Mr X, like zoolgremlin suggested, I think the input screwiness you found is related to Quality_of_Life\Run_After_Area_Change; it causes quite a few little issues. If you discover that's not the case let me know, but that one's on my list of stuff to get reviewed for improvement.
I assume regarding Moon Energy you're pointing out how huge of a damage bonus it is where it makes other buffs unnecessary?
Regarding undead enemies getting special cased on resists, it's certainly possible. In vanilla, I think the only code that actually makes a weakness affect damage is in the spell functions where a weakness becomes a 2x multiplier on incoming damage. If you wanted analyzer to not show the salamando undead enemy as being weak to undine, that'd get more complicated, and / or if you wanted the undead enemy to literally not be weak to undine (undine not set in their weakness flags), same deal, but if you just wanted spell immunities / weaknesses / non-weaknesses to use those rules it wouldn't be too difficult.
zoolgremlin, where are you seeing lots of NOPs in bank 0x00? If it's ~4 or ~8 between code like STA $004203 and LDA $004217, those aren't dummied out code, they're intentional delays to wait on the SNES's multiplication / division. Multiplication needs 8 cycles (3 NOPs + the result read) and division 16 cycles (7 NOPs + the result read) to completely finish (there are circumstances where an early read is valid, but I won't get into that here). While you can do work between starting a multiply / divide and reading the result, there does need to be time wasted for the result to be calculated so there has to be something between the write that starts them and the read of the results (at least on real hardware and in more accurate emulators).
hmsong, all enemies have a weapon, and some have two that they switch between, though most don't have a visible graphic. The issue with transforming an enemy into a mimic is if their weapon did have a visible graphic, it'd still be visible even after they turned into a mimic. So an Ice Thug holding a little fork would turn into a mimic with a little fork floating beside it. Most of an enemy's attack damage comes from their weapon instead of their base stats, if I recall correctly.
Colors for spell stuff is really annoying. It's almost always a 1-byte number that indicates which palette to use (and I think they're generally 5-color palettes), but said numbers are sort've all over the place. The spell icons are defined at the end of bank 0x18: starting at $D8FD7E, each spell is 2 bytes starting there with Earth Slide, where the first byte is the icon number and the second is the palette number. Colors for spell animations is worse: sometimes it's the first byte per group of 8 bytes in the spell data for animations (in your Grave Flare example, the "89" at the start of 6 of the lines is the palette number), but some animations don't use that value and instead have hardcoded palette numbers in their animations (which are primarily but not entirely in bank 0x11). Look at the annotated spell data for Firenado (can search for $D02DE0 in the ZPS); it's a lot to take in, but may give you more info to go on.
I don't currently plan to let Inferno Barrier also cast Blaze Wall; merging the two behaviors just seems to be too complicated to make work sensibly. Inferno Barrier's spell function is set up to be extensible to apply other status effects in the future (it doesn't actually currently, but it's easy for me to add), so for example, it could be set up as a "Medusa Skin" buff that could be given to Basilisks that they cast on themselves and / or allies where meleeing them would petrify the attacker, etc. Tying in an offensive spell would currently require hardcoding the offensive spell behavior. I did give it some consideration though.
One issue with swapping a detrimental spell and a beneficial spell is the enemies won't automatically change who they target; so if Fire Bouquet becomes Speed Up, enemies who originally cast Fire Bouquet on the players will instead cast Speed Up on the players. You'll have to find the spell cast command in their AI and change the target (and if changing from beneficial to detrimental, possibly also add AI code so they locate a target before casting). The spell target value in spell data affects who the players can target; enemies ignore it.
The spell names being in ALL CAPS means you have Text\Proper-caser Off, but it does make sense for Relocalized to also trigger Proper-caser behavior even without Proper-caser, so I'll change that.
Is [Enemy_Type_And_Element_Changes] an okay name? I've been trying to avoid using non-text characters in feature names (- hyphen has been the only exception). Added the "Intended for use with..." note to the [Alternate_*_List] descriptions.
kethinov, a Dryad Saber sounds neat. Probably a fair bit of work but maybe I'm overthinking it. Vanilla doesn't provide for a Dryad-specific saber spell function, but it doesn't take much code in the vanilla scheme to add one. More complicated might be working out the specifics of having it mesh with the Mana Magic logic, but I expect in the sprite's Mana Magic spell function it could be changed to check if the Dryad Saber is applied and use that as the basis for overwriting with Mana Magic. Probably even enough space by repurposing the girl's Mana Magic spell function to fit whatever new code is needed.
I don't think it makes sense to try and work out a scenario where both a Dryad Saber and Manaburn could coexist. It's not unthinkable to have patches be mutually exclusive.
I hadn't considered that someone would want vanilla Evil Gate and Firenado at the same time. Seems crazy to me but whatever, I'll support that combination. Ignoring that the first attempt at a generic Evil Gate function failed (not actually sure what I messed up on as it worked during testing, so I must've botched something when rewriting some of it), a generic form is still my plan if both More Useful Evil Gate and Firenado are On at the same time since it's ~73 bytes of waste otherwise, plus more if someone wants to add a third element variant, etc. And to that end, More Useful Evil Gate and Firenado both being on by default make the generic form the default; if in the future the Map Display Settings table ends up fully populated, I simply wouldn't support some mixed application of the three (something would win out and force the others On/Off to avoid the overlap).
I'm more interested in knowing your thoughts on removing the multi-target penalty so that use on normal enemies has more value (at its exorbitant mana cost) and / or overhauling the damage logic so that level 0 isn't functionally the same as level 8 once you hit bosses with enough health. My initial thought is a damage cap based on spell level so that the mana cost can be brought back down; in my play testing over the past couple days where I had boss health drastically pumped up, I found it unusable at 12 mana and wound up using Fire Bouquet, Acid Storm and sacrificial Burst for spell damage more than anything. I also wound up fast-forwarding during level 8 Burst spell animations, which is a bad sign. -_- Gave myself a new longterm goal of removing the full-pause from spell animations that have them (definitely one of the things I hate in Trials of Mana is the spell animations stopping the game).
I share hmsong's opinion that worrying about missable monsters when changing transform / shape shifter results will lead to less worthwhile transform / shape shifter results.
SFX Channel Limit is Off by default due to an abundance of caution; I don't know exactly how it affects the resulting audio, and my preference is to not lose any sound effects (though I don't even know if that happens with it applied; just guessing based on supposedly limiting sound effects to 3 channels). Considering that testing with it on, I still note music channels cutting out and don't notice any missing sound effects, I'm not even clear what it's managing to do.
And finally, thanks for the bug reports.
The Functional Fashion palette issues are known except for possibly the no starting equipment interaction (if you mean the cutscene girl appearance, that won't be "fixed" because that girl isn't the real girl, and applying the real girl's palette to the fake cutscene girl is way too much work for too little result, likewise with any other fake cutscene versions of the playable characters).
Fixing the poison palette is on my to-do list, but I also plan to optimize it so it's only replacing the character's flesh and hair colors, instead of their entire palette like it currently does.
Fixing the stamina blink will likely happen but is low priority, because that's non-default configuration, and because at a cursory glance it looked like a difficult fix.
The accidental hotkey triggering when leaving menus, etc. should be side effects of Run_After_Area_Change, and will hopefully be improved.
As for the gather party event, which edge of the screen are the party members stuck at (I assume that's what's happening). To clarify, you have Bug_Fixes\Gather_Party_Fix On? I should have fixed it for the left screen edge, so are you describing it occurring at the top edge? I may not need a save state if the circumstances match my hypothesis (top screen edge, since in that case I know what I'm looking for) but conversely, having one wouldn't hurt (using one is how I figured out the details to fix it for the left edge). Ideally, if the save state is made with Safe_Save_State On, and the game currently hard paused, or alternatively, along with a Turbo INI with the exact same settings set as were in use during the save state, but I can generally cope regardless.
As for what I've been up to, over the weekend I got through analyzing all 1024 doors. My findings are that only 21 are unused (with one, 0x0BF already claimed for Farmable_Sword_Orb), with some arguably having a "natural" (i.e. matango cave pitfall trap) purpose that removes them from availability for generic reuse (by which count I'd say there are ~10 unused with the others used to recreate cut content). Also pretty confident in some conclusions on cut content from the analysis:
- the wind palace had preliminary work done on a dungeon; any areas made for it were repurposed and reassigned to the earth palace dungeon
- the lighthouse interior had preliminary work moved to the light palace on gold isle
- the matango cave was reworked at least a little; it has an unimplemented pitfall trap and probably had another room that was scrapped
- there's an unused door that dumps you out in the empire castle moat; this could be a red herring but it's fun to imagine what it might have been for
- there was possibly plans for another intro cutscene scene; no map data exists, but an unused door between two other intro cutscene doors pointing to an empty map suggests they kept space for one
- the grand palace probably had one more room at some point
- some doors were simplified (i.e. animations removed) so that 2+ wide doors could share a single door table entry
All of that was to find a door or doors I could use to make the Goblin Village be a connection to the east side of Gaia's Navel. I can imagine using as few as 1, or as many as 3. Given how few unused doors there are I may opt for the 1-door approach. Next I need to give Gaia's Navel tidy map treatment to prepare for the addition of a trigger tile, probably far east under the log bridge. Then I need to hook my Goblin Village exit up to an event and set up the new west exit door. And find space in the trigger array for stuff. I'll probably want to make combat in the village functional and get time of day functional there. Oof, this is going to be a handful.