Name | dcss:brainstorm:misc: Traps |
---|---|
Summary | About dungeon traps and the Traps&Doors skill |
Further information | Trap creation from wands http://sourceforge.net/tracker/?func=detail&aid=2579528&group_id=143991&atid=757516 |
Added by | b0rsuk |
Added on | 2010-09-13 19:10 |
Overhauled | 2011-10-10 by dpeg |
Example: ......# Upon seeing the ogre, you will want to engage it. .@...O# ......# If you move towards it... ......# ...a trap like the marked one may be spawned. You can proceed slowly (walk, wait etc.) ..@^.O# but then you have less time to pelt the ogre with stuff. Or you can press onwards and ......# suffer a little (net, additional summons, item loss, noise). ^.....# You can try to avoid this by luring the ogre towards you. However, then a trap may still .@..O.# spawn in your vicinity. This is of no concern against a lone ogre like here (you kill the ......# ogre and the trap disappears if not yet detected) but ogres come in bands.
Traps which are interesting on their own (i.e. without monsters) like shafts or summon traps can use a special ruleset (some care is needed to make it scum-proof and to avoid punishing players who use autoexplore):
Reasoning: We don't want to place standalone traps in advance (i.e. during level building) in order to prevent players from boring themselves by slow exploration.
Example: ######################################### A miniature Crawl level. Some potential zones are indicated: #.<.....#.....11..######.........######## #..##.#.#.....11.....####..............## Suppose 1 is for a shaft, 2 is for a Zot trap. #...###.#.##..##.......##........####..## #.###....##...##..............#####....## Entering any square marked 1 triggers a T&D check. #.............11....####..#......#....### After the check, no other shafts are tested - all 1 marks are #.####........1########..#.#..._.#...#### removed. Same for all 2 squares. (The zones could overlap.) #.#..#....##############..#......#.....## #.#..+.....##############.....######...## The idea is that no matter how you explore (even if extremely #2#22#2222222#####2222##2222222222222222# conservatively), you will have to cross the zones. #.#..#..........1.......................# #.####..........1............##+++##....# #...............1............#.....#....# ###+############1............#.....#....# #..............+.............#..>..#....# #..............#.............#.....#....# #########################################
I like the idea, but the zones themselves may need some thought. I feel as though using a wall-to-wall strategy as presented above could still cause players to alter their behavior, with the rationale being that if they fully skirt the perimeter of a level before heading inward, they're guaranteed to not hit any standalone-type traps in the interior. Perhaps better to define zones as irregularly-shaped, contiguous regions, like so:
######################################### #.<.....#.....2222######.........######## #..##.#.#....22222222####..............## #...###.#.##22##2222222##........####..## #.###....##222##22222.........#####....## #......2222222222222####..#......#....### #.####...222222########..#.#..._.#...#### #.#..#....##############..#111...#.....## #.#..+.....##############11111######...## #.#..#.......#####....##1111111.........# #.#..#................11111111111.......# #.####..............111111111##+++##....# #....................1111111.#.....#....# ###+############........111..#.....#....# #..............+.............#..>..#....# #..............#.............#.....#....# #########################################
As long as the zones are sufficiently large enough to fill the thoroughfares they inhabit, and as long as the number of zones per level is random, this doesn't punish autoexplorers nor does it give players a predictable means of trap avoidance. — Wensleydale 2011-10-23 03:20
This proposal heavily redefines the purpose of traps, but it is (relatively) short and simple.
Shafting is a great effect, but shaft traps only work if you don't know about them. Introduce a shaft-like monster spell that sends the player a few levels down, maybe even into a different branch. Resistible with MR, of course.
Now for the most radical part of this radical proposal. I do not expect devs to like this part, but it's a part that can easily be left out while keeping the rest of the proposal. I want a shaft-like effect to replace banishment. Distinct from the other shaft effect, it should send the player to a location unrelated to their current one (this is necessary, otherwise you'd go to D:1 to unwield distortion), probably a random level out of a hardcoded set (sends you to a level between D:20 and D:27, for instance). The flavour of this is space warping around you - think of it as a teleport on a larger scale - so it fits perfectly with getting hit by or unwielding distortion.
The best part of this and my major motivation is that it makes “banishment” much less annoying. Abyss is trivial for strong characters, and strong characters will have already cleared most of the dungeon, which means it only wastes a few seconds of their time instead of ten minutes. Against weaker characters, the effect is very similar balance-wise.
Cutting mechanical traps and secret doors will require revision of many vaults. I personally volunteer to do this. — minmay 2011-11-29 17:29
This is a purposefully vague list of new trap ideas. (The last discussion got bogged down in a number mess.) Bear in mind that traps are supposed to be interesting on their own, and that avoiding them via T&D should be useful. All of the following traps have a single charge: like shafts, they disappear after use. Note that traps like portcullis are already in use via lua (check des/traps/ for inspiration).
For all of the new traps, both damage (which can be nominal — it is for flavour reasons) and destruction potential should increase with depth.
“Cloud traps” could have a broad range of effects in general; miasma, poison, meph, petrify, mutation. Also see “Tactical Usage” for how to make these actually a resource. — mumra 2011-09-04 17:26
galehar: We can use the existent expose_player_to_element function which does the damage and the item destruction. It uses a strength parameter and each item has strength% chance of being destroyed. clouds have a strength of 7 and lava a strength of 14. Maybe we can randomise the dungeon level a bit (10 - d(10-N) for example) and use it for the strength of the effect. What is interesting about this function is that all items have the same chance of being destroyed. So it means the more you carry, the more items are going to get destroyed. And carrying junk items doesn't help at all.
Optional: detected Fire/Ice/Magic traps can be set off with a thrown projectile (or something more subtle if we want, e.g. needing fire dart, flame/fire wand, fire spell to trigger a Fire trap etc.). This is gimmicky and not the crucial point of this proposal. (In any case, they can be disarmed like mechanical traps. No idea if you should get something if successful.)
There are cloud traps in trunk now. The mephitic/petrify ones are great, the poison ones are okay, but the fire and ice ones feel really cheap. They don't create interesting tactical situations, they just…destroy your items. Granted, I really dislike item destruction in general, but it is especially bad here since it doesn't even feel like it's your fault that your scrolls are getting burned. — minmay 2011-09-26 02:49
In a discussion on ##crawl-dev the consensus seemed to be that cloud traps are only marginally more interesting than mechanical ones, and those with permanent effects (fire, ice, mutagenic) add problems on their own. If you see a fire drake or a fire crab, you can at least run away, these traps just destroy items without you having anything to say. This can be significantly avoided by frequently stashing potions and scrolls, as having 1 potion on your person makes you lose only 10% of what you'd lose if you had 10. This just adds tedium and frustration while hardly ever being interesting. — kilobyte 2011-09-26 02:54
There are no mutation clouds. It also appears there might be a bug which causes a turn to be skipped when clouds appear (elliptic commented on this), so actually item destruction should happen on the next turn, giving you a chance to avoid it. Another option would be to introduce a delay of 1 or 2 turns after the trap is triggered before the cloud explodes. — mumra 2011-09-26 08:17
These all seem to create interesting situations that you have to work around (without the danger of just killing you outright, and without being able to simply rest it off) - so should be good.
Miasma - more clouds - see previous section.
Acid - could be quite harsh, especially with enchant scrolls being relatively rare. Maybe acid could be a new cloud type; and ensure that you have at least 1 turn to take action (i.e. blink, or put on rCorr) before anything bad happens.
Curses - Ashenzari could shield you from the curse, just like death curses. — mumra 2011-09-04 17:34
The spider basket trap is a fun example of this, what would be great is a whole range of different types of things that can be summoned, and tactical situations in which it can arise. I wrote a sample vault file that summons shadow creatures in a bigger area of LOS, and it's easy to make the summons depth and level appropriate by adding any number of vaults with different DEPTH: parameters. The script might need some revision but it's at: http://pastebin.com/LTg8CAmD — mumra 2011-09-04 17:34
These are perhaps extreme; and being watched isn't really any different than a more powerful alarm trap. — mumra 2011-09-04 17:34
From an idea by wensleydale on IRC. It's similar to a summons trap, but themed around spawning elemental monsters relevant to the local environment (and depth; slightly OOD of course). Two ways this could work: either analyse local tiles and based the spawns on that, or use hand-crafted vaults themed around a particular element. Ideas include:
I started coding some LUA for such trap vaults; might need a wide range of themes and layouts to prevent them being hugely spoilery.
Both sound good to me, and easily doable; teleport cap can be implemented as ammo when random traps are created. — mumra 2011-09-04 17:34
I mocked up an example IOOD trap to show how it's done; not suitable to be used in the game as-is, but it shows how to place an orb at A, targetted at B: http://pastebin.com/P6m10bQF
One area we may want to explore is the interaction between allies and traps. Summoners, for example, are a bit binary: they find many fights trivial, and some quite hard. Balance could be better. I think more attention should be paid to how new game elements will affect summoners. Traps are just one example. Characters with allies might be more interested in dealing with traps effectively, and this can be expanded.
A much more vague idea: some items could be hidden and only detected with T&D skill. Since we don't want players to re-visit all areas, this idea works best for portal vaults and Pan (but there it is interesting). Best to only have a single attempt at finding the loot — we don't want players to meticulously investigage every nook and cranny in a cleared portal vault.
The same could apply to hidden devices, which could turn off all traps on a level, or find all secret doors etc. (this idea is more fancy and has less appeal than hidden loot).
This is an attempt to make T&D more useful for players spending xp into it, by giving it an active use. Before going in any details, the basic ideas are:
I'd be against making wands for traps just cause it feels weird, and doesn't make sense since then why not use rods. Why not simply code a new device just for making a trap? A new class could even start with one or two of 'em. — greepish 2011-12-06 15:15
A new idea; being able to detect and use traps tactically could add new value to the Traps & Doors skill, making traps potentially an interesting and valuable resource, rather than a one-off hazard to just avoid or disarm when detected.
The possible mechanics and scenarios I'm thinking about are:
— mumra 2011-09-04 17:15