Bugs and issues in 4.1.2

[Nick]

Nick, did the disconnected map I posted about earlier check out?

Thanks for the reminder. It is indeed disconnected (it's a "moria" level, new in 4.1), and I've filed that as a bug.

[geoff_tewierik]

Nick, did the disconnected map I posted about earlier check out?

[Nick]

Would you consider providing these two features?:

1. Showing the "slowed" status of a monster (c.f., color on monster health bar and stating the monster is "slowed" when “looking” at or “targeting” the monster (c.f., confusion and stunning)?

2. Separating alchemist gloves of slaying and power from regular slaying or power gloves for purposes of ignoring. Classes with CUMBER-GLOVE could then safely ignore those on penalized gloves without the risk of ignoring them on alchemist gloves which they could use without penalty.

Yes to both; the first actually needs some work to make the whole thing a bit less ad hoc in the code, and the second probably along the lines that Estie is suggesting.

[Estie]

2.) should probably be done in line with great weapons and dragon armor by putting them in a special class ("splendid handwear" or so, including alchemist gloves and caesti).

[Ingwe Ingweron]

Would you consider providing these two features?:

1. Showing the "slowed" status of a monster (c.f., color on monster health bar and stating the monster is "slowed" when “looking” at or “targeting” the monster (c.f., confusion and stunning)?

2. Separating alchemist gloves of slaying and power from regular slaying or power gloves for purposes of ignoring. Classes with CUMBER-GLOVE could then safely ignore those on penalized gloves without the risk of ignoring them on alchemist gloves which they could use without penalty.

[Derakon]

That can happen with very weak artifacts (e.g. Elvagil or one of the Paur* gauntlets) if the randart rolls low on its power variance (so even weaker than the standart) and then spends its power on to-hit/dam instead of abilities.

The Glaive of Pain is a standart that doesn't have any powers, but nobody complains because it's actually useful.

[Sphara]

I don't know if this is a bug even an issue to report.
Nevertheless, playing 4.1.2 on angband.live, I found an artifact that has no powers or abilities whatsoever.
Well, its indestructible. Guess that counts as a power


the Trident of Imlas (1d10) (+5,+12)
Dropped by Sangahyando of Umbar at 1250 feet (level 25)

Cannot be harmed by acid, fire.

Combat info:
2.0 blows/round.
With +2 STR and +0 DEX you would get 2.3 blows
With +0 STR and +2 DEX you would get 2.3 blows
Average damage/round: 40.

Clears rubble in 13 turns, magma veins in 80 turns, and doesn't affect quartz veins.

[Ighalli]

Is the competition on one of the older versions? There was a bug for a while where waybread's heal didn't work. I think that was fixed several versions ago; at least, waybread works for me in 4.1.2.

If not, I'm guessing that the randart somehow pulled in whatever the waybread code was before the bugfix. Which would be a bug in randart generation that needs fixing.

It's the competition code, but it has all the commits that master does, so there must be a regression in the magic branch. Or a problem when I compile it.

[luneya]

I'm playing a randart game for the competition, and I found an amulet with this power:

When activated, it restores 4d8 hit points and neutralizes poison.

I've seen it heal 2 hp and 0 hp on activations since I was watching closely.

I found it in the artifacts.txt:


namef Bereth
base-object:amulet:Pendant
graphics:"ark
info:20:3:28000
alloc:19:29 to 79
power:0:0d0:2:7:0
flags:FEATHER
values:WIS[3]
values:RES_ACID[1] | RES_FIRE[1] | RES_COLD[1] | RES_CHAOS[1]
act:FOOD_WAYBREAD
time:32+1d4
desc:Random amulet of power 133


Waybread seems like it should heal 4d8, according to activation.txt:


name:FOOD_WAYBREAD
aim:0
power:4
effect:NOURISH
dice:7500
effect:CURE:POISONED
effect:HEAL_HP
dice:4d8
desc:restores 4d8 hit points and neutralizes poison


If this thing really activates for waybread, it should probably also mention that it nourishes you. Not sure why the healing isn't working as advertised though.

Is the competition on one of the older versions? There was a bug for a while where waybread's heal didn't work. I think that was fixed several versions ago; at least, waybread works for me in 4.1.2.

If not, I'm guessing that the randart somehow pulled in whatever the waybread code was before the bugfix. Which would be a bug in randart generation that needs fixing.

[Ighalli]

I'm playing a randart game for the competition, and I found an amulet with this power:

When activated, it restores 4d8 hit points and neutralizes poison.

I've seen it heal 2 hp and 0 hp on activations since I was watching closely.

I found it in the artifacts.txt:


namef Bereth
base-object:amulet:Pendant
graphics:"ark
info:20:3:28000
alloc:19:29 to 79
power:0:0d0:2:7:0
flags:FEATHER
values:WIS[3]
values:RES_ACID[1] | RES_FIRE[1] | RES_COLD[1] | RES_CHAOS[1]
act:FOOD_WAYBREAD
time:32+1d4
desc:Random amulet of power 133


Waybread seems like it should heal 4d8, according to activation.txt:


name:FOOD_WAYBREAD
aim:0
power:4
effect:NOURISH
dice:7500
effect:CURE:POISONED
effect:HEAL_HP
dice:4d8
desc:restores 4d8 hit points and neutralizes poison


If this thing really activates for waybread, it should probably also mention that it nourishes you. Not sure why the healing isn't working as advertised though.

[PowerWyrm]

Summoning (scroll/staff) now appears to always bring up out of depth uniques. Is that deliberate? These chaps have a tendency to instakill...

Ok I think this comes from the fact that the summon parser has all entries reverted so S_ANY is in fact S_UNIQUE (although scroll of summon undead seems to work fine). Or just off by one. Or using an empty entry zero.

I have implemented the latest changes in PWMAngband while keeping the parsing function similar to others and I don't have this problem.

[Sky]

noo .. don't take away my summons!

[PowerWyrm]

A bug. Commit 9c5b10b, mon-summon.c, summon_specific_okay()

In fact that's a totally unrelated bug. It just allows WRAITH to summon any W, not just unique Ws (Ringwraiths).

[Anthon]

Exploitable bug ignoring non-ID items

If you know many (or all) of the available items from previous runs, you can identify single items by toggling the ignore status of a all items of a particular type via the "item ignoring setup" (= in rogue mode), then select ignore item (^D in rogue mode), select the un-ID item. In its ignore options the b) option will toggle from "b) All" to "b) Unignore all".

The above assumes you did not ignore the unidentified flavor. What b) should always indicate for non-ID-ed items is the ignore status of the flavor.

Details on how this can be exploited:

If you don't want to go through toggling (e.g.) each potion to see what your unidentified flavor matches with, you can just set to ignore anything you don't want to drink up front. Then do not try anything you have not identified, until you checked if you can unignore it (then don't) or not (feel free to try).

Practical example: As a mage I don't ever want to drink PoContemplation. So I would normally buy it when available and then discard (or sell depending on the birth settings). With the exploit, I just set to ignore PoContemplation before ever encountering it. Once I stumble on an un-identified potion, I initiate the ignore-an-item sequence to see if option b) is unignore or not. If it is un-ignore, don't drink it.

[PowerWyrm]

Summoning (scroll/staff) now appears to always bring up out of depth uniques. Is that deliberate? These chaps have a tendency to instakill...

A bug. Commit 9c5b10b, mon-summon.c, summon_specific_okay():

/* A valid base and no match means disallowed */
	while (bases) {
		if (race->base == bases->base) [B]return true[/B];
		if (bases->next == NULL) return false;
		bases = bases->next;
	}

should probably be:

/* A valid base and no match means disallowed */
	while (bases) {
		if (race->base == bases->base) [B]break[/B];
		if (bases->next == NULL) return false;
		bases = bases->next;
	}

to keep the old behavior.