Positional Attack Magic Implementation [Now Playable]


Although the central place for design discussion is ##crawl-dev on freenode, some may find it helpful to discuss requests and suggestions here first.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Wednesday, 13th November 2019, 05:17

Positional Attack Magic Implementation [Now Playable]

The previous thread on the topic was a brainstorming thread that concluded with some workshopping of design of specific spells and the beginnings of a sketch for a big magic overhaul in 0.25.

I've been hacking away, and in this thread I'd like to gather specific feedback about the results. For those of you who compile at home, the draft branch is now available in the main crawl repo and the continued progress can be tracked in the draft pull request.

As this converges towards something that might end up in trunk, the initial design doc might age out of date, but it's currently reflective of my thinking and the state of implementation.

In this thread please don't comment with: complaints about nerfs just because they're nerfs, complaints about your favorite spell being removed, or off-the-wall theory crafting. Since this is still a work in progress, if you want to brainstorm further, head back to the brainstorm thread.

Once this gets a bit more mature playable experimentals will be online.

This is now a playable experimental on:
  • CPO (thanks chequers)
  • CKO (thanks floraline)
  • More coming soon
Last edited by ebering on Tuesday, 19th November 2019, 00:31, edited 2 times in total.

For this message the author ebering has received thanks: 5
chequers, gammafunk, Ge0ff, Implojin, sdynet

Ziggurat Zagger

Posts: 8786

Joined: Sunday, 5th May 2013, 08:25

Post Wednesday, 13th November 2019, 20:21

Re: Positional Attack Magic Implementation

The purpose of the Sting change is lost on me - it's still a manually targeted spell with range greater than 1, so the interface remains just as cumbersome as before, surely? Searing Ray still has that issue as well. Really glad to see IMB and Dazzling Spray fixed, though.

I really think you should try to go all the way in getting rid of spells with manual targeting. Single-target spells can still exist without manual targeting. For example, imagine a spell that shoots at the closest monster (picking randomly if there are multiple closest monsters) and stops on that square. I confess I'm at a loss for how specifically to salvage spells like Airstrike and LRD though.

For this message the author duvessa has received thanks:
tealizard
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Wednesday, 13th November 2019, 20:31

Re: Positional Attack Magic Implementation

I tried the foxfire spell. I recommend increasing the range by 1 across the board, making the fire things trigger on adjacent monsters so monsters can't attack or get blocked by them, and allowing the player to walk through them, but not switch with them (currently they block player movement -- not good). Works pretty well otherwise.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Spider Stomper

Posts: 200

Joined: Sunday, 11th May 2014, 11:26

Post Thursday, 14th November 2019, 11:31

Re: Positional Attack Magic Implementation

I'll withhold judgment on the other new spell. I think I'll have to do this a bit more time to make an accurate decision. But let me tell you about Starburst right now. I used this spell a bit, but found it to be a serious flaw.
The problem of this spell:
1. In order for the player to aim the enemy, he must move his seat. It's a very lethal and dangerous act. This is especially true when players are attacked by ranged attackers.
2. Because the spell orbit is not constant, you must approach the enemy to hit the spell.
3. It is helpless against enemies in diagonal passages.
4. In an environment(Besieged) where spell works as intended, it is effective to run away. Since you are a wizard, you can easily die if you are surrounded by enemies.
5. This spell is not strong enough for players to withstand and use all these problems.
Frankly speaking, this spell is cumbersome, tired and less effective than using Bolt of Fire. I think you need to make spell more convenient, or much more powerful.
Last edited by sdynet on Thursday, 14th November 2019, 17:56, edited 1 time in total.
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Thursday, 14th November 2019, 17:15

Re: Positional Attack Magic Implementation

Also re: foxfire, there's a problematic interaction with monster invisibility. The fire guys stay around for a long time and block the invisible monster, tank attacks.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Friday, 15th November 2019, 04:45

Re: Positional Attack Magic Implementation

Thanks for the feedback re: Foxfire. Having monsters trigger them when moving into them turns out to be a huge technical headache; (letting the player do it was fairly easy and has been added to the branch, Doesnty also pointed this out). To keep them from tanking invisible things I've made them SInv. A "seek out the invis monster" property seems too strong for an L1 spell (don't derail the thread with general complaints about monster Invis thanks), so I'm not 100% happy with this solution.

Since the Foxfires are fast they tend to go first in the monster move, since they dissipate it often looks like monsters are walking into them, so maybe this suffices.

Starburst needs more work, I agree. I'll circle back to it once I get the rest of the first-draft spells into the branch.

For this message the author ebering has received thanks: 2
chequers, sdynet

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Friday, 15th November 2019, 04:52

Re: Positional Attack Magic Implementation

Regarding "going all the way" and getting rid of single target spells outright. I hope to look at prototyping Implojin's "action button system" using clua once I merge a PR adding a targeting prompt to clua (you might think this is in the wrong direction, I ask you withhold judgement until you see the actual interface system). I remain convinced that there is some design space for single target spells (as long as they are sufficiently different from ranged), that space is tangled up with cumbersome UI, and that there is a possible UI improvement to free that design space. I'm aware that with a generic claim like this I have the onus to write some proof of concept code; patience, I intend to try at some point.

For this message the author ebering has received thanks:
Implojin

Ziggurat Zagger

Posts: 8786

Joined: Sunday, 5th May 2013, 08:25

Post Friday, 15th November 2019, 05:39

Re: Positional Attack Magic Implementation

ebering wrote:A "seek out the invis monster" property seems too strong for an L1 spell
corona negates monster invis and is considered one of the weakest level 1 spells in the game, it's probably fine. especially when the book has sticky flame anyway

also, glancing at the inner flame change again, maybe conjure flame should just be removed? the new-inner-flame way of creating flame clouds is way cooler and doesn't have the cast-in-advance problem of conjure flame.

For this message the author duvessa has received thanks:
petercordia
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Friday, 15th November 2019, 05:43

Re: Positional Attack Magic Implementation

About starburst, you should avoid patterns that will require the player to count past 2. Level 1 spells that detect invisible monsters are nothing new, I wouldn't worry about it. [okay, duvessa got there first on this one]
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Friday, 15th November 2019, 16:11

Re: Positional Attack Magic Implementation

Having recently made a thread about ally control/orders, it's occurred to me that positional attack spells will similarly interact very poorly with some gods, especially Hep/Yred. It's something of an abrasive interaction to have spells that depend on positioning while being denied agency for positioning. Unless it's intentional design that some spells/schools aren't usable simply based on this limitation, rather than as a built-in god conduct...but that would be odd in this context.

I do still think that better ally control is a preferable way to alleviate this compared to most alternatives, possibly all.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Sunday, 17th November 2019, 05:04

Re: Positional Attack Magic Implementation

Some further updates.

Starburst now fires unfuzzed bolts in the 8 crawl directions at range 5 (6 with Veh). Anisotropy and fuzzing seemed cool on first draft but they're not good.

Eringya's Noxious Bog is implemented to round out the poison revision.

I think this will be in experimental-ready state soon, stay tuned.

For this message the author ebering has received thanks: 2
chequers, sdynet

Spider Stomper

Posts: 200

Joined: Sunday, 11th May 2014, 11:26

Post Sunday, 17th November 2019, 13:58

Re: Positional Attack Magic Implementation

Hailstorm
The problem of this spell:
1. It is very difficult to use. In order to use it, it must be possible to block enemy access or allow users to move the distance. And in fact, if the user is not Sp or Ce, the only way to stop it is to stop it with Ice Beast. This spell forces us to use different spell.
2. Despite being very difficult to use, the damage is too low. It is much stronger and more stable to continue using Summon Ice Beast than this.
3. The spell's intent is probably to be a powerful force when surrounded by enemies, but the user is easily at risk because he cannot actually attack the beside enemy.

I distributed the executable to the Korean DCSS community and received feedback from people. I'm sorry to say this to you, but... People's feedback on this spell is all negative.

Ziggurat Zagger

Posts: 8786

Joined: Sunday, 5th May 2013, 08:25

Post Sunday, 17th November 2019, 19:41

Re: Positional Attack Magic Implementation

Minimum ranges aren't going to be a good fit for DCSS unless you get rid of speed 10 monsters. Consider why all the suggestions to penalize launchers at close range are (rightly) dismissed out of hand.

sdynet wrote:I distributed the executable to the Korean DCSS community and received feedback from people. I'm sorry to say this to you, but... People's feedback on this spell is all negative.
Although I oppose this particular spell too, I don't think this is really meaningful; feedback from existing players on all of these spells is almost certainly going to be overwhelmingly negative. There is practically no incentive for players to give feedback on things that they think are fine.

Spider Stomper

Posts: 200

Joined: Sunday, 11th May 2014, 11:26

Post Sunday, 17th November 2019, 20:12

Re: Positional Attack Magic Implementation

This is additional feedback from the DCSS community in Korea.

[Fire]
Foxfire
1. This was not produced when I was surrounded by enemies, so I couldn't do anything.
2. It is understandable that players were not allowed to be replaced for fear of Kiting, but it seems harsh for players to suffer damage when they are replaced.
3. This feels like using a summoning magic.
4. I felt tired when I used it in a place with complex topography. It was too much trouble.
5. It had a starting range of three, and it was frustrating to have to be near the enemy.
6. It felt like playing Wu Jian.

Conjure flame
1. It is very cumbersome to use.
2. To use this magic, you have to make two turns. Such an act in front of a long-range attacker is suicide.

Inner flame: It's hard to feel the change.

Starburst
1. It's very interesting, it's a tactical. This is well balanced.
2. Using this spell in a place with a wall of magical reflexes is suicide. This is probably an unintended bug, right?
3. The animation seems to be a bit long.

[Poison]
Sting: Good.

Eringya's Noxious Bog: A kiting spell. There is no reason to use this in the presence of a long-range attacker.

[Conjuration]
Searing Ray: Strong, convenient.

Dazzling Flash
1. The noise is red, and if you faith in Vehumet, it's purple. This is too loud.
2. This is less likely to happen as the farther away it is, and since there is no damage, the situation to use it is likely to be limited.

Iskenderun's Mystic Blast: Interesting, not bad damage, and good knock-back effect.

Overall opinion: The direction of change is interesting, but the overall decline in convenience is noticeable.

For this message the author sdynet has received thanks:
ebering
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Sunday, 17th November 2019, 20:53

Re: Positional Attack Magic Implementation

re: Foxfire and the huge technical headache, I strongly encourage you to power through it. The despawning thing would be useful to have around.

About the stencil style spells, hailstorm in particular, these would have a lot of added utility/interest with the gradius option style replacement for battlesphere I suggested in the other thread. Without something along those lines, I tend to agree about minimum range.

edit: just tried conjure flame, that shit is balanced af, outstanding work
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Monday, 18th November 2019, 15:22

Re: Positional Attack Magic Implementation

Conjure flame might be more "balanced" than 0.24, but it's a non-trivial increase in player tedium. Both in direct inputs and needing ~spoiler information or calculations in order to ensure flames are where you want them to be before monsters get there.

For this message the author TheMeInTeam has received thanks: 4
Aean, duvessa, nago, sdynet
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Monday, 18th November 2019, 18:00

Re: Positional Attack Magic Implementation

For a spell like this to have the role in the game it does, it cannot create a gap. Now if people want to talk about raising the spell level, that's different. If the spell fails because monsters are too fast or whatever, that's good -- conjure flame already fails because monsters are too fast or too close, you just don't think to include that in your analysis (of course). It is false that it is more tedious. Obviously false.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Monday, 18th November 2019, 19:56

Re: Positional Attack Magic Implementation

I'd be 100% for raising spell level, weakening the spell's damage + making monsters just walk into it, or alternative fewer input/calculation solutions. If the argument that conjure flame is too strong for level 3 (a reasonable assertion) these are preferable.

I'm thinking easiest is probably the proposal of "flames appear on character and burn you if you stay there" in addition to either raising spell level or making the flame do less damage. Let monsters walk in then.

conjure flame already fails because monsters are too fast or too close, you just don't think to include that in your analysis (of course). It is false that it is more tedious. Obviously false.


On the contrary, it is objectively more tedious:

- Right now, you know you can't place the flame because the monster is already there. No calculation or planning required. The yak is there, so the fire can't be. Proposed implementation allows the spell to 1) go off but 2) be cancelled by said yak, forcing the player to consider future yak movement + possibility for random energy...even when the Yak is not already there. This is computation that is much less frequent right now than under proposed solution.

- On more careful analysis I was (partially) mistaken on #inputs, in many cases if you've set a macro in both cases they'll be very similar or identical (you are using movement keys to place flame either way, casting spell either way), depending on how you want to structure/place the conjured flames (non-contiguous favors current model).
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Monday, 18th November 2019, 20:46

Re: Positional Attack Magic Implementation

Counterpoint: Calculation and planning is actually good.

What is really going on here is a spell that already required calculation and planning to use in many situations has been nerfed so that it now takes calculation and planning to use in more situations. A common situation with classic cflame is that you need to retreat to a choke point or cast more than once to block a monster. The monster's speed, energy randomization, and ambiguity about where a monster is in its "move cycle" already apply in these situations.

Now if you want something that has approximately the same effect as classic cflame, I go back to my original proposal of placing adjacent to the player nearest to the nearest monster, not placing on already clouded tiles, and randomly breaking ties among eligible options. Unfortunately, that ran afoul of the mistaken belief that randomized targeting of potentially targetable effects is bad.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Monday, 18th November 2019, 21:09

Re: Positional Attack Magic Implementation

I'm aware that cflame already has some issues. My point is that the proposed update somewhat exacerbates those issues rather than improves upon them.

Using ~spoiler information (exact monster move speed + random energy) to estimate monster movement nearly every time you use this spell (which will be more often, given the delay onset requiring more space) will make even practiced usage of the proposed cflame take longer than the present cflame. Since we both agree even the present cflame has undesirable properties, this does not seem an attractive direction to take the spell.

Your proposal of randomized placement would be a slight improvement over present model (fewer inputs due to not having to manually target the spell --> added convenience), but unless monsters willing enter flame would still allow gap creation/most of the present utility which is allegedly problematic. So while I'd appreciate the small reduction in micro demand it's not what seems to be in the cards for a significant rework for the spell.

I don't see why randomized targeting is considered bad. Isn't it already the case for say Qaz's disaster area, Ru's Apocalypse, some Nemelex destruction effects, and lamp of fire? These are all still useful abilities and already present in the game...random square in direction of monster is actually more controllable than these.
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Monday, 18th November 2019, 22:17

Re: Positional Attack Magic Implementation

I remind readers that creating a gap is when you go from a @M type situation to @.M type situation. Classic cflame only allows you to maintain a gap, not create one.

I mistakenly failed to say above that my proposal would not place clouds on top of monsters either, essentially it replicates a common use-case of classic cflame. It also cannot create gaps.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Monday, 18th November 2019, 23:01

Re: Positional Attack Magic Implementation

Okay, in that case it does meet your requirements and I'd say it's a strict, if mild, improvement over what we have now. I'd certainly prefer it to the janky "make a cloud under character that can be stomped out" interaction.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Monday, 18th November 2019, 23:27

Re: Positional Attack Magic Implementation

TheMeInTeam: You often write long posts arguing for some position, but just about every time I read one you've taken some strange definition or a "problem" that applies to all of crawl specialized to a spell to argue against it.

In the case of conjure flame: using it requires understanding monster movement. This is true of old and branch cflames, as well as every aspect of crawl combat. If there are spoiler issues, they don't have anything to do with the spell: they have to do with monster movement.

In the case of spells that require you to think more about the decision "use this or move": you seem to think that having to move to make something (more) effective is tedious. This isn't the devteam's definition, in fact its quite the opposite. Our goal is a game that involves tactical combat and the essence of tactics is the decision "act or move". The goal of this branch is to make that choice more relevant for magic users, while also differentiating magic from ranged combat.

For this message the author ebering has received thanks:
duvessa

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Monday, 18th November 2019, 23:40

Re: Positional Attack Magic Implementation

tealizard: Re chasing the technical demons in crawl code. There are a lot of branches with good ideas in the main repo that died when they hit a technical wall; I'm trying to avoid the trap many of my foredevs fell into. The best I can say is "we'll see".

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Monday, 18th November 2019, 23:42

Re: Positional Attack Magic Implementation

sdynet: Thank you for sharing this around with the Korean community! The feedback is very helpful.

One clarification question. I think this is a language barrier issue. By decrease in convenience do you mean "decrease in when you can cast a spell without moving" or something else?
User avatar

Mines Malingerer

Posts: 42

Joined: Wednesday, 18th July 2018, 23:51

Post Tuesday, 19th November 2019, 02:34

Re: Positional Attack Magic Implementation [Now Playable]

Some initial thoughts. Minor theorycrafting, sorry. Played using 0.25-a0-140-g3f2c7a7243 local tiles.

General: Most of the starting books feel like they're missing something now. I assume there will be some more additions at some point, but I haven't seen this explicitly stated. My personal thought is that getting rid of *all* ranged target spells is really overkill. I'm also surprised to find that some people find targeting so unpleasant or tedious; in the list of things that bother me about crawl, targeting spells doesn't even register.

Foxfire: Works more or less fine, I think you should be able to walk through them without taking damage, even if this causes them to just disperse.

Conjure Flame: I think it would be better if the flame always sprang up after one turn. I get that this makes it easy to make gaps in hallways so this might be too strong. I rather agree with TheMeInTeam's initial criticisms though: it definitely increases tedium and requires too much pre-planning for just placing a fire cloud. Yes, having more decisions about moving or doing something else is more tactical, but I feel like the reward for having to make such a decision should be more beneficial than what CF offers. The fact that foxfire also impedes movement currently exacerbates this.

Inner Flame: I barely noticed the actual flame cloud on damage triggering. If you made the clouds it created much more likely to happen, and last longer, it could replace Conjure Flame. An inner-flamed monster could even regularly create "smouldering flames" that function like CF does now as they walk.

Starburst: You can still cast it without Z even if it won't hit anything. It's a minor issue and you might be aware of it already, but it should be fixed at some point. When I first looked at your initial design doc this is the spell that had me most worried about how the new magic system would play out, so I'm glad you've changed it from the original idea. The 8-way burst works well.

Hailstorm: Seems fine. EDIT: Kind of weak, as discussed in my game.

Airstrike: Seems like a good enough change, Lightning bolt works well in hallways so it's fine if Airstrike is better in the open.

Earth Spells: I find it "funny" that the most universally used single target spells have stayed in the game. Not a complaint.

Sting: Seems like a good change and counterpoint to Poisonous Vapours.

Noxious Bog: It's fun but I'm having a hard time imagining it will be useful enough to be worth training a two-skill level 6 spell, especially when one of them is poison and Poison Arrow and Venom Bolt both no longer exist. Poison in general seems extremely weak now.

Searing Ray: Good change.

Dazzling Flash: Good change for conjurers, bad for enchanters because of the noise. I guess maybe that's OK. EDIT: Wait I didn't realize that this never does any damage. I guess it's potentially useful but I doubt I would ever use it on a typical conjurer.

IMB: I guess?

Agony: I think no one is ever going to use this spell if it's range 1.

Bolt of Draining: Removal of this and Venom Bolt makes it harder to attack things from behind undead allies. I realize that there are still some spells that can do this, but it feels disappointing. One idea would be to increase the range of agony to 2 or 3 but have it smite targeted, but I imagine that would be too strong.

edit: I don't know how you plan to introduce these spells into trunk, but in my opinion, Sting, Searing Ray, Dazzling Flash, Airstrike, and maybe Hailstorm (replacing Icicle) and IMB could just go into trunk right now. They all seem like very straightforward changes that would be good regardless of where the rest of these positional magic experiments go. Perhaps you want to do it as more of an all-in-one change though.
Last edited by kitchen_ace on Tuesday, 19th November 2019, 08:33, edited 2 times in total.
EthnicCake on CKO. It's an anagram.

For this message the author kitchen_ace has received thanks: 2
ebering, Stairdancer

Spider Stomper

Posts: 200

Joined: Sunday, 11th May 2014, 11:26

Post Tuesday, 19th November 2019, 07:30

Re: Positional Attack Magic Implementation [Now Playable]

I'll explain it so that there's no room for misunderstanding.
The 'decrease in convenience' I mean a lot of keystrokes in one use, and a lot of environmental factors to consider. In other words, it is meant to cause a lot of fatigue. I'll give you an example.

Flame tongue(good convenience)
1. Move to a location where you can attack.
2. Press the z key and select the this spell.
3. Aim at the enemy and press on the Enter.

Foxfire(bad convenience)
1. Move to a location where you can attack.
2. Make sure there is enough room to summon the foxfire.
3. Press the z key and select the this spell.
4. Press . to wait or move to avoid disturbing the foxfire.
5. You should be aware that replacing your seat with a foxfire can cause damage.
6. It is also less predictable because it cannot designate an enemy. It is possible that it simply disappears without hitting the enemy.
In other words, there are many things to think about. I must say in advance, I think this spell is interesting and fun. I just think it's a little annoying to use.

Starburst(good convenience)
1. Move to a location where you can attack.
2. Press the z key and select the this spell.

Conjure Flame(existing - good convenience)
1. Press the z key and select the this spell.
2. Determine where to create and press on the Enter.

Conjure Flame(rework - bad convenience)
1. Make sure there is more than two empty space from the enemy.
2. Also, Check the enemy's move speed.
3. For two turns you are defenseless. Make sure there is a risk of the long-range attack.
4. Press the z key and select the this spell. Then move the position one space after that.

Hailstorm(a rather bad convenience)
It's not painfully uncomfortable. But it relies on different magic by birth, which is quite annoying. I'm afraid I feel like I'm interfering with you. But I will carefully tell you my personal opinion. As I see it, there are two directions.
1. maintain the current constraints. Instead, it gives different effects to the 3x3 range. I think Freezing clouds or Metabolic Englaciation looks fine. In this case, it will be a unique spell with two effects(damage or debuff) coexisting.
2. remove restrictions. Normal 5x5 attack spell. far more powerful damage. However, the user is frozen. It goes well with the spell, Ozocubu's Armour, which should not be moved.

These are not bad things, but I think it should not be tired to use them, even if the magic is a little strong. As far as I know, the development team has moved towards eliminating the hassle of games. I know your throwable rework have the same intent. But some of the new spell, I think, makes me feel like I'm back with 0.12.
I think Starburst is the most complete of the things made to date. This is not tiring, simple to use, and tactical. I like this :D.

For this message the author sdynet has received thanks:
ebering

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Tuesday, 19th November 2019, 08:06

Re: Positional Attack Magic Implementation [Now Playable]

I like foxfire as a spell, but one thing I notice is that if you get stuck with something next to you in a hallway (or at a corner) at low level FE, you have no offense (You can't use either conjure flame or inner flame as offense against adjacent creatures) an inexperienced or unluckly player can get a monster on either side of them in a hallway (or accidentally walk down a dead-end) and be left with nothing but untrained melee attacks to try to kill their way out.

I realize that you can usually beat an early creature with untrained unarmed or a weapon but I suspect the experience probably feels pretty bad for a new player (not to be able to attack with their trained skills/primary offense) particularly with races that are weak physically.

I also agree with the comment earlier that foxfire makes the early game feel a bit like a later-game Wu game, which is Ok, but a bit complicated. It seems awkward to start the game no ranged offense *and* have limited types of positioning, where you can't easily take advantage of terrain that lets you make fights 1:1.

It would be nice if foxfire had say a longer initial range and was level 2, and there was some level 1 spell one could use hallways and/or when surrounded.
Spoiler: show
As an aside what sprung to mind for a level 1 fire spell was a channeling spell, which could only damage adjacent creatures and started with no or little damage and increased as long as you channeled.

Given the way foxfire relates to other spells, I would also say it should probably be straight fire, rather than fire/conjuration.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

For this message the author Siegurt has received thanks:
sdynet

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Tuesday, 19th November 2019, 08:20

Re: Positional Attack Magic Implementation [Now Playable]

sdynet wrote:Flame tongue(good convenience)
1. Move to a location where you can attack.
2. Press the z key and select the this spell.
3. Aim at the enemy and press on the Enter.

Actually with a proper setup, you can use a single key to attack with a targed conjuration (see automagic)

sdynet wrote:Foxfire(bad convenience)
1. Move to a location where you can attack.
2. Make sure there is enough room to summon the foxfire.
3. Press the z key and select the this spell.
4. Press . to wait or move to avoid disturbing the foxfire.
5. You should be aware that replacing your seat with a foxfire can cause damage.
6. It is also less predictable because it cannot designate an enemy. It is possible that it simply disappears without hitting the enemy.
In other words, there are many things to think about. I must say in advance, I think this spell is interesting and fun. I just think it's a little annoying to use.


You're using foxfire inefficiently, you can just make sure there's enough space and cast it over and over again (You may end up casting it one extra time doing it that way, but it's effective.)
Also if you want to avoid "wasting" a spell and also not waste a turn, you can just hit the thing with a melee attack. If you're just attacking over and over, you can also map this to a single keypress similar to conjure flame.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Tuesday, 19th November 2019, 08:32

Re: Positional Attack Magic Implementation [Now Playable]

Also it's a little weird that you can get your foxfires to attack with any length of action, this leads to some strange interactions (fast/slow movement, doing quick actions like wielding/unwielding things) Not explicitly broken in any specific way that I've identified, just weird.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Tuesday, 19th November 2019, 17:11

Re: Positional Attack Magic Implementation

ebering wrote:TheMeInTeam: You often write long posts arguing for some position, but just about every time I read one you've taken some strange definition or a "problem" that applies to all of crawl specialized to a spell to argue against it.

In the case of conjure flame: using it requires understanding monster movement. This is true of old and branch cflames, as well as every aspect of crawl combat. If there are spoiler issues, they don't have anything to do with the spell: they have to do with monster movement.

In the case of spells that require you to think more about the decision "use this or move": you seem to think that having to move to make something (more) effective is tedious. This isn't the devteam's definition, in fact its quite the opposite. Our goal is a game that involves tactical combat and the essence of tactics is the decision "act or move". The goal of this branch is to make that choice more relevant for magic users, while also differentiating magic from ranged combat.


The whole is a sum of its components. "All of crawl" may or may not apply positive/negative effects to a particular mechanic, or those effects may be noticeable or irrelevant to a particular mechanic. Factoring random energy/nonspecific monster movement speed is more computationally expensive on player time when trying to plan cflame placement than, say, walking away from a 10 aut monster, or attacking a faster one because you know you can't escape from it. These are important choices, but they don't require meticulous planning with player weighting probability of a turn being wasted mid-combat like proposed cflame, and they don't require you to make said calculations dozens to hundreds of times like using cflame.

Movement *is* an action, as are other choices. The issue with conjure flame specifically is the proposed change significantly adds to IRL time burden to utilize it, relative to alternative proposals or even the sub-ideal status quo. If that doesn't fit the definition of "tedious" I'm not sure what would.

Magic is already objectively differentiated from ranged combat in a few important ways. Making it more so might be interesting, depending on what is done. At best, cflame proposal will have players memorize patterns/spacing so they get utility out of it. At worst, it's going to sink enough time to significantly detract from the rest of crawl/alternative spells. It's good for positioning to be important in crawl. I'm not convinced it's good for using a spell to be a hassle.

For this message the author TheMeInTeam has received thanks: 4
Aean, duvessa, petercordia, sdynet

Abyss Ambulator

Posts: 1233

Joined: Wednesday, 23rd April 2014, 21:57

Post Tuesday, 19th November 2019, 19:42

Re: Positional Attack Magic Implementation

tealizard wrote:I remind readers that creating a gap is when you go from a @M type situation to @.M type situation. Classic cflame only allows you to maintain a gap, not create one.

I mistakenly failed to say above that my proposal would not place clouds on top of monsters either, essentially it replicates a common use-case of classic cflame. It also cannot create gaps.

Classic CFlame can create gaps, by creating a cloud on the other side from the monster, then walking through it. But it comes at a cost:
1) 10 aut of potential monster damage
2) 10 aut of CFlame damage

The damage is correspondingly less for faster movement/action characters.
User avatar

Mines Malingerer

Posts: 42

Joined: Wednesday, 18th July 2018, 23:51

Post Wednesday, 20th November 2019, 03:07

Re: Positional Attack Magic Implementation [Now Playable]

I've just played a bit with a VM that got Eringya's Noxious Bog, so a few more thoughts:

Bog is really good, and I think a great example of a positional spell that rewards players for thinking about positioning. My previous attempt I didn't really get that monsters will go in if they have nowhere else to go, i.e they treat it basically like water. So, enemies are really dumb, but it works out OK. Two thumbs up for this spell.

The player is also immune to their own bog if flying; I don't know whether this is too powerful or not. It certainly makes it a nice Tengu/Gargoyle spell.

VM is probably the strongest book background now that it has Ignite Poison. Previously with VM I would often go to Orc before Lair, because there are much less poison-resistant enemies there. This time I had no qualms about heading into Lair as soon as I saw it.

Also, the ease of now creating flaming clouds with Meph + Ignite makes the Conjure Flame changes seem even worse in comparison.
EthnicCake on CKO. It's an anagram.
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Wednesday, 20th November 2019, 03:23

Re: Positional Attack Magic Implementation [Now Playable]

It would probably be easier and good enough to make monsters swap places with foxfire guys. Probably has some weird side effects, but whatever.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Thursday, 21st November 2019, 18:54

Re: Positional Attack Magic Implementation [Now Playable]

New sting feels a lot like a better old-sting with shorter range, I still have the complaint about poison magic (generally) that the venom brand seems to apply more poison status more quickly than even dedicated poison spells do, but that's likely more a complaint about the venom brand being overpowered.

The balance in the new VM starting book between the shorter-range-but-more-immediate damage sting and longer-range-slower-damage-and more costly poisonous vapours feels much better than it did with the old version, there's reasons for both spells now. Ignite poison being added to the starting book is a good call, it makes the early game smoother (I think ignite poison should lose the transmutation school though, it's too limited-use to be a 3 school level 3 spell, compared specifically in the same book to meph cloud, which is much better at the same price)
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Thursday, 21st November 2019, 20:09

Re: Positional Attack Magic Implementation [Now Playable]

Hailstorm shouldn't prompt you to really attack your ice beast since it's immune.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Thursday, 21st November 2019, 20:47

Re: Positional Attack Magic Implementation [Now Playable]

Hailstorm works pretty well with ice beasts, but is terrible if you memorize it first (without blockers it's frequently a bad spell) I suggest making it level 5 (with proportionally more damage)
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Friday, 22nd November 2019, 01:50

Re: Positional Attack Magic Implementation [Now Playable]

Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Friday, 22nd November 2019, 03:53

Re: Positional Attack Magic Implementation [Now Playable]

Thanks for the report, I've pushed a fix.

The issue with the targeter is known, there are some similar issues with other friendly checks in the new spells; the right solution is a bit involved code-wise so it will be a little bit.

Lair Larrikin

Posts: 27

Joined: Monday, 23rd May 2011, 02:56

Post Friday, 22nd November 2019, 23:21

Re: Positional Attack Magic Implementation [Now Playable]

Reposting here since it got deleted on reddit...

I have about 300 wins to my name and have been playing since 0.4. I do play online, but these days I only play online exclusively to get a token 15 rune win for the tournament on the latest branch.

I feel these changes are badly misguided and represent a misunderstanding of both the functional and fantasy elements of a caster. This design is missing the forest for the tree.

Functionally, casters have been getting worse, both implicitly and explicitly since 0.10 or so. Around the time, Casters were undoubtedly the best archetype in the age of most things with 'normal' movespeed, moderate monster HP, control teleport, and strong wands (just to name a few things). The variance introduced by casting was easy to mitigate via the insane toolkit that was offered to the caster kits.

Then nerfs started happening. Control Teleport was taken away (this is a good thing), wands were taken away, a lot of spells got nerfed in some way or another. Monsters with Executioner-like properties were added (Caustic Shrieks / Shoals monsters). Monsters with a LOT of hp and resistances were added. Monsters that made managing finite resources complicated were added (Convokers, Wardens etc.). Mutations rules were changed, adding additional variance on top of variance. Available mid-game XP got nerfed several times. The list goes on and on.

The point I want to get at is this:

We are way past a point where playing a caster as anything but an experienced player is enjoyable.

In a tournament scenario, I used to reach for caster archetypes almost exclusively. They were really fun, and in the pre 0.12~ branches, they were also pretty busted. In modern Crawl, unless I have enormous amounts of time, there is no way I'd ever touch a caster. The real crime here though is that it's not necessarily hard if you know the game well. It's just tedious. It's a game about mitigating risk as much as you can at every step, which, in a game that can last between 20,000 to 100,000+ turns, is a nightmare. We have gotten to a point where casters are not fun to play and are flat out inaccessible to newer players, and these changes do not help that at all.

I love the abstract of this idea: "Tactical Positioning should matter for magic", but it already does, even on stuff you propose to remove. This is another forest/tree problem. Bolts are highly sensitive to positioning to realize efficiency gains, but you are axing two of them? Creative use of Conjure Flame is getting a severe nerf because FEs have something in the early game to carry them into mid? What exactly is an IE supposed to do into the mid game without their core nuke kit? Is every Elementalist expected to just pick up IMB? Is the game actually better for that? How is a newer player supposed to discover that?

The beauty of the design in the original elementalist kits was that they were self-contained, but still had an identity. You could roll a FE and basically never touch any other school until the late game, when you would ostensibly have more knowledge about what the game expects out of you. This was flavorful, and more importantly, accessible. The game gave you immediate access to the list of spells you have access to and gives you something clear to work toward by way of the spellbook you start with, and those spells by themselves could carry you through the midgame. With the proposed redesign, that elegance and accessibility is badly compromised.

Removing bolts is another pretty serious problem for a lot casters, but I feel like if I haven't convinced you so far, then writing 3 paragraphs about the importance of Bolts isn't going to go anywhere.

I feel strongly that the proposed changes damage the game a great deal in a space where the game might need more help than hurt right now. Compromising the elegance of the caster design for some higher-level goal is folly for the long term health and accessibility of the game. If you want to add spells that ask for stronger / different positional requirements, I'd encourage that, but you can easily do that by just adding them, not pruning away every Elementalist's core damage kit. Casters are a far cry from the godstate they used to be, and could use some love, especially for newer players.

Ziggurat Zagger

Posts: 8786

Joined: Sunday, 5th May 2013, 08:25

Post Saturday, 23rd November 2019, 01:25

Re: Positional Attack Magic Implementation [Now Playable]

thats a lot of writing for someone who didnt read the OP

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Monday, 25th November 2019, 17:46

Re: Positional Attack Magic Implementation [Now Playable]

That depends on whether the comments were based on experience or not.

Some of the changes are okay, others are rough. It also doesn't alter the considerations for casters as much one might expect. Foxfire sounds cool for example, but in most cases you just walk into monster range and smash the macro key for it repeatedly, just like with flame tongue. Only now you don't stand in a hallway while doing it, which means you might have to move a little bit. That's kind of annoying, but the extra strength on the spell makes up for it.

Predictably, I found myself barely using/mostly skipping conjure flame, instead spamming foxfire --> sticky flame to get off the ground. Granted, this is strong enough to carry though most of early-mid game after fireball...but how many interesting decisions per unit of player time were really introduced by this?

And yet some of the spells are pretty nice and any intention of making spells less annoying to cast is welcome. I do feel the changes achieve that in some cases and somewhat regress in others.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Sunday, 1st December 2019, 19:32

Re: Positional Attack Magic Implementation [Now Playable]

It turns out allowing monsters to swap with foxfire and having foxfire attack on this swap were of equal technical difficulty, and that difficulty was lower than I initially expected. Monsters will now walk in to foxfire (taking the hit) to try to get closer to the player, so it no longer can be used as a pseudo-blocker.

Travel prompts with foxfire have also been fixed.

For this message the author ebering has received thanks:
sdynet

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Monday, 2nd December 2019, 06:21

Re: Positional Attack Magic Implementation [Now Playable]

Update: The branch has been merged into master. Continued feedback for tweaks &c. is of course still welcome. Thank you all for the feedback in the development of this.

For this message the author ebering has received thanks: 2
chequers, sdynet

Spider Stomper

Posts: 200

Joined: Sunday, 11th May 2014, 11:26

Post Monday, 2nd December 2019, 08:40

Re: Positional Attack Magic Implementation [Now Playable]

The fox fire is marked with a question mark.
User avatar

Mines Malingerer

Posts: 42

Joined: Wednesday, 18th July 2018, 23:51

Post Monday, 2nd December 2019, 12:57

Re: Positional Attack Magic Implementation [Now Playable]

I missed the spellforged servitor change before now. It's a good idea but could OOD be moved down to a secondary spell? It's kind of annoying/dangerous to have on a servitor, and will still lead to (very slight) altering of your spell list for the servitor's sake.
EthnicCake on CKO. It's an anagram.

Spider Stomper

Posts: 200

Joined: Sunday, 11th May 2014, 11:26

Post Monday, 2nd December 2019, 19:01

Re: Positional Attack Magic Implementation [Now Playable]

I tried playing mages with the trunk version. Overall, it was satisfactory. The starting magic book has an intuitive theme. Except for Ice magic school. Here's the feeling I got from the starting book.

Fire: Explosives that should be handled with care.
Ice: No theme. Or odds and ends?
Air: Bouncing. Powerful but dangerous.
Earth: Stable and intuitive.
Conjuration: Crafty and Very tactical.
Poison: Applying a wide range of school skills.
Necromancy: Versatile, but requiring special resources.

I'll talk about Fire starting book before I talk about Ice starting book. Fox Fire, Conjure Flame, Inner Flame, Fireball are magics that are at risk of being swept away by the player. This theme is relatively consistent and completely consistent with the starting book. A player can easily understand which position to fight in if he plays only a few times. "Ok, I will keep steadily away from the enemies to avoid being swept away by the explosion."

The Shock, Lightning Bolt bounces on the wall. Air Elementalist sees the bounce of spell and can quickly understand that he must fight near the wall.
Venom Mage can quickly understand that when they find the undead, they must burn the poison cloud to fight it or run away altogether.
Other wizards are similar. It is relatively clear which tactics should be used primarily.

Ice starting book is heterogeneous.
Freeze attacks the neighbouring enemy. Ozocubu's Armour breaks when it moves. Players can summon Ice Beasts. "Um ... Is the ice magic concept a melee wizard?"
Hail storm should not bring enemies close. Summon Ice Beasts to block access. "Is this magic school a concept that can effectively damage enemies in the middle distance?"
So, why does Freese exist? What are the main tactics of the ice school? When I use Freeze and Hail Storm together, I feel like I'm using different school magic. The two belong to different themes. Ice school seems to be a versatile school that can do everything at first glance, but a closer look shows that the impression is vague.

The starting book determines the overall impression of the magic school. Magic does not have to be completely bound by the theme, but the starting book thinks the theme should be constant.

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Monday, 2nd December 2019, 19:55

Re: Positional Attack Magic Implementation [Now Playable]

Yeah, I still think the starting FE needs an ability to attack adjacent things in corridors with magic, it makes starting FE much worse than all the other books.

Personally, I would move Foxfire to level 2 and create:
  Code:
Immolate - (Fire) Level 1

[i]Encase yourself in fire, as long as you channel mana (1MP to cast, 1 MP per turn after the first, use '.' to channel) all creatures that attack in melee you take fire damage before landing their blows. Multiple attacks can smother the flames.[/i]

Damage increases the longer you channel, max turns channel-able is determined by spellpower (as is base damage) up to 5 turns at max spellpower (25), each time beyond the first it hits a creature it has a chance of cancelling the spell.


It preserves the positional intent of the branch, is un-targeted, and allows trapped FE to attack with their magic in corridors, it also goes well with sticky flame later (Slightly discouraging the "sticky flame and walk away" which is boring but presently optimal once you get it.)

Foxfire is a pretty cool spell, but it would be nice if it was level 2 and could scale longer.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

For this message the author Siegurt has received thanks: 2
petercordia, Stairdancer
User avatar

Mines Malingerer

Posts: 42

Joined: Wednesday, 18th July 2018, 23:51

Post Tuesday, 3rd December 2019, 12:39

Re: Positional Attack Magic Implementation [Now Playable]

Went to use Starburst against a hydra and got a message that my line of fire was blocked by a broken pillar, which wasn't actually blocking the Hydra. No problem actually hitting it, of course.

https://i.imgur.com/euVkqnf.png

(Is there a preferred place for bug reports etc. regarding positional magic changes?)
EthnicCake on CKO. It's an anagram.

Dungeon Master

Posts: 250

Joined: Thursday, 27th November 2014, 19:12

Post Tuesday, 3rd December 2019, 13:59

Re: Positional Attack Magic Implementation [Now Playable]

Now that it's in trunk Mantis is a fine place for reports.

For this message the author ebering has received thanks: 2
kitchen_ace, sdynet
Next

Return to Game Design Discussion

Who is online

Users browsing this forum: No registered users and 37 guests

cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group.
Designed by ST Software for PTF.