rolling boulder trap can be generated on a
> lava pool. mkroll_launch() in trap.c, line 1584 checks only for pools
> of water.
64-bit linker had a number of warnings that the 32-bit linker did not.
It turned out to be because the 64-bit compiler is more picky about
declaring DLL exports 2-different ways, even if they are essentially
the same. The 32-bit linkder didn't complain. It is described here:
http://support.microsoft.com/kb/835326
The warnings that this suppresses are:
nhdefkey.c
Linking hdefkey.dll
Microsoft (R) Incremental Linker Version 10.00.40219.01
Copyright (C) Microsoft Corporation. All rights reserved.
nhdefkey.o : warning LNK4197: export 'ProcessKeystroke' specified multiple times; using first specification
nhdefkey.o : warning LNK4197: export 'NHkbhit' specified multiple times; using first specification
nhdefkey.o : warning LNK4197: export 'CheckInput' specified multiple times; using first specification
nhdefkey.o : warning LNK4197: export 'SourceWhere' specified multiple times; using first specification
nhdefkey.o : warning LNK4197: export 'SourceAuthor' specified multiple times; using first specification
nhdefkey.o : warning LNK4197: export 'KeyHandlerName' specified multiple times; using first specification
Creating library o\nhdefkey.lib and object o\nhdefkey.exp
nh340key.c
Linking h340key.dll
nh340key.o : warning LNK4197: export 'ProcessKeystroke' specified multiple times; using first specification
nh340key.o : warning LNK4197: export 'NHkbhit' specified multiple times; using first specification
nh340key.o : warning LNK4197: export 'CheckInput' specified multiple times; using first specification
nh340key.o : warning LNK4197: export 'SourceWhere' specified multiple times; using first specification
nh340key.o : warning LNK4197: export 'SourceAuthor' specified multiple times; using first specification
nh340key.o : warning LNK4197: export 'KeyHandlerName' specified multiple times; using first specification
Creating library o\nh340key.lib and object o\nh340key.exp
nhraykey.c
Linking hraykey.dll
nhraykey.o : warning LNK4197: export 'ProcessKeystroke' specified multiple times; using first specification
nhraykey.o : warning LNK4197: export 'CheckInput' specified multiple times; using first specification
nhraykey.o : warning LNK4197: export 'NHkbhit' specified multiple times; using first specification
nhraykey.o : warning LNK4197: export 'SourceWhere' specified multiple times; using first specification
nhraykey.o : warning LNK4197: export 'SourceAuthor' specified multiple times; using first specification
nhraykey.o : warning LNK4197: export 'KeyHandlerName' specified multiple times; using first specification
Creating library o\nhraykey.lib and object o\nhraykey.exp
Add dupstr() as a substitute for strdup() so that out-of-memory
handling will be consistent with the rest of nethack, and make it aware
of nethack's heap logging. It's treated like alloc() so that its caller
can be logged for NH_HEAPLOG.
I put it into use in a few places, but there are lots more candidates
besides the existing calls to strdup() that should be replaced.
Fix some expressions that were supposed to use bitwise '&' but were
accidentally using logical '&&', pointed out by Keni's lint tool. All 3
instances are in post-3.4.3 code, so don't affect the branch and don't
need a fixes entry.
The wizard mode sequence
load bones? y, unlink bones? y, die, save bones? y
works, but
load bones? y, unlink bones? n, die, save bones? y, replace old bones? y
fails if/when external compression is in use. The file gets uncompressed
before being opened to check its existence, then immediately closed, and
re-compressed, changing the file's name, before the deletion attempt takes
place. Then delete_bonesfile() can't find it via the uncompressed name
and the bones saving code reports "cannot unlink old bones".
The code involved doesn't seem to have changed since the current cvs
repository was set up, so this bug has gone unnoticed for a long time.
There's no reason this fix shouldn't go into the branch too, other than the
fact that I don't have that checked out on this machine. If someone wants
to apply it there, be my guest (and move the fixes entry to fixes34.4).
(This covers some thing that Pat found and some things I found while working
on those.)
Unscramble duplicate use of GREPPATH and GDBPATH symbols.
Add some more info to config.h.
Make missing SYSCF_FILE a fatal error.
Make a parse error in SYSCF_FILE a fatal error.
Rename PANICTRACE_GLIBC (et al) to PANICTRACE_LIBC (et al) since FreeBSD
and Mac OS X (at least) also implement the needed API.
Allow SYSCF_FILE to be unreadable by the user (for setgid installs).
If SYSCF, do NOT fall back to the compiled in WIZARD account.
Put WIZARD into sysopt and remove special cases in authorize_wizard_mode().
Part II of the bones tracking patch. Change umpteen different outrip()
routines to handle its new time_t argument, and use formatkiller() instead
of directly accessing killer.{format,name} and killed_by_prefix[]. The
latter is now static within formatkiller().
The many sys/* and win/* changes are untested....
[See cvs log for include/rm.h or doc/window.doc for more complete description.]
Attach hero info, death reason, and date+time to a level that's being saved
as bones. Read such data back when loading a bones file, then treat it as
part of that level for the rest of the game. Dying on a loaded bones file
will chain the new hero+death+date to previous one(s) if new bones get saved.
outrip() now takes an extra argument of type time_t, and interface-specific
implementations of this routine need to be updated to handle that.
Quite a long time ago, the developer/administrator of the 'hearse'
bones respository asked to have bones files augmented so that they could
be correlated with logfile entries. He was forced to approximate it by
comparing file date+time with logfile date, which won't work well if there
are multiple deaths at roughly the same time, or perhaps even on the same
day. This adds character name plus role, race, gender, alignment, the
cause of death, and date plus time of death to the bones file when it is
saved, and reads that data in when a bones file is loaded, then retains
it as part of that level for the remainder of the game. Dying on a level
that was loaded from bones will chain the new dead hero info to whatever
was there from the previous one(s). It's written as fixed length strings
padded with spaces before writing the map and its messy details, making
it easy to spot with a simple file browsing tool rather than requiring
something which can interpret nethack level files. This may need to be
tweaked if players start shelling out of nethack to see whether the
checkpoint file for a newly entered level contains bones info, but at the
moment I'm not going to worry about that.
TODO: I wanted the bones and topten date to match, so am obtaining
the current date+time in done() and passing it to both of those and also
to outrip(). Hence the latter now has an additional argument. So far only
genl_outrip() and hup_outrip() in src and the three outrips in win/chain
have been taught about that; interfaces that supply their own outrip()
need to be updated and probably won't compile right now. Also, code for
formatting the cause of death has been moved from topten() into a separate
routine so that the new bones code can share it. genl_outrip() now calls
it too; the various other outrip() routines should be changed to call it
instead of continuing to duplicate that core code. (I probably should
have made topten.c's killed_by_prefix[] be static in order to force that,
but haven't done so.)
TODO too: there ought to be some way of viewing the data for a loaded
bones file from within nethack. I'll probably add something to the dungeon
overview code to treat it as an implicit annotation, as least in wizard mode.
Showing it in normal play once a level is sufficiently discovered would be
useful, but I'm not sure what criteria should control that. Neither ghost
nor grave is guaranteed to be present, particularly for levels that were
saved as bones, loaded into a subsequent game, then became new bones when
the second hero died there, which can occur an arbitrary number of times.
gcc reports "comparison between signed and unsigned". One header file
uses unsigned long for tty flags, another uses [signed] short. 'unsigned'
seems like the best compromise, but this might accidentally introduce lint
for some other configuration.
Remove date.h and patchlevel.h from win/tty/wintty.c, win/X11/winX.c,
and sys/share/pcmain.c (caveat: the latter two are untested) so that they
don't get recompiled every time any other source file changes and triggers
creation of a new date.h. Only version.c needs to be recompiled in that
situation. Also, Makefile.src was missing a reference to botl.h.
Fix a couple of signed vs unsigned and unused paramater warnings
that pointed to actual bugs. uid values were being handled as int, even
though "modern" systems use type uid_t which could be bigger and is almost
certainly unsigned. There haven't been any reports of nethack falsely
claiming that the wrong user is trying to restore, so in practice this
hasn't mattered, but switch from int to unsigned long to make the chance
of problems be even smaller.
The code to save message history was ignoring the 'mode' argument so
would have attepted to write even when asked to free memory instead. It
isn't currently called by freedynamicdata() so the problem was theoretical
rather than real.
The 'UNUSED' macro is inadequate to handle parameters which are used
by some conditional configurations and unused by others, so there are
still several warnings about unused parameters from save.c and restore.c.
From the newsgroup: if someone adds too many new special levels, dlb
creation during install will give a warning but still exit with success,
and the subsequent installation won't know that the excess files need to be
placed in the playground separately. The result is that some files will
be missing when nethack tries to access them. The newsgroup thread states
that slash'em increased dlb's default limit of 200 files to 300, and the
unnethack variant increased it to 250 and also changed the overflow message
into an error that causes 'make' to quit. (The thread was initiated by
someone working on his own, not affiliated with either variant, who asked
for help figuring out why nethack couldn't find files at the end of the
alphabet. My answer didn't help much; I thought he was working with
separate files rather than with a DLB container.)
I started to go with the too-many-files-is-an-error fix, but instead
went the GNU route ("no arbitrary limits") and made the number of allowed
files become dynamic. It starts at 200 and expands by increments of 40
when necessary.
For text data processed by makedefs at install time, change all
printf and scanf calls that use %lx format to deal with unsigned long
variables, replacing the makedefs hack of a few days ago. It's not as
clean as I would have liked (quite a few casts), because the values
involved are derived from ftell and/or passed to fseek, which deal in
signed longs. But it clears up a few format check warnings by gcc in
rumors.c and pager.c in addition to the previous one in makedefs.c and
uses the right data type even in the places where no warning was issued.
Teach ``make depend'' about the new win/chain code so that the build
rules for that aren't blown away, and then run make depend to get things
up to date. I think hack.h/$(HACK_H) missing botl.h and pcmain.o missing
date.h were the only things significant that turned up.
The comment for CSOURCES says it should have all sources, but the
value had $(SYSSRC) rather than $(SYSCSRC). I've taken the comment at its
word and inserted the missing 'C'; I wonder whether that'll break anything.
Does anybody use ``make tags'' these days?
gcc warned about comparing signed with unsigned for one particular
write() that used an expression for the size argument, and there was already
conditional code to try to handle it for a couple of other compilers. But
this simpler fix should handle it for everybody.
gcc doesn't complain about using %lx to write out a signed long, but
it does complain about using it to read into a signed long. Technically
it's right about the latter, so fix this properly rather than just suppress
the message with a cast.
This is the code I built trying to figure out the large window size issue.
It completely compiles out if not needed (see -DWINCHAIN in hints/macos10.7)
and except for one call during setup has zero overhead if compiled in and
not used. See window.doc for more info.
Defs for UNUSED parms. I know this has been controversial, so use is isolated
to the chain code and windows.c (where it shouldn't be intrusive and saves about
50 warnings).
Hints file for 10.7, but the build process still needs to be migrated from
the branch.
Pat noted that I neglected to drop the SCCS lines on the files I've been
committing, so clean up those and any others I could find where the SCCS
line date is out of date.
From a bug report, if you used the apply command while
not carrying anything applicable except for unknown potion, you would get
"you don't have anything to apply" if that potion wasn't oil but an
inventory selection prompt (with '*' as the default since you wouldn't
have anything considered to be a likely candidate) if that potion was oil,
giving away information.
This fix makes carrying any unknown potion yield the inventory
selection result, unless oil is already discovered and hero can see that
the unknown potion isn't oil.
Conditional on H2344_BROKEN which can be disabled at the top of the file if
necessary. This appears to handle all the cases I was able to reproduce from
Pat's list, with the caveat that msg_window=f fails on very wide windows on
Mac OSX Lion with the default terminal emulation. When the emulation is set
to ANSI, it works (the failure mode is the output from a simple putchar() loop
across the contents of history called up from ^P wraps at about the middle
of the window).
Provide a mechanism for cleanly moving between tentative window system
selections during startup. Now, before a second (or later) system is selected,
the first will be notified that it is losing control. See window.doc.
From the newsgroup: casting spell of drain life at Stormbringer
(or Excalibur or Staff of Aesculapius) would reduce its enchantment just
like any other weapon. Drain resistance should protect against that even
when not actively wielded.
