nullcgt / splicehack-rewrite Goto Github PK

A rewrite of SpliceHack.

Perl 0.48% Lua 5.70% CSS 0.02% C 88.11% C++ 4.31% Shell 0.16% Assembly 0.25% OpenEdge ABL 0.03% GAP 0.01% sed 0.01% R 0.10% Max 0.01% Makefile 0.19% Batchfile 0.07% Awk 0.05% Roff 0.24% DIGITAL Command Language 0.23% AppleScript 0.04%

splicehack-rewrite's Introduction

Kestrel Gregorich-Trevor

🔭 I’m currently working on a number of small projects, only some of which are on GitHub.
🌱 I’m currently learning vim shortcuts.
💬 Ask me about Roguelikes!
😄 Pronouns: She/Her/Hers!
⚡ Opinions expressed here are my own, and do not reflect the views of any company or organization that I am employed by or have been previously employed by. Additionally, all public repositories found below are not managed or the property of any company or organization, current or previous.

Reference

🖌 Dotfiles: A repository for my personal dotfiles. Useful if you want to make your desktop look like mine.

Hobbyist Projects

🧪 Mutagen: My entry in the 2017 Fighting Game AI Competition. Placed 4th overall.
🛡 SpliceHack: A fork of NetHack. Actively developed, with an active playerbase.
🛡 Zenzizenzizenzic: A traditional roguelike built from scratch in C.

splicehack-rewrite's People

Contributors

Stargazers

Watchers

Forkers

rojjacebolla actual-nh umbire

splicehack-rewrite's Issues

Replace all &mons[] array accesses with checks against mondata->omnum.

So far, I've replaced the chcecks in mondata.h, but for the sake of completeness and fixing bugs, all instances of this should be changed.

Cannot wish for specific types of blood.

This is clearly a wish parser issue. I believe one of the dnethack devs has a fix for it.

Last call: Name of repo/wiki page/abbreviations

Last call to rename freely! This is the moment when DNH wound up committed to the "dNAO" repo instead of the more fittingly-named "dnethack" repo; are lost DNH players out there, blindly playing from the wrong repo? This is the moment that doomed UnNetHack to an unsearchable forked repo; were contributors lost due to inconvenience?

With only 4 watchers/stars and no server host, "spl-r" can be moved. Renamed. Or even merged in as a new branch on the old repo, with no shared commit history to the other branches.

I hope that SpliceHack-Rewrite isn't the final name. I haven't come up with any wonderful replacements. But for lack of something brilliant, we could use:
FreshSplice. NewSplice. NiceHack. NetKes. PartyHack. Something else... or even simply SpliceHack.

(writing this as an "issue" just so that it can be a Project card)

Failed monster sanity checks exposed by fuzzer.

This is a very persistent bug, and I have not been able to discover the source. It occurs fairly frequently when fuzzing, and yet I've never found it during regular gameplay.

Sample output:


Suddenly, the dungeon collapses.
illegal mon data 555555a03568; mnum=317 (fmon)
  Generating more information you may report:

[00] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(+0xea318) [0x55555563e318]
[01] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(+0xeaad1) [0x55555563ead1]
[02] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(panic+0x26e) [0x555555640ea8]
[03] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(+0x15fb30) [0x5555556b3b30]
[04] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(mon_sanity_check+0x6c) [0x5555556b44b0]
[05] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(sanity_check+0x23) [0x5555555e000a]
[06] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(moveloop+0x11d9) [0x5555555b312e]
[07] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(main+0x8b2) [0x5555557eb29b]
[08] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf3) [0x7ffff7c340b3]
[09] /home/username/Workspace/SpliceHack-Rewrite/src/splicehack(_start+0x2e) [0x5555555b1c8e]

Program received signal SIGABRT, Aborted.
__GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
50	../sysdeps/unix/sysv/linux/raise.c: No such file or directory.
(gdb) bt
#0  __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
#1  0x00007ffff7c32859 in __GI_abort () at abort.c:79
#2  0x000055555563eac8 in NH_abort () at end.c:206
#3  0x0000555555640ea8 in panic (str=str@entry=0x555555876918 "illegal mon data %s; mnum=%d (%s)") at end.c:705
#4  0x00005555556b3b30 in sanity_check_single_mon (mtmp=mtmp@entry=0x555555b2ab00, chk_geno=chk_geno@entry=1 '\001', msg=msg@entry=0x555555833eb9 "fmon") at mon.c:63
#5  0x00005555556b44b0 in mon_sanity_check () at mon.c:205
#6  0x00005555555e000a in sanity_check () at cmd.c:3012
#7  0x00005555555b312e in moveloop (resuming=<optimized out>) at allmain.c:493
#8  0x00005555557eb29b in main (argc=<optimized out>, argv=0x7fffffffe658) at ../sys/unix/unixmain.c:335

Any help with this issue would be greatly appreciated.

Object materials detection bug

See k21971/EvilHack#67 - this also affected xNetHack and this variant. Easy fix.

Change flanking code to allow for players to flank monsters with their pets.

Adding this card for tracking purposes, even thought it is mostly complete.

Moldiers and other infesting monsters spawn adjacent to the player, rather than the infesting monster.

This isn't a bug so much as it is a behavioral issue. We previously spawned the monster at the exact location of the infesting monster, but that was causing it to get clobbered.

Can likely look to the revive code for inspiration.

just jotting down this important looking warning before i go to sleep

this seems important:

makemon.c: In function ‘m_initweap’:
makemon.c:648:9: warning: this ‘if’ clause does not guard... [-Wmisleading-indentation]
  648 |         if (!rn2(4))
makemon.c:650:13: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the ‘if’
  650 |             break;

and this attempt to block exploders from becoming explosive, looks like valid syntax based on other things in this file. but it isn't:

makemon.c: In function ‘is_valid_template’:
makemon.c:2805:21: error: ‘struct monst’ has no member named ‘mlet’
 2805 |         return !mtmp->mlet == (S_EYE || S_LIGHT);

Phoenix Egg revival via fire and lava does not work.

Phoenix eggs revive very quickly, but the code forcing revival via exposure to extreme heat does not work.

Hand Grenade Explosions are sometimes the wrong type.

This issue occurs when a monster throws a frag grenade. The explosion type should be fiery, but instead it appears as magical. This issue is not present whenever the player throws a grenade.

The issue is fairly clearly the fault of the grenade_explode code in zap.c, reproduced here:

/*
 * Note: obj is not valid after return
 */
void
grenade_explode(struct obj *obj, int x, int y, boolean isyou)
{
    int ztype;
    int otyp = obj->otyp;
    int yours = isyou ? 1 : -1;
    boolean save_mon_moving = g.context.mon_moving;

    g.context.mon_moving = yours ? FALSE : TRUE;

    if (obj->oartifact == ART_HAND_GRENADE_OF_ANTIOCH) {
        ztype = isyou * ZT_SPELL(ZT_MAGIC_MISSILE);
        explode(x, y, ztype, d(50,6), WEAPON_CLASS,
            isyou * -1 * EXPL_FIERY);
    } else if (otyp == FRAG_GRENADE) {
        ztype = isyou * ZT_SPELL(EXPL_FIERY);
        explode(x, y, ztype, d(3,6), WEAPON_CLASS,
            isyou * -1 * EXPL_FIERY);
    } else if (otyp == GAS_GRENADE) {
        ztype = isyou * ZT_SPELL(ZT_POISON_GAS);
        explode(x, y, ztype, d(3,6), WEAPON_CLASS,
            isyou * -1 * EXPL_NOXIOUS);
    }
    g.context.mon_moving = save_mon_moving;
    wake_nearto(x, y, 400);
}