<email deleted> wrote:
> If more monsters fall through a trap door than can fit on the
> level below, when you go down the stairs, you get the following
> message:
> "Program in disorder - perhaps you'd better #quit.
> rloc(): couldn't relocate monster"
> This message seems to appear once for every monster-too-many that
> fell through the hole. I originally found this while
> intentionally completely filling a level with black puddings
> (there was a trap door I didn't know about). I also confirmed it
> in a wiz-mode test using gremlins and water.
[confirmed: moveloop -> deferred_goto -> goto_level ->
losedogs -> mon_arrive -> rloc -> impossible]
This patch:
- causes rloc() to return TRUE if successful,
or FALSE if it wasn't.
- adds code to mon_arrive() in dog.c to deal with
the failed rloc()
- allows the x,y parameters to mkcorpstat() to
be 0,0 in order to trigger random placement of the
corpse on the level
- if you define DEBUG_MIGRATING_MONS when you build cmd.c
then you'll have a debug-mode command #migratemons to
store the number of random monsters that you specify
on the migrating monsters chain.
> NetHack feedback form submitted by
> <email deleted>
> on Tuesday, September 9, 2003 at 06:41:34
> Hi, Just thought I'd point out a sort of inappropriate word
> choice or typo that I came across in Juiblex's Swamp. I got this
> message, after pushing a boulder into the swamp: There is a large
> splash as the boulder falls into the moat. Obviously it's a swamp
> and not a moat so that sounds a bit wrong. It says the same sort
> of thing when I #dip a scroll in the swamp as well.
Bones loading was only checking to see if a
monster was marked extinct, it wasn't adding
up the born count of a species in the current
game with the number of that species on the
bones level being loaded. That made it possible
to exceed the correct number of nazgul and
erinys via bones.
This adds a common routine called propagate()
that makemon() and restmonchn(ghostly) share,
for incrementing the born count and checking for
extinction, etc.
When a bones level is loaded, restmonchn()
will flag an illegal monster (duplicated unique,
or too many of a species) by setting the
individual monster's mhpmax to the cookie
value DEFUNCT_MONSTER. Before getbones() finishes
loading the bones level, it will purge those
monsters from the chain.
This old conditional was targeted for Apple compilers in a Unix
environment (probably A/UX, which hasn't been sold in over 10
years). It now interferes with installations under MacOS X.
Support compilation of multiple flavors for the Macintosh:
1. A Carbon port, text only and backwards-compatible with
older versions of the MacOS.
2. A termcap/Unix port for MacOS X only.
3. A Qt port, tiled and MacOS X only.
The flavor can be selected at compile-time by specifying the
appropriate prefix file for the compiler.
- fix ring of protection from shape changers causing
real monster to be created.
- add ability to get the character class monster from
genus() or the species.
- use the character class monster when animating
guardian corpses.
> The lighting store doesn't sell oil (probably never added when
> POT_OIL was added). There currently isn't enough room in the
> shclass struct for another item, but that could be expanded. I
> think adding oil there would be useful for the post-3.4.x
> version.
The following files are gone (sys\wince):
- stat.h (moved to ceinc\sys\stat.h)
- fcntl.h (moved to ceinc\fcntl.h)
- errno.h (moved to ceinc\errno.h)
- assert.h (moved to ceinc\assert.h)
The following files were added:
- mhtxtbuf.c
- mhtxtbuf.h
- menubar.uu
- ceinc
- ceinc\sys
- ceinc\fcntl.h
- ceinc\errno.h
- ceinc\assert.h
- ceinc\sys\stat.h
CE notes:
Fixes:
- added new options "wraptext", "fullscreen" and "softkeyboard"
- CE341-1 fix ("wraptext" option)
- hide map scrollbars on Smartphone
- added View->Options menu
- PocketPC: added "Fit to screen" and "Show/Hide keypad" icons
on the menu bar
- Smartphone: '<', '>' keys were mapped incorrectly
- build: use source files directly from <buildroot>\sys\wince instead of
copying them to <buildroot>\wince\
- If you have Gnome installed on solaris, the GETRES support wouldn't build.
I don't have access to a solaris system with Gnome installed, but hacked
unixconf.h to force the GETRES code itself to be compiled. So, I believe the
unixres.c change will work for folks really using Gnome on Solaris.
Whether the rest of the gnome code will build there is beyond me.
- I accidentally left TIMED_DELAY defined and the Solaris build failed.
While solaris has usleep(), this is not part of SVR4 as far as I can tell.
But, SysV does have poll, so I implemented msleep() for SysV systems in
terms of poll. So, you can now define TIMED_DELAY on any SYSV build.
<Someone> wishes to add a couple of new options to the wince port ("run fullscreen" and "do not use CE software keyboard").
The wincap field was full, so this adds a second field for
additional options.
[B04003 and B04004 are still marked "Reported"]
<Someone>:
> You aren't very skilled at reaching from the saddled blue dragon.
> Continue your attempt to set the land mine? [yn] (n)
> You begin setting your land mine.
> There is the trigger of your mine in a pile of soil below you.
> KAABLAMM!!! The air currents set your land mine off!
> I somehow suspect that it'd be more than the air currents, if I were
> trying to arm a land mine from dragonback while not very good at
> controlling it.
<Someone>:
> What do you want to use or apply? [cmu or ?*]
> You aren't very skilled at reaching from the saddled warhorse.
> Continue your attempt to set the land mine? [yn] (n)
> You begin setting your land mine. You escape your land mine.
> Is "escape" really the right word here?
gcc -c -O -I../include -DDLB -DUSE_TILES -oo/cmd.o ../src/cmd.c
../src/cmd.c: In function `rhack':
../src/cmd.c:1800: warning: case value out of range
../src/cmd.c:1801: warning: case value out of range
Finally got around to installing OpenBSD (rev 3.3) in a vmware partition.
Found that several #if BSD's were inappropriate for modern BSD's. Haven't
installed FreeBSD or NetBSD, but based on reading their man pages,
these changes are needed there too. Mostly due to POSIX time() signature.
Switch the default Linux build behavior to use random instead of lrand48,
since lrand48 exhibits some obviously non-random behavior. random() has
been in glibc for a long time. Even if no other changes are made to
nethack's random number generator, this will improve the Linux behavior.
This is derived from the proposed patch and feedback to it. This applies
the last-position cache behavior without an option, making the behavior
more like it is for interfaces with a mouse, where holding the mouse still
acts the same way as the travel cache. The code is not #ifdef'd either.
This allows the use of the right mouse button to
look at things on the screen when the
'clicklook' option is set.
Concept came from a patch for 3.4.0
that I saw referenced on r.g.r.n
[see http://www.steelskies.com/nethack.php]
but the implementation is different.
A while back, I noticed that there was a custom use of the obj->recharged
flag in mkobj.c for tracking corpses on ice. It's much more obvious for
those of us that don't have the entire source base memorized to follow the
usual convention of adding a #define to obj.h. That's what this change does.
The number_pad option can now optionally hold a value
{0,1, 2 } for {off, on, DOS-mode} but plain number_pad and
!number_pad in config files still work as before.
When number_pad:2 is set, iflags.num_pad_mode is set to 1
which triggers the following behaviour:
> '5', M('5') and M('0') are mapped in rhack()
>in cmd.c, only when they are entered as a command. When used as a
>number, like in the 'n' command, no mapping takes place. '0' is
>already mapped to 'i' by the core. The
>only difference [<Someone>] left in (deliberately) is when you press Ctrl-0;
>this used to map to C('i'), which is an invalid command; now
>keep it '0' (which is interpreted as 'i' by the core.)
I encountered a look vs pickup cockatrice corpse
bug today.
If you looked at a location with ':', you
would instantly get
"Touching the cockatrice corpse is a fatal mistake..."
but if you used "m," you got the full list of
things at the location to choose from.
This patch makes the behaviour consistent
and more informative to the player.
You now get the partial list of things felt
up until the cockatrice corpse is encountered,
and then you get the
"Touching the cockatrice corpse is a fatal mistake..."
Before, the code was never displaying the partially
built list because the feel_cockatrice() call was
happening before the window display call.
Mostly `gcc -Wwrite-strings' complaining about passing string
literals to safe_qbuf(). `gcc -Wformat' didn't catch the type mismatch
of formatting the return value of strlen() with %d, presumeably because
size_t is defined as unsigned int on this system and it treats int and
unsigned int as interchangeable as far as printf/scanf checking goes.
I'm not sure whether the sizeof() values being passed to safe_qbuf()
ought to have casts. Any system where size_t isn't the same width as
unsigned int is bound to support prototypes, but might possibly warn about
the implicit conversion of unsigned long (or even unsigned long long these
days) to unsigned int.
this particular fix has been sitting around my inbox for a while although
we've had reports of FreeBSD build problems for a long time. While it's
untested, it certainly looks like the unfixed system.h had a case that
could not be reached. bsdi seems like it needs to be handled the same way.
This solution is mostly a band-aid. Make sure information set by join_map
that is overlaid by the MAP is cleared out. This ensures that place_branch
will never consider invalid data. A new function, remove_rooms(), with a
helper, remove_room(), takes care of this, but only for rooms created by
join_map, which addresses the only known case that causes this problem.
There's a possibility that some other strange behavior, especially in
minetn-6, will be fixed by this as well. The problem of disconnected caves
on minetn-6 is not yet addressed.
Also, add a check to lev_comp.y to make sure the required fg semantics of
joined levels (fg must be ROOM or CORR) are actually met. Doesn't affect
any levels currently included in the distro, but might address levels
others are trying to make.
add another bit to the flags passed to launch_obj so it can print
the initial "rumbling" message at the appropriate time rather than
having the caller print the message, possibly out of order.
1) killer reason for scattered land mine shrapnel used "a" or "an" prefix
even when multiple projectiles hit as a group -- one of various things
From a bug report.oextra field) --
noticed while investigating #1 and later From a bug report.4.0 due to an unintentional side-effect of missile killer reason
handling in 3.4.1 (removal of redundant "poisoned" prefix by m_throw()
confused the poison handling routine) -- noticed while investigating #3.
> The `prompt' buffer in tty_yn_function still only holds QBUFSZ
> characters. But fixing the tty incarnation wouldn't be good enough;
> all the other interfaces would need to handle it too. I think it
> should be fixed in the core instead. Prompt strings simply should
> not be allowed to become so lengthy.
Another step in the fight against prompt sting buffer overflows.
The goes after the ones that may not have been found yet.
This makes yn_function a real core function and removes
the #define yn_function macro.
The yn_function validates the prompt string buffer being
passed prior to calling (*windowprocs.win_yn_function)(),
and if necessary, truncating it and adding "...".
This won't help if the overflow occurs in the core in
a buffer that is still QBUFSZ in size, but it will help if
a BUFSZ buffer is being passed to one of the query
functions.
This patch gives game and savefile compatibility
whether GOLDOBJ is defined or not.
You can build with GOLDOBJ defined or not, and
still load your saved games. Rebuild with the
opposite, and load the same game.
That way GOLDOBJ can be experimented with
more easily.
1. Leave the "you" struct and the "monst"
struct the same under the hood between
GOLDOBJ and !GOLDOBJ.
2. Always write out gold as an
object on the player and monster
inventory chains.
On a restore of the savefile with GOLDOBJ
not defined, take the gold objects out of
the inventory chains and put it into u.ugold
or mtmp->mgold as appropriate.
On a restore of the savefile with GOLDOBJ
defined, nothing special is done.
<Someone> noticed that is_damageable used the address of is_rottable
instead of calling it. Unfortunately, there doesn't seem to be a good
compiler warning to detect uses of function addresses as booleans.
Another fix to address the complaints about two-handed weapons being
rendered useless by 3.4.1's change to require free hands in order to apply
containers. Some players now fear to wield two-handed weapons because a
curse would make accessing their bag impossible, which is doubly nasty if
that's where they have scrolls of remove curse or potions of holy water
intended to deal with cursed items. The same situation applies for cursed
one-handed weapon combined with cursed shield, so some are now claiming
that 3.4.1 has made two-weapon combat be even more attractive than before.
This implements #tip, a new command that causes a container at the
current location or carried in inventory to have its contents emptied
onto the floor. Hero's hands don't need to be free at the time but tipping
a floor container requires limbs; tipping an inventory container doesn't
need hands or even limbs. The contained items don't pass through inventory
during the process, so don't cause objects (loadstones, crysknives, scrolls
of scare monster?) to go through their special handling unless it's part of
normally dropping to the floor. Tipping a bag of tricks behaves the same
as applying it (one monster is released, and it only becomes empty if
that happened to be the last charge) and items tipped out of a cursed bag
of holding have their normal cursed bag chance (1/13) of being destroyed.
Tipping an inventory container while levitating or during unskilled riding
behaves similar to normal drop--from a height, so some fragile items break.
Players have wanted this feature to get gray stones out of chests or
heavy corpses out of ice boxes but I didn't care much about that; losing
access to your bag is more significant. I'm pretty sure that there was a
user patch to do something like this floating around at one time, but I
couldn't find it when I looked, so I implemented #tip totally from scratch.
Bug? Extended commands which lack meta-key shortcuts are not listed
in the help files displayed by the '?' command....
Move a couple of instances of container contents manipulation into
their own routines. Behavior for items disappearing from cursed bags of
holding isn't quite identical but is effectively the same. I think its
use of stolen_value (or simply the behavior of the latter) is buggy, but
I haven't tried to fix that. (Cursed bag of holding destroying a player
owned bag containing shopped owned items definitely doesn't work well.)
<Someone> noted that if you dropped a box while levitating, nothing would
break. This is true for sacks too, which shouldn't be immune either.
Throwing didn't break contained objects either, not mentioned in his report.
Refactored out the code in kick_object that handled damage for objects
contained in normal containers and added calls in hitfloor and throwit.
kick_object still only calls it for boxes, but other callers will call it
for any object letting it decide if damage is required. BoH objects aren't
affected, since traditionally the inside of the BoH is "somewhere else".
