Bugs and complaints on current master

[Nick]

I realized if I remove the curse on my ring of teleportation I will have a ring of speed <+2> so I found a ?Remove Curse in the dungeon and tried to remove the Teleporation curse (power level 100) from the ring of Teleportation, but I could not get the scroll to activate. When I read the scroll, it would give me the option to remove the curse from the ring. I'd select the ring and then it would ask me if I want to remove the curse from the ring? so I'd hit <Enter>, but nothing would happen. I couldn't get the scroll to be read. I tried all the letters on the keyboard as a response in case there was some visual bug, but nothing worked. I simply was unable to read the scroll. And so my ring of Tele is still cursed, and somewhere on the dungeon floor is a scroll of remove curse that I wasn't able to successfully read.

Curse power 100 means it's permanent. So not letting you read the scroll is correct, but it really should just not even give the ring as an option.

[PowerWyrm]

Yeah, I did a different hack this time round, and rounded up the value of those things. The calculations are changed a bit, mainly in armor and ability bonuses, where I approximated the existing method in a way that was more amenable to going into the edit file.

FYI, I was playing around with the values for ability bonuses and compared the results with the artifact set:
- total modifier bonus of -70 and modifier bonus divisor 16 (current values in object_power.txt): delta around -600
- total modifier bonus of -39 and modifier bonus divisor 19: delta = 18 (the lowest value I could get)

[Nick]

FYI, I was playing around with the values for ability bonuses and compared the results with the artifact set:
- total modifier bonus of -70 and modifier bonus divisor 16 (current values in object_power.txt): delta around -600
- total modifier bonus of -39 and modifier bonus divisor 19: delta = 18 (the lowest value I could get)

Nice

I'll use that at some point.

[PowerWyrm]

Not sure, but I think commit 867258e introduced a bug:

static int object_power_calculation_FLAG_POWER(void)
{
	struct obj_property *prop;
	prop = lookup_obj_property(OBJ_PROPERTY_FLAG, iter);
	assert(prop);
	return of_has(power_obj->flags, iter + 1) ? prop->power : 0;
}

Since iter has been rebased to 1..OF_MAX, this is probably:

static int object_power_calculation_FLAG_POWER(void)
{
	struct obj_property *prop;
	prop = lookup_obj_property(OBJ_PROPERTY_FLAG, iter);
	assert(prop);
	return of_has(power_obj->flags, [U]iter[/U]) ? prop->power : 0;
}

[Grotug]

I read a ?RR on DL99 (forced descent off) and then descended a magical staircase. The game then crashed.

Macbook Air early 2015. Yosemite 10.10.5

I had saved the game right before reading said scroll of recall, so I read the scroll and descended and the game loaded DL100 just fine. So the bug was not repeatable.

[Grotug]

Did not find any scrolls of identify for about 40 dungeon levels (from about DL85 if not earlier to DL127). Has their rarity been adjusted in the Nightly versions?

[Nick]

Did not find any scrolls of identify for about 40 dungeon levels (from about DL85 if not earlier to DL127). Has their rarity been adjusted in the Nightly versions?

Yes, they're rarer.

[Nick]

New build up on the nightlies page that fixes some issues with the randart saving system. Here's how it now works:

1. If you choose the randarts birth option, a new file called randart.txt will be written into your user directory
2. This file is used to read the randarts until your character dies, when it will be moved into your (new) user/archive directory, and renamed randart_1.txt (then randart_2.txt, etc)
3. The option for keeping previous randarts at birth is gone. If you wish to use a previous set of randarts, copy it from your archive directory back into your user directory, and rename it artifact.txt (not randart.txt).

This system is not perfect; in particular, things will go wrong if you try to play two different randart characters at once. Otherwise it should work for now, and I will fix that later.

[PowerWyrm]

Did not find any scrolls of identify for about 40 dungeon levels (from about DL85 if not earlier to DL127). Has their rarity been adjusted in the Nightly versions?

Allocation range for scrolls of identify is 5-70, so you won't find any past DL70.

[Grotug]

Allocation range for scrolls of identify is 5-70, so you won't find any past DL70.

Is this... good? I suspected equipment I was finding had pStun but I was never able to confirm it because I couldn't find any scrolls of ID.

I love the new randart archive system.

[kandrc]

Here's a new one in 4.0.3-606-gec71e34:

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff6eb06a1 in __strlen_sse2_pminub () from /lib64/libc.so.6
(gdb) bt
#0  0x00007ffff6eb06a1 in __strlen_sse2_pminub () from /lib64/libc.so.6
#1  0x00000000004e5bdf in my_strcpy (buf=0x7fffffffbb50 "DRAIN_CHARGES", 
    src=0x150 <Address 0x150 out of bounds>, bufsize=1024) at z-util.c:174
#2  0x00000000004e2dbc in vstrnfmt (buf=0x7fffffffc480 "drop-artifact:", 
    max=1024, fmt=0x4fb2c6 "drop-artifact:%s\n", vp=0x7fffffffc8a0)
    at z-form.c:496
#3  0x00000000004e2112 in file_vputf (f=0xafa6a8, 
    fmt=0x4fb2c6 "drop-artifact:%s\n", vp=0x7fffffffc8a0) at z-file.c:642
#4  0x00000000004e20b0 in file_putf (f=0xafa6a8, 
    fmt=0x4fb2c6 "drop-artifact:%s\n") at z-file.c:624
#5  0x0000000000448ba3 in write_lore_entries (fff=0xafa6a8) at mon-lore.c:1545
#6  0x00000000004e5646 in text_lines_to_file (
    path=0x7fffffffd2d0 "/home/sheaffer//.angband/Angband/lore.txt", 
    writer=0x448735 <write_lore_entries>) at z-textblock.c:714
#7  0x0000000000448e39 in lore_save (name=0x513155 "lore.txt")
    at mon-lore.c:1611
#8  0x00000000004aed04 in save_game () at ui-game.c:516
#9  0x00000000004aca88 in new_level_display_update (
    type=EVENT_NEW_LEVEL_DISPLAY, data=0x0, user=0x0) at ui-display.c:2183
#10 0x000000000041b1bf in game_event_dispatch (type=EVENT_NEW_LEVEL_DISPLAY, 
    data=0x0) at game-event.c:43
#11 0x000000000041b472 in event_signal (type=EVENT_NEW_LEVEL_DISPLAY)
    at game-event.c:142
---Type <return> to continue, or q <return> to quit---
#12 0x000000000041d0a8 in on_new_level () at game-world.c:710
#13 0x000000000041d4e6 in run_game_loop () at game-world.c:841
#14 0x00000000004aeb0e in play_game (new_game=false) at ui-game.c:433
#15 0x00000000004e6ebf in main (argc=1, argv=0x7fffffffdcd8) at main.c:524
(gdb)

[kandrc]

Same version, text in game:

Something makes a high-pitched shriek.
It looks even faster!

[kandrc]

Another infinite loop:

(gdb) bt
#0  scatter (c=0xb9edf8, yp=0x7fffffffd760, xp=0x7fffffffd75c, y=64, x=107, d=1, need_los=true) at cave.c:293
#1  0x0000000000431185 in vault_monsters (c=0xb9edf8, y1=64, x1=107, depth=98, num=2) at gen-util.c:731
#2  0x000000000042ab54 in build_room_template (c=0xb9edf8, y0=58, x0=107, ymax=10, xmax=17, doors=3, 
    data=0x9cb898 '#' <repeats 18 times>, ".x.#9x.#.x9#.x.##...#...#...#...##123#321#123#321#%", '.' <repeats 15 times>, "%%", '.' <repeats 15 times>, "%#321#123#321#123##...#..
.#...#...##.x.#9x.#.x9#.x.", '#' <repeats 18 times>, tval=0) at gen-room.c:1010
#3  0x000000000042add9 in build_room_template_type (c=0xb9edf8, y0=64, x0=188, typ=1, rating=1) at gen-room.c:1074
#4  0x000000000042e7a8 in build_template (c=0xb9edf8, y0=64, x0=188, rating=1) at gen-room.c:2384
#5  0x000000000042fedb in room_build (c=0xb9edf8, by0=0, bx0=0, profile=..., finds_own_space=true) at gen-room.c:2990
#6  0x0000000000423cec in modified_chunk (depth=97, height=64, width=188) at gen-cave.c:1688
#7  0x0000000000424260 in modified_gen (p=0x9ca2b8) at gen-cave.c:1791
#8  0x000000000041f2d1 in cave_generate (c=0x74d320 <cave>, p=0x9ca2b8) at generate.c:864
#9  0x000000000041d45a in run_game_loop () at game-world.c:840
#10 0x00000000004af233 in play_game (new_game=false) at ui-game.c:433
#11 0x00000000004e75ed in main (argc=1, argv=0x7fffffffdc48) at main.c:524
(gdb)

always fails the check for !square_in_bounds_fully() in scatter().

[Nick]

Thanks for those.

First crash is probably the new randart code, second one is like the bug PowerWyrm found recently. Will try and chase them down this weekend.

[Ingwe Ingweron]

Is this... good? I suspected equipment I was finding had pStun but I was never able to confirm it because I couldn't find any scrolls of ID.

I love the new randart archive system.

yeah, I reported this issue as a bug some time ago, but I don't think it was ever fixed, that or maybe Nick didn't think it was a bug. Personally, I feel that ?ID should be able to be found all the way down to the level just this side of hell (DL127), even if they are rare.