Charms Overhaul
This is a second draft of this design doc, based on feedback and conversations spurred by the first and channges in 0.25. See the revision history for the earlier work.
Deflect Missiles was left as is after the removal of RMsl with the hopes of a coming charms revision, until finally being removed in 0.25. The future charms revision was promised a while ago, in the interleaving time several forks have experimented with a variety of "permabuff" systems. These reduce MaxMP to have a charm on, which possibly drains MP when the effect happens; spell failure and power are factored in when the spell effect happens somehow. Ultimately, these begin to feel much more like items. This points me (ebering) in a direction: remove the remaining charms that pose "tedious optimal" issues, and re-introduce them as items.
Based on discussions in ##crawl-dev and elsewhere regarding an earlier draft
of this document, the following design principles have emerged:
- The current list of Charms spells is more a historical relic than the result of intensionality. The fact that they are grouped together now should not be used to guide this overhaul.
- Having Charms be a skill that's a mix of "these items and these spells" is not good.
Bad charms are charms that are optimally cast out of or just before entering combat with no downside other than the mp cost (or a downside that's easily mitigated). These are:
- Shroud of Golubria
- Excruciating Wounds
- Ring of Flames (rC-- is not enough of a restriction, nor is the brief cloud surround after it expires)
- Infusion (1 mp is a limit on usefulness, but once cast one starts swinging and stops casting)
- Battlesphere (elliptic points out that the MP cost is somewhat more relevant than other charms as it's a tradeoff between this and spells)
- Spectral Weapon (unwielding to avoid damage sharing)
- Not a Charm But: Darkness.
- Not a Charm but: Invisibility. Not quite a bad charm because of the Contam brakes.
These are worth keeping, perhaps they can become single school so that Charms as a skill is only for items. For these dual school spells if the single school version is too strong they can gain a level bump.
- Song of Slaying (noise ramp up and reset. might not be in the best place but is at least workable).
- Swiftness (not an all the time spell that requires some thinking about when to use it).
- Ozo's Armour (breaks on movement, limits to lighter armor, again requires some thinking not an always on charm).
- Death's Door (def. not always on).
Based on the above guiding principles, each charm spell will get separate consideration and either be moved out of the school, turned into an item effect, or removed. The current draft divides these into "revise significantly" and "make single school" with an "undecided" section for things that are unclear. Before getting into details, moving charm-effects to items leaves the question of what to do with the charms skill.
With no Charms spells, Charms doesn't make sense as a "school of magic". The items are still in flux, but for those where the effect makes sense to scale with a skill level (if there are any) there's a decision to make. Either Charms becomes another "interact with magical items" skill, or we use Evocations.
In the case that Charms is retained as a skill for interacting with items, a clear dividing line is needed: either Charms is for passives or Charms is for equipped items; leaving Evocations for either activated items or devices in inventory. This decision depends somewhat on what the final set of items looks like so for now those items that require skill scaling will simply say "skill". n.b. in the current list below has only one definite design that scales with skill and one that might tentatively, which points to using only Evocations.
The obvious answer here is removal. Another possibility is an amulet or ring of infusion that provides the infusion effect (scaled with skill?) decrementing MP with melee attacks. This would be a tradeoff with other alternatives in the different equipment slots. It's not a good weapon ego: we already have a skill scaling damage ego (pain) and plenty of non-scaling damage egos (it would need to be stronger than them to justify the MP cost, but then would incentivise swapping when out of MP).
An amulet of infusion would offer a tradeoff between extra early-fight damage at the cost of MP and the other possible amulets; now that amulet swapping is slow switching away when out of MP wouldn't be an issue like there would be with a ring of infusion. Skill scaling again steers towards pain, but without it the amulet probably would not be viable long term (which could be ok).
I (ebering) am not wholly convinced of the merits of this amulet.
Remove. As an item the shroud effect becomes effectively either a straight +HP item or, if there's some use of MP, trading MP for effective +HP. We don't have many +HP items, and guardian spirit does a much better job in the second category.
I (ebering) am currently unsure of what to do with this one. It could be moved to a different school (though thematic fits are hard to find—perhaps a transmutations reflavor?) but the effect is active and has a drawback which doesn't make for a good passive item. It could work as an activated item but the effect doesn't seem more compelling than alternatives in any item slot.
Spectral Weapons are cool, but as a pure Hex spell they would retain the "cast in almost every fight" problem. The staff of Battle has shown that tying a battlesphere to an item works nicely, so let's use this approach. As an item effect, the spectral property would automatically call the spectral weapon (scaled with skill) when needed. This could be either an amulet or an ego (restricted to certain weapon types, elliptic suggested staves or short blades).
ebering after more thought: staff-unique weapon ego feels like the way to go here.
Make this into an (unrand?) amulet fire enhancer with rF++ rC-- and the cloud trail. Cool! Fun! The slow amulet swapping makes for tradeoffs wrt cold vulnerability!
This isn't a Charm but does have some "maybe always up issues". Make it into an amulet/scarf/cloak of shadows with a spell de-enhancer (like the old lantern!).
Each of these has a good mechanic and is dual school, they could all work single-school with a possible level adjustment.
- Swiftness
- Ozocubu's Armour (separately kate- has discussed making the AC scale with armour weight instead of a hard cutoff)
- Excruciating Wounds. Instead of a level adjustment, could get -Wiz in place of or in addition to its noise cost since the noise cost is mitigated well with stairs.
- Death's Door
From ##crawl-dev while discussing an earlier draft:
10:34:51 <elliptic> ebering: I think battlesphere is a bit different from the other charms you list because it is used on characters for which running out of MP in a battle is a serious concern, and depending on the layout of the battle it can easily be a wasted cast (because it can't reach the enemies) or die before doing much
10:34:51 <elliptic> and then the MP you spent on it is an actual cost
10:35:27 <elliptic> it's different from spectral weapon in that the spectral weapon users aren't going to be using MP after casting spectral weapon regardless, so it's not a cost if the spectral weapon doesn't turn out to be useful in the given fight
This was in response to the earlier draft, which proposed:
- Charm of Iskenderun: Spawns a battlesphere when danger is present.
In a different direction kate wrote:
I'm not sure that Battlesphere would work very well as something tied to a general piece of equipment - after the spell targeting changes the number of characters that would use it is much smaller, so it'd be a pretty narrow effect. If it's removed as a spell it could probably remain exclusive to the staff of battle (and could be removed as a monster spell too? I don't think it's a great monster spell regardless). - kate
In light of the positional spell changes, removing battlesphere (leaving it on the staff of battle) or making it pure conjurations seem to be the best two options, but it's not immediately clear which of the two is better.
A powerful effect provided as a potion, an evocable, and a spell! It has a good brake, but perhaps we should cut down on the sources of it. No firm plans as of yet.
Change here is generous. The resulting Skald book, such as it is, lacks a unifying spell school and much at all in the way of spells (Ozo's armour and possibly SoS the survivors). In light of these changes they've sung their last. If any of the resulting items aren't too overpowered at D:1 they could be added to a rare Wanderer roll.
If Charms is retained as the "use equipped magical devices" skill, these should use Charms in place of Evocations.