Skill Bug -- Land Mine

[Andrue]

Roar. :o

There's a pretty nasty bug that's been around for ages. The issue is with the effective range of the Land Mine trap skill. Land Mine, like other traps, has a trigger radius of 3x3, but the damage radius is only 1x1. This means that Land Mine will always fail to damage (assuming that an enemy/monster is walking towards the trap.)

Lovely visuals (with colors!):
OOO
OOO
OOO

Green: Trigger Range
Red: Damage Range

I'm uncertain on the official mechanics behind the skill... I presume that if the trap has a trigger radius of 3x3, it should also have a damage radius of 3x3.

[Koma]

Oddly enough, I was a Land Mine hunter back during the iRO beta and do not recall it ever missing in the long hours I used it on Stormy Knight MVPing. We'll check into this but if anyone has any other details on this, please let me know.

Apparently, this is related to an almost year old change (October 24th, 2008) and also an older one:

Revision: 13323
Author: L0ne_W0lf
Date: 9:26:14 PM, Friday, October 24, 2008
Message:
* Just some random updates and bugfixes!
- Corrected Landmine's placement range. (bugreport:2305)
- Corrected comment in merc skill db. (bugreport:2324)
- Corrected the casting range for traps. (Based on visial comparison from iRO)
----
Modified : /trunk/db/Changelog.txt
Modified : /trunk/db/skill_db.txt
Modified : /trunk/db/mercenary_skill_db.txt
Modified : /trunk/db/skill_unit_db.txt

Revision: 12180
Author: skotlex
Date: 4:07:47 PM, Wednesday, February 06, 2008
Message:
- Fixed Freezing Trap doing no damage (changed type from misc to weapon)
- Changed suiton's element to water (apparently fixes not being able to use water-based skills on it)
- Fixed Landmine doing no damage (decreased trigger range to 0 to match its splash range)
- Fixed the disguise commands using the wrong variable when attempting to disguise as an npc.
- Fixed a compilation warning on CELL_CHKNODAMAGE (which is a totally bad-ripoff of basilica cells anyway)
- Corrected and optimized npc_remove_map's npc array cleanup (@reloadscripts no longer report 'too many npcs per map and related dangling pointer crashes are fixed)
- Removed suspicious, yet totally unused npc_data variable 'n'
- HP loss item scripts can no longer kill you.
- Corrected the item-drop-rate check to avoid duplicate entries in @whodrops after a @reloadmobdb
- Cleaned up and optimized map_addnpc (there are never any gaps in the npc array so npc_num always points to the last valid entry)
----
Modified : /trunk/Changelog-Trunk.txt
Modified : /trunk/src/map/pc.c
Modified : /trunk/src/map/atcommand.c
Modified : /trunk/src/map/map.c
Modified : /trunk/db/Changelog.txt
Modified : /trunk/src/map/battle.c
Modified : /trunk/src/map/mob.c
Modified : /trunk/src/map/map.h
Modified : /trunk/src/map/npc.c
Modified : /trunk/db/skill_db.txt
Modified : /trunk/db/skill_unit_db.txt
Modified : /trunk/src/map/charcommand.c

[Andrue]

Actually, those fix reports contain some information that helps, I suppose. If the splash range is truly a 1x1 range, then the trigger range should be the same, too.

The issue with trying to make this change in the skill_unit_db.txt would require setting the range to 0, which means the traps could be placed side by side. To keep the trap placement(every other cell), but reduce the trigger range, you can modify skill.c:

Find(Line 6801; part of struct skill_unit_group* skill_unitsetting):

Code: Select all

layout = skill_get_unit_layout(skillid,skilllv,src,x,y);

Below this, add:

Code: Select all

if(skillid == 116)range=0;

So this will keep trap spacing, but reduce the trigger range of Land Mine to the exact cell that the trap is on. I actually posted this fix over at eAthena before.. but it was never picked up. Perhaps a lack of caring.

Once again, this is if we assume that the splash range of Land Mine is 1x1. Is there anyone that can test this on an official/aegis server? :o

[Koma]

I could always bring up an AEGIS server so we can test this, otherwise, it's easy enough to make a hunter on iRO Valk. We'll see.

As far as eA not picking up a bug fix request, that's quite common these days. A number of their developers have quit due to internal strife and arguments regarding how things were run and certain dev "mindsets"/corruption. I personally don't see eA continuing for much longer especially considering that even the more recent updates have primarily been less core driven. It may end up like Freya/Nezumi soon enough. My two cents on that though and hopefully it's not the case. I've run into too many of their core developers in the past year alone and I guess that's where my opinion comes from at least (five off the top of my head come to mind along with a couple of their admins too). It's very visible from /stable/ support dropping over the past couple of years and the lack of activity in /trunk/ as well.

[Andrue]

I don't think I have enough time for the grind of iRO Valk, but I have some friends that play on the server already. I'll ask them if they have a hunter that I can test the trap with, or if they have a guild member that I can bother to watch them use Land Mine. :O

I'll report my findings when I have the time.

[Andrue]

Okay, I've talked to a few people with a Hunter/Sniper on iRO Valk. From what I've been told, the trigger and damage range are both 3x3.

I'm fairly certain that Land Mine doesn't have a stacked damage multiplier, but I'm not sure if it damages one monster(the one that triggers the trap) or multiple monsters in the radius. I'll try to create a hunter and see what I can find out. Afterward, I'll see if I can code a fix that works as intended.

[Koma]

http://svn.eathena.ws/bugs/changeset/14101" TARGET="_blank

Key thing to note in this change which will come to us during our scheduled maintenance is this update:

- Updated Land Mine to have a splash effect. (Provided by Daegaladh)

Looks like they finally listened to you maybe. =)

[Andrue]

Ah neat. :D

It's about time... :(

[Andrue]

I would mark this as resolved. :o Thanks for all the help.

[Koma]

I'm going to leave it open until you've had a chance to confirm the results. I wouldn't want to prematurely close this if its not actually fixed. =)