Disclaimer: This is a minimal recipe, just to get someone else
started if they have a desire to get a full cross-compile of
NetHack-3.7 going for the Amiga. Some NetHack code bitrot was
corrected, and it does seem able to compile the game itself
to a point. See caveats below.
- If you want to obtain the cross-compiler and tools/libs for Amiga
https://github.com/bebbo/amiga-gcc
To our knowledge, a pre-built copy isn't available, so you have to
obtain the source via git and build it on your system.
The build prerequisite packages for Ubuntu are easily obtained:
sudo apt install make wget git gcc g++ lhasa libgmp-dev \
libmpfr-dev libmpc-dev flex bison gettext texinfo ncurses-dev \
autoconf rsync
The build prerequisite packages for macOS are apparently easily
obtained via homebrew, but that was not tested:
brew install bash wget make lhasa gmp mpfr libmpc flex gettext \
texinfo gcc make autoconf
After installing the prerequite packages and the cross-compiler
it was a straightforward build:
git clone https://github.com/bebbo/amiga-gcc.git
cd amiga-gcc
make update
[Note that you may have to take ownership of the files in the
bebbo repo via chown before succesfully carrying out the next
steps]
make clean
make clean-prefix
date; make all -j3 >&b.log; date
The compiler pieces are installed in /opt/amiga by default which
was satisfactory for our initial attempt, but if you want you can
alter the prefix before you build if you want. That is all
spelled out on the page at: https://github.com/bebbo/amiga-gcc
The Amiga cross-compile can then be carried out by specifying
CROSS_TO_AMIGA=1 on the make command line.
For example:
make CROSS_TO_AMIGA=1 all
make CROSS_TO_AMIGA=1 package
You can explicitly include tty and curses support if desired, otherwise
you'll end up with a tty-only cross-compile build. The SDL1 pdcurses
support has not been tested.
make WANT_WIN_TTY=1 WANT_WIN_CURSES=1 CROSS_TO_AMIGA=1 all
Also note that building the amiga targets using the make command
above, does not preclude you from building local linux or macOS
targets as well. Just drop the CROSS_TO_AMIGA=1 from the make
command line.
The cross-compiler hints additions are enclosed inside ifdef sections
and won't interfere with the non-cross-compile build in that case.
CAVEATS: The original NetHack Amiga build steps included the source for
some utilities that were built and executed on the amiga: txt2iff and
xpm2iff as part of the NetHack build procedure on amiga. Those did not
compile out-of-the-box on the linux host. They will either have to be:
- ported to build and run on the linux or macOS cross-compile host
or
- their functionality will have to be rolled into amiga NetHack
itself and executed on the target Amiga the first time the game
is run, perhaps.
Good luck amiga aficionados, perhaps you'll be able to take this
initial effort forward and get NetHack-3.7 available on the amiga or
amiga-emulator. Let us know if you do, and we can roll changes in
if you provide them.
- If you want to obtain the djgpp cross-compiler and tools/libs for MSDOS,
which is available for linux and macOS, you can use the following script
to obtain it:
sh sys/msdos/fetch-cross-compiler.sh
That script won't install anything, it is just file fetches. It will
store the cross-compiler in subfolders of lib and the hints files are
configured to find it appropriately there.
Note: Both the fetch and the msdos cross-compile package target require
unzip and zip to be available on your host build system.
Cross-compiler bits:
https://github.com/andrewwutw/build-djgpp
and the pre-built binary for your platform from:
https://github.com/andrewwutw/build-djgpp/releases/download/v3.0/
and a DOS-extender (for including in msdos packaging) from
http://sandmann.dotster.com/cwsdpmi/csdpmi7b.zip
and pdcurses from:
https://github.com/wmcbrine/PDCurses.git
The MSDOS cross-compile can then be carried out by specifying
CROSS_TO_MSDOS=1 on the make command line.
For example:
make CROSS_TO_MSDOS=1 all
make CROSS_TO_MSDOS=1 package
You can explicitly include tty and curses support if desired, otherwise
you'll end up with a tty-only cross-compile build:
make WANT_WIN_TTY=1 WANT_WIN_CURSES=1 CROSS_TO_MSDOS=1 all
Also note that building the msdos targets using the make command
above, does not preclude you from building local linux or macOS
targets as well. Just drop the CROSS_TO_MSDOS=1 from the make
command line.
The cross-compiler hints additions are enclosed inside ifdef sections
and won't interfere with the non-cross-compile build in that case.
Expand the use of the sys/unix Makefiles to be used for both normal
local builds and installs, as well as cross-compiles for other
platforms/targets.
Up until now, the primary unix Makefiles have treated util/host-side
component compiles, links and target object files just the same as
the game component compiles, links, and target object files.
Unfortunately, that meant that cross-compile effort typically had
to re-invent Makefiles specific to the cross-compile, creating a
maintenance burden and deviation from the typical local unix build
and providing a daunting obstacle to those that want to establish
build for a target environment/platform.
This change distinguishes between util/host-side component builds,
links, and component builds and targets object files destined for
the game (and other target platforms) in the Makefiles.
In theory, this will ease the effort for people that want to try to
resurrect NetHack perhaps on an old platform where it is no longer
viable to build NetHack-3.7 on the platform itself using old, outdated
compile tools, possibly with an old, outdated C dialect.
Some details:
- Game-related targets in the Makefiles (as opposed to util/host-side
targets that will be executed on the host), which could be destined
for another platform in a cross-compile scenario are prefixed with
$(TARGETPFX) so that they are distinguished.
The default scenario where no cross-compiler is involved, is to
define TARGETPFX to nothing, and therefore meant to have no effect.
- Game-related compile and link commands in the Makefiles and their
associated command line flags are distinguished from util/host-side
compile and link commands in the Makefiles by using $(TARGET_CC),
$(TARGET_CFLAGS), $(TARGET_LINK), $(TARGET_LFLAGS), $(TARGET_CXX),
$(TARGET_CXXFLAGS), $(TARGET_LIBS).
Those are used in the Makefile in place of $(CC), $(CFLAGS), $(LINK),
$(LFLAGS), $(CXX), $(CXXFLAGS), $(LIBS).
The default scenario where no cross-compiler is involved, defines
the TARGET_ version of those Makefile variables to match their
typical non-TARGET_ ounterparts.
- The dependency lists in the Makefiles includes the $(TARGETPFX)
prefix for stuff that would potentially be produced from a
cross-compile build.
- It adds pregame targets and $(PREGAME) variable, so that hints files
can add some additional stuff if required for a cross-compile
scenario.
The default scenario where no cross-compiler is involved doesn't
do anything for $(PREGAME).
- It adds $(BUILDMORE) target and variable, so that hints files
can add some additional things to be built for a cross-compile
scenario.
- It adds a "package" target and $(PACKAGE) variable, so that hints files
can add steps for the target platform in a cross-compile
scenario.
The "install" target assumes local build and placement and
isn't really applicable to a cross-compile scenario where the results
really just need to be bundled up for transport to the target platform.
- Also, this adds a pair of include files that can be updated with some
cross-compile recipes as they evolve. They are named "cross-pre.2020"
(for stuff to be included in the PRE section) and "cross-post.2020"
for stuff to be included in the POST section via sys/unix/setup.sh.
Those are included in sys/unix/hints/linux.2020 and
sys/unix/hints/macOS.2020 hints files.
Consider the following scenario: There's a level where there's a zone of
des.replace_terrain() between the stairs and some other objective, and
the terrain is something non-walkable like trees. There's a chance that
the path is entirely blocked off by random replace_terrain, so you
make the level set terrain to '.' along a randline (or normal line, or
whatever) between the randomly placed stairs and the other side of the
replace_terrain zone. The problem: this overwrote the stairs with a '.'
as well.
This can be worked around in the lua file by first picking the desired
location of the stairs, then setting the terrain that overlaps with the
stairs, then doing des.stair() after that, but this is awkward and hard
to read.
So this makes it impossible for anything calling SET_TYPLIT (only called
in sp_lev.c) to overwrite stairs. I can't really think of a situation
where a level designer would want to define stairs, then maybe overwrite
them.
This was leading to problems with special themed rooms which were
irregular. Walls of ordinary rooms count as part of the room, and
irregular rooms should be no different.
This also makes doors on the room edge be considered as part of the
room, which affected special room entry messages and filling.
All irregular rooms' walls getting marked as SHARED instead of their own
room is probably a latent bug in upstream NetHack, but will prevent
future issues for when/if themed rooms that involve special
rooms/subrooms get added.
This is a simple && vs || bug. The clear intention of the code is that
stairs aren't supposed to generate in themed rooms unless there is no
other choice.
Fixes#348
Aimed at fixing the problem where the player knows they're going to
forget a spell in a few thousand turns, so they go back and get the
book... only to find out that they "know it quite well already", and
need to wait an indeterminate amount of time until they are on the verge
of forgetting it (< 2000 turns) before the book will let them read it
again.
This commit simply removes that 2000 turn limit, so the player can fully
restore their memory at any time with the spellbook. Naturally, this
still consumes a read charge, so the book won't ultimately last as long
if you keep rereading it early.
If you do have more than 2000 turns left, the game will prompt you to
confirm that you do want to refresh your memory anyway. As before,
rereading with fewer turns will not prompt.
This is an omission in the filled/prefilled unification. The default for
filled on regions now being 0 meant that regions that had previously had
no need for any fill declaration at all (regions' prefilled defaulted to
0 before this, the effect being to fill them) now failed to get filled.
The rule of thumb is that all des.regions with a type for which filled
is meaningful (e.g. special rooms) should declare the fill status. I
added it to a bunch of temples even though this doesn't really seem to
affect anything there (the priest and altar come with the altar
definition). I assigned temples filled=1 and filled=2 loosely based on
if there is ever being some other generation that would put other
furniture or items in a temple, but the distinction should not affect
anything right now.
Cases fixed where non-temple regions weren't getting filled:
- Barracks, a graveyard, and shops in Tou-goal
- The beehive in the Wizard's Tower
This unifies the two separate special-room-stocking code paths, one in
the standard dungeon generator and one in the special level generator
(neither of which reacted to themed rooms, which is the reason for this
commit) into the end of makelevel(), placing the special room stocking
as the very last step of level creation.
Under the new system, when a regular or special level decides to create
a special room, it sets that room's rtype, but the room is not stocked
until later. It already worked this way for special levels, so the main
difference here is in the normal level generation, where the mkroom
family of functions identifies and marks a room as a special room, but
stops short of filling it. (I suppose perhaps the mkroom, mkzoo, mkshop
family of functions would be better off changing their names to
"pickroom" and so on.)
This also restructures makelevel() itself a bit, but the only real
change is that the paths that call makemaz don't return immediately
afterward; they continue to the special room stocking code. Also, this
code was lifted from fill_special_rooms, which is now not used
anywhere, so it has been deleted.
I don't really like how fill_ordinary_room is in mklev.c and
fill_special_room is in sp_lev.c; they seem like they'd be better off in
mkroom.c, but in the interest of not making unnecessary code changes,
I'll just recommend it.
The plan is to unify special room filling code and cause special rooms
to be filled as the very last stage of level creation. Since this will
occur after fixup_special, it was necessary to address the one remaining
piece of code in there that affects special room filling. (The Medusa
code remaining in there doesn't have to do with special rooms.)
There is code in fixup_special for stocking Medusa's lair with statues
of players from the leaderboard. It makes two assumptions: that there
will always be at least one room defined on Medusa's level, and that
the statues should be placed in the first room defined. In the process
of removing prefilled, some of these rooms suddenly became non-rooms,
and this caused problems. This commit ensures that the regions for
turning into rooms to hold the statues are present and come first.
In the process of writing this commit, I discovered a bug: the statue
stocking code for medusa in fixup_special naively chooses the spot at
which to place its final statue by selecting independent x and y
coordinates with somex and somey. This is responsible for a statue
occasionally being embedded in a wall or in iron bars on medusa-2 and
medusa-4: the rooms defined to receive statues are irregular, and some
of the possible coordinates happen to be walls, bars, and water.
The proper fix here is to add lua functionality so that the level
designer can specify that they want a leaderboard corpse or statue, and
remove the medusa special case from fixup_special, but that's rather
out of scope for what I'm doing here.
The fill_special_rooms function was only stocking two types of rooms:
top-level rooms in g.rooms, and their immediate subrooms. If there were
a special room 2 or more levels down, it would not get filled. (No
special levels currently define such a special room, so this bug is
latent.)
To address this, I changed fill_special_rooms to iterate only through
the top level rooms, and fill_special_room to recurse through all of its
subrooms (if any) before filling itself.
The existing system was a confusing mess of competing names (filled,
needfill, prefilled, etc) that had varying semantics, with prefilled
being the worst offender as it meant at least three different things in
various contexts. This commit unifies everything in the code under
"needfill", and everything in Lua under "filled", which defaults to 0
everywhere.
This also removes the second argument to fill_special_room; that
function now just checks the needfill of the room it's passed. As
before, a filled == 2 value is used for a special room to indicate that
the room should set the appropriate level flag, but shouldn't actually
be stocked with anything (for instance, King Arthur's throne room); the
difference is that this now comes directly from the lua script instead
of being manipulated within sp_lev.c.
The prefilled argument had one use case that is occasionally used in the
level files: if the level designer had specified an ordinary region with
prefilled = 1, it would become a room to control monster arrivals on a
level -- monsters that arrive within the bounds of a room are supposed
to stay there.
However, not all of the places where the comments indicated this was
being used were using it correctly; I tested this by letting a few
monsters fall through the knox portal (they're supposed to be
constrained to the entry room) and waiting a hundred turns, then going
through the portal; they were not constrained to the room and had
"wandered" through its walls.
Instead of trying to maintain this special case, I have added an
optional "arrival_room" boolean argument to des.region, which forces it
to create a room for the purposes of constraining monster arrival.
I have gone through and replaced occurrences of prefilled in lua files
with the appropriate filled option (or arrival, as needed). In some
cases, that resulted in questionable regions such as a filled ordinary
area in a non-themeroom (I just dropped the filled=1), or an area which
didn't do anything, not even lighting (which I deleted).
No code in this function would actually do anything if it were called on
a themed room, but since it is pretty clearly intended for the "regular"
special rooms, it's probably best to explicitly avoid themed rooms as
well.
I noticed that any subrooms created within a themed room were bare -
they never had any monsters, objects, traps, or anything really,
regardless of whether filled = 1 was set on them. As a result, they're
pretty boring.
It turns out that the code in makelevel() responsible for stocking
ordinary rooms with stuff only looped through g.rooms, and completely
ignored subrooms. (Subrooms would not get stocked with items by virtue
of being part of the larger room; I tested this by dialing the item
generation in rooms way up, and none of those items ever got placed in a
subroom.)
To fix this, I've extracted the code that populates an ordinary room
into its own function, fill_ordinary_room, and made it recurse into its
own subrooms. (I also renamed fill_rooms and fill_room to include the
word "special" in their names, because they only deal with special
rooms.) Note that since special rooms follow a separate codepath, an
ordinary subroom of a special room won't get stocked; perhaps these
functions should be unified in the future.
The fill_ordinary_room code is pretty much a verbatim cut and paste from
makelevel, so there is not currently any consideration for the size of
the subroom or the fact that it is a subroom with respect to how many
monsters, traps, objects, etc get placed.
I'm not sure whether other things such as stair selection will ever
select themed room subrooms, or whether they too only look at g.rooms.
This commit is intended to fix the bug where a pet will get fixated on
an unmoving monster and stop moving itself. I described the cause in the
github issue; the gist is that the pet AI chooses the unmoving monster
as its ranged target, doesn't do anything when it calls mattackm
(because it doesn't have ranged attacks), then returns a value
indicating it didn't move and can't take further actions.
I initially implemented a fix that refactored mattackm to distinguish
between "attacker missed" and "attacker did nothing", which the pet AI
could then use to determine whether the pet could continue doing things.
But then I realized that if mattackm is called with non-adjacent
monsters, a return of MM_MISS more or less unambiguously indicates that
the attacker did nothing (because the ranged functions it calls like
breamm don't actually check to see whether the target was hit, just
whether the monster initiated the attack.) So, this only really needed
to check whether mattackm returned with MM_MISS.
I also found a probable bug in mattackm, in that the thrwmm call isn't
treated the same as breamm or spitmm. In the latter two, mattackm
returns MM_HIT even though it doesn't check whether the ranged attack
actually hit its target. But there was no logic doing the same for
thrwmm, so this commit also adds that. (Otherwise, a pet could possibly
use a ranged weapon attack and then get to keep moving on its turn.)
Some changes to the YnDialog widget used when popup_dialog is On.
If a button is labelled with a space, it just looks like an
unlabelled button. Switch to "Spc" for space, "Ent" for \n and
"Ret" for \r. (The last two aren't completely logical but I
haven't seen any dialogs that need them and they'll be better
than "^J" and "^M" if there are such.)
For yn#aq dialogs, preload a grayed-out "#" in the count widget.
Just for show; has no tangible effect.
The count widget should return long rather than plain int.
The short exclamations ("Gasp!", "Why?", &c) led to ambiguity
about which monster was vocalizing them. Use full sentences
which refer to the speaker. It can become quite a bit more
verbose but is less likely to lead to confusion. Perhaps it
should cut those off after a modest number of them have been
issued?
"When a room is created and passed down to a contents function in
Lua, the width and height properties of that room are computed by
subtracting lx from hx and ly from hy, which means e.g. a room
which is 8 floor squares wide and 5 tall appears to the contents
function as having a width of 7 and height of 4. This patch fixes
that off-by-one."
I don't understand the details here: should a room's dimensions
include its boundary walls or just the inner amount? This change
didn't seem to cause any problems so I've put it in.
Closes#345
Replace the blank placeholder icon with individual placeholders
for Stone, Slime, Strngl, Deaf, Lev, Fly, and Ride. They're just
40x40 tiles showing solid color (different for each) holding white
block letters spelling the condition. For the first four of those,
the text runs from upper-left to lower-right, for Lev and Fly the
text runs from lower-left towards upper-right, and for Ride it's
horizontal. Not particularly exciting but better than blank. We
still need real artwork to make them be similar to the older
conditions.
Also moves the two petmarks and the pilemark from qt_xpms.h to
qt_map.cpp. The marks and the assorted status icons are all
static arrays, and including that header in two source files
meant that they were all duplicated unless the compiler or linker
was smart enough to discard the unused ones.
For Qt, experience points weren't shown when enabling 'showexp'
option because they were conditional upon '#if EXP_ON_BOTL'. That
got eliminated prior to 3.6.0 so wasn't defined for qt_stat.cpp.
When displayed, show Exp as Level:Xp/Exp instead of as a separate
status field. This has the intentional side-effect of omitting it
when hero is polymorphed and status shows HD instead of Xp.
Label the six characteristics in mixed case instead of all upper
case: Str, Dex, and so forth.
As reported in https://github.com/NetHack/NetHack/issues/391
if make was invoked with -j, makedefs instances could end up running in
parallel and could trample on each other's grep.tmp tempory files.
Default to using mkstemp(); allow a port runtime library implementation
that lacks mkstemp() to define HAS_NO_MKSTEMP to revert to the old behaviour.
Provide a work-alike mkstemp() implementation for windows Visual Studio build
in mdlib.c so there is no requirement to define HAS_NO_MKSTEMP there.
Fixes#391
The code for peaceful monsters witnessing the hero attack another
peaceful monster and getting angry had a 20% of making them gasp in
surprise or exclaim "why?" in shock. It was only requiring them to
have humanoid shape rather than checking for speech capability, so
peaceful zruty or minotaur, possibly other animals, could exclaim
comprehensibly. Other things which shouldn't talk, like mummies,
would behave similarly.
This categorizes how a bunch of MS_foo types should react. It has
only been lightly tested.
Any key bindings in player's run-time config file will already be
in place by the time the Qt menus are constructed. Those menus
will use the new key assignments rather than the defaults, so not
a bug.
The previous teleport scroll fix was mislabeled with this pull
request number. Too late to fix that now; should have been
Closes#307
Now... Interaction between voluntarily busy hero (resting,
searching, and so on) with approaching monsters to decide whether
to stop had some inconsistencies.
Really closes#386
Since teleporation gives a "you matrialize" message even when
arriving close by, the old behavior of not learning a scroll of
teleportation when you land quite close to your original spot
no longer made sense. Always [almost] discover teleport scroll
when reading it.
Also adds one-shot teleport control when reading a blessed scroll
of teleportation. I changed that to be prevented when hero is
stunned, same as with full-fledged teleport control.
I reworded or reformatted several of the comments. And removed
the EDITLEVEL increment in patchlevel.h; save and bones file
contents are not affected.
I've also added an unrelated comment about reading mechanics to
doread().
Closes#386
The #enhance menu revealed a couple of menu problems for Qt.
Items flagged with "*" or "#" were showing tiny "..." instead of
the flag character. An existing problem rather than something
caused by yesterday's overhaul patch.
The "(Skills flagged by "*" may be enhanced when you're more
experienced.)" legend line was causing the regular entries to be
formatted strangely (their skill name column was much too wide).
That was caused by me dropping something (special case for header
lines during tab-separation handling) in yesterday's patch that
I mistakenly thought wasn't needed.
The save file selection widget was issuing a complaint to stderr:
|QLayout: Attempting to add QLayout "" to QDialog "", which already
| has a layout
This shuts that up, but doesn't fix the broken save file selection
so I haven't added a fixes37.0 entry.
handle preselected item in pick-one menu; picking it returns that
item rather than toggling it off and returning nothing, picking
something else only returns the other thing (was returning first
of the chosen item or the preselected item, foiling core's attempt
to deal with both and giving wrong result whenever the preselected
one came first--like pick-an-attribute for menu colors);
when handling typed input, check selector letters before menu
command keys so that special "letters" '-' (fingers, hands, self)
and ':' (look inside container) that are specified by a few menus
can be chosen by keyboard;
menus were using default line heights which are excessively tall,
effectively making them be double spaced and using more screen
space than should have been needed; reduce height to 60% of what
it was, still a bit taller than regular spacing; look at ^X--which
is rendered via menu--before and after to see the difference;
start with count column empty instead of 6 spaces; grow it as counts
get entered; reset to empty if [all], [none], or [invert] is used;
treat intermediate counts as long rather than int; right justify
formatted count values;
simplify creating menu return data (pick-one doesn't need separate
handling);
for pick-one menus,
enable [ok] button if there is one preselected item,
enable [all] button if there is only one item (may never happen),
enable [none] if there is a preselected item (menu remains active
if [none] is used to clear the preselection);
enable [invert] if there is one item (may never happen; should
allow two items if one of them is preselected--definitely does
happen--but that wouldn't work as intended without code changes);
honor pending count if an item is selected by clicking its checkbox
(already done for typing its letter or for clicking another part
of item's menu line);
accept <delete>/<rubout> in addition to <backspace> when backing out
a digit as a count is being typed;
accept ^[ as well as ESC key for cancelling count or entire menu;
honor 'menucolors'=false to ignore any defined menu color patterns.
Record a bunch of Qt stuff before I forget it all. It's probably
too late: I'll bet I've left some things out.
Doesn't include several menu problems that I've already fixed but
not checked in yet. (I don't have any other Qt changes pending.)
Qt's map hadn't been updated to draw unexplored locations with
the unexplored glyph so was still using solid stone instead.
Column 0 should be removed but I'll leave that for someone more
adventurous; I did it for curses and for X11 but am going to
pass here. It's very noticable after magic mapping but is only
"bad" (by wasting space) if clipping is being performed.
Qt's text window for 'history' was unexpectedly wide and it turned
out that that was to fit in one excessively wide line. tty has
been formatting it by splitting it into one normal sized line
followed by an excessively short line.
Similar to how the pick-an-attribute menu for menu colors and
status highlights shows the attribute names using the attribute
so that you can see how it looks (or whether it is supported),
have the pick-a-color menu show the color names in the
corresponding color. Does so by temporarily removing any
user-specified menu colors and setting up another list of such
for matching color names.
Forces the 'menucolors' option On while the pick-a-color menu is
in use, then restores the previous setting along with the user's
menu colorings. Might need some way to avoid setting that for a
configuration where colors don't work.
Avoid use of GNU make's 'override' feature by requiring
'make CCFLAGS=-O' to replace -g from the make command line instead
of 'make CFLAGS=-O'. Note the extra 'C' in the spelling.
Revert the previous umpteen MORECFLAGS+= back to normal CFLAGS+=.
