|Anonymous | Login | Signup for a new account||2019-07-19 08:45 CEST|
|Main | My View | View Issues | Change Log | Wiki | Tavern | News|
|Viewing Issue Simple Details|
|ID||Category||Severity||Reproducibility||Date Submitted||Last Update|
|0007415||[DCSS] Bug Report||minor||have not tried||2013-07-31 04:56||2015-11-26 14:58|
|Status||resolved||Product Branch||0.13 ancient branch|
|Summary||0007415: Mephitic Cloud's autotargetter does not work properly|
|Description||When you cast Meph, it will not autotarget at creatures outside of its range by 1 space, the way Fireball autotargets. The spell can be selected with z instead of Z, but it will center on the player instead. This happens for both poison-vulnerable enemies and poison-resistant enemies. (Additionally, Mephitic Cloud probably should not autotarget enemies with poison resistance.)|
|Tags||No tags attached.|
DCSS-Meph_Targetting.png [^] (26,901 bytes) 2013-07-31 04:56
edited on: 2013-08-01 23:47
This happens because Mephitic Cloud isn't guaranteed to splash into every adjacent cell unless it is at max power. In fact, once you max out the power, it starts targetting "correctly".
This happens because the edge-of-LOS explosion targetting code (directn.cc: _find_monster_expl() ) only considers cells that the explosion is guaranteed to hit. This check is necessary to avoid doing things like targetting IMB near the edge of LOS (where it probably won't explode anyway), or placing a firestorm with the monster only in the outer "maybe" radius. Changing the behaviour for meph should be done in a way that doesn't cause problems for these other spells.
One possibility would be to add a new affectedness type AFF_PROBABLY, and have targetters use that rather than AFF_MAYBE if thee probability of affecting the cell is sufficiently high.
For reference, for each of the eight surrounding cells, meph at power 0 has a 56% chance to place a cloud there, at power 25 a 67% chance, at power 50 a 78% chance, and at power 100 a 100% chance.
By comparison, firestorm at power 75 has a 0% chance of the larger radius, at power 100 a 24% chance, at power 150 a 49% chance, and a power 200 a 62% chance.
|Possibly fixed by doy in https://github.com/crawl/crawl/commit/62b5303cbeee92; [^] we could keep an eye out to see if anything goes wrong because of it.|
|I went ahead and fixed this in 62b5303. While it's true that targeting the edge of LOS where it's unlikely to hit isn't great, it's not like leaving the targeter on the player is particularly better, and this situation only occurs when there is no better place to target anyway. I don't think trying to determine what "sufficiently high probability" means is going to result in anything that's particularly better at being understandable or predictable.|
|Looks like 0010103 has a case where the autotargeter targets 3 spaces away from a target with firestorm, but the player could put one two spaces away; possibly this behaviour could be more of a last resort than it is now.|
|Ah, yeah, good call. It should probably try first to target at max range, and then fall back to min range as a separate step. I'll see about fixing that this weekend.|
|2013-07-31 04:56||magicpoints||New Issue|
|2013-07-31 04:56||magicpoints||File Added: DCSS-Meph_Targetting.png|
|2013-08-01 23:42||neil||Note Added: 0023620|
|2013-08-01 23:47||neil||Note Edited: 0023620|
|2015-11-24 16:48||wheals||Relationship added||has duplicate 0008994|
|2015-11-24 16:49||wheals||Note Added: 0029825|
|2015-11-24 16:49||wheals||Status||new => resolved|
|2015-11-24 16:49||wheals||Fixed in Branch||=> 0.18 development branch|
|2015-11-24 16:49||wheals||Resolution||open => done|
|2015-11-24 16:49||wheals||Assigned To||=> wheals|
|2015-11-26 08:41||doy||Note Added: 0029844|
|2015-11-26 14:24||wheals||Note Added: 0029848|
|2015-11-26 14:24||wheals||Relationship added||related to 0010103|
|2015-11-26 14:58||doy||Note Added: 0029851|
|Mantis 1.1.8[^] Copyright © 2000 - 2009 Mantis Group|