Final Fantasy Hacktics

Modding => Hacking/Patching Tools => Topic started by: Vanya on May 31, 2009, 11:25:28 am

Title: FFTPatcher Blank & Unknown Research
Post by: Vanya on May 31, 2009, 11:25:28 am
Help us make our community better!! ^_^

As many of us know there are a number of options in melonhead's patcher that are marked unknown or are blank. We need some thorough testing to decipher what these things do if anything at all. Anyone who would like to contribute to the community or just has some extra time on there hands is encouraged to participate in the testing of these values.

As we make discoveries they can be added to the wiki for convenience.

I'll begin by listing the relevant options from the patcher and my own notes.
Title:
Post by: Vanya on May 31, 2009, 11:25:53 am
Unknowns & Blanks:
(listed top to bottom, left to right)

Abilities Tab[/b][/u]
Unlabeled AI Box 1
(under skill type pull-down menu)
01) Action?---------->Controls weather the ability's name is displayed before execution. On=display name. [contributed by Kamek]
02) <empty> 1st
03) Unknown 1st---------->Controls weather the ability is considered for use by an enemy unit. On=usable by enemy. [contributed by Kamek]
04) Unknown 2nd
05) Unknown 3rd
06) <empty> 2nd
07) <empty> 3rd
08) <empty> 4th
09) <empty> 5th
10) Unknown 4th---------->*Controls the target's side-step animation if this ability misses. On=trigger side-step animation.[contributed by R999]
*Note that this requires that the 'Evadable' boolean at the bottom and the Unknown 3rd boolean in the AI Behavior box are unchecked(off).

AI Behavior Box
01) Ignore Range?
02) Unknown 1st
03) Unknown 2nd---------->Informs AI that this ability is affected by the Faith statistic.[contributed by R999]
04) Unknown 3rd---------->Informs the AI that this ability is evadable. On=trigger side-step animation.[contributed by R999]
05) <empty>

Unlabeled AI Box 2
01) <empty> 1st
02) <empty> 2nd
03) <empty> 3rd
04) Normal Attack?

Items Tab[/b][/u]
Shop Availability Pull-down Menu
01)5 <Blank>.
02)Literally hundreds of Unknown (##).

Unlabeled Shop Box
(under availability pull-down menu)
01)Unknown

Unlabeled Weapon Type Box
(upper right-hand side)
01) <empty> 1st
02) <empty> 2nd

Unlabeled Attack Type Box
(weapons only)
01) <empty> ------------------------>This is required for the weapon to show up in the Throw menu.

Items Attributes Tab[/b][/u]
All 3 Status Boxes
01) <empty>
02) Dark/Evil Looking
(really just want clarification and confirmation of the effects of 02)

Jobs Tab[/b][/u]
Equipment Box
01) <empty>

Action Menus Tab[/b][/u]
Pull-down Menu
01) Blank 1st
02) Unknown 1st
03) Blank 2nd
04) Unknown 2nd

Status Effects Tab[/b][/u]
Hex Value 1st
01) Unknown

Hex Value 2nd
01) Unknown

Unlabeled Attribute Box
01) <empty> 1st
02) <empty> 2nd
03) <empty> 3rd
04) <empty> 4th
05) <empty> 5th
06) <empty> 6th
07) Blank
08) <empty> 7th
09) <empty> 8th
10) <empty> 9th
11) <empty> 10th---------->This bit determines whether the status is canceled by the special status "Immortal". [contributed by Zodiac]
12) <empty> 11th

Inflicted Statuses Tab[/b][/u]
Unlabeled Status Type Box
01) <empty> 1st
02) <empty> 2nd
03) <empty> 3rd
04) <empty> 4th

ENTD Tab[/b][/u]
Unlabeled Attribute Box
(split into 2 boxes)
01) <empty> 1st
02) <empty> 2nd
03) <empty> 3rd

Unknown Box
01) hex value 1st
02) hex value 2nd
03) hex value 3rd
04) hex value 4th
05) hex value 5th
06) hex value 6th
07) hex value 7th

Distribution:
□□4□
□257
136□

Jobs Unlocked Pull-down Menu
01) Unknown

Initial Direction Pull-down Menu
01) Unknown0x30
02) Unknown0X33

Move-Find Item Tab[/b][/u]
Unlabeled Attribute Boxes
01) Unknown 1st
02) Unknown 2nd
03) Unknown 3rd
04) Unknown 4th
Title:
Post by: Vanya on July 21, 2009, 07:39:56 pm
Phuckin' bot...
Title:
Post by: Mental_Gear on July 22, 2009, 04:18:39 am
Thing is, if some of these are unused spots they can be used to add instead of replace (though they wont show up in TacText) but if they work in the background they can cause glitches.

I think unused slots like Midgar Zolom should be put here too.
Title:
Post by: Vanya on July 22, 2009, 03:10:58 pm
Well the blank/unused skill & skillset slots don't really pertain to this. This list is for the bytes you can set within these slots.
We pretty much know exactly what skills and skillsets are used and which are not by examining job & ENTD entries.
(Midgar Zolom is one of Serpentarius' skills.)

What we really need is some serious testing to see what the unknown booleans and values do if anything. There are some values that I already have leads on, but most of these haven't been closely examined. The point of this is not only for documentation, but also to find out what values are truly unused. The ones that are not used at all can be exploited for some really awesome ASM hacks.
Title:
Post by: Pickle Girl Fanboy on August 07, 2009, 06:49:14 pm
I just realized, you can use the Gameshark tab in FFTPatcher to get the codes for these, which could make for easier testing.  If I didn't live in a tent right now, I'd be testing.
Title:
Post by: Vanya on August 09, 2009, 03:09:29 am
Interesting point, Ditrta. I'll keep it in mind.
Title:
Post by: Kamek on August 25, 2009, 08:49:20 pm
In the "Abilities" tab, there's a checkbox marked "Action?". This value appears to control whether or not the skill's name is displayed before being executed. Most skills have it on, but there are a few that don't. If you've ever wondered why the Byblos' Energy and Parasite never show the skill name, this is why -- their "Action?" flags are missing.

Counter Tackle and Counter don't have the flags, either, and those two reaction skills don't display skill names.

I discovered this while trying to make two new skills for a custom job class, and wondered why the AI would never use said skills, or why the skill name never showed, then I realized I was missing two checkboxes -- the "Action?" flag, and the first Unknown flag. Then I noticed the lack of "Action?" flags on Energy and Parasite, and came to this conclusion. Sure enough, the Byblos skills now display their names when being used.
Title:
Post by: SilvasRuin on August 25, 2009, 09:37:53 pm
Nice discovery.

I still have yet to test it, but I think the first of the last three values in the new animation tab controls what type of effect it is, such as the energy gathering for spells, the wisps of power when summoning, etc.  I notice it is a different value for each of these action types.  It would probably be a quick test, but I'm kind of lazy about making and applying patches just for testing purposes.  I suppose all I'd have to do though is put the number the summons have in for one of the Squire's abilities and see if those wisps show up...
Title:
Post by: Kamek on August 25, 2009, 09:55:23 pm
Yeah, copy the values from a summon spell to say, a black magic spell, and you get the summon spell green orb thingies. Basically seems to be the "preparation" part of the spell, whether the guy stands still, stands in a "ready" pose, chants, sings, dances, etc, plus the charge-up animation or whatnot.
Title:
Post by: Vanya on August 26, 2009, 01:58:30 pm
@Kamek: Nice work! I added your info about the action flag. About the  1st unknown flag... if I understand you correctly, it means that if it's off that an enemy unit won't use the skill at all?

@SilvasRuin: You are indeed correct. When I researched these values months ago I verified this. Basically all animations have 2 parts. The first part is a prep animation and the second is the action animation itself. For example; most spells consist of the energy gathering part (first value) then the casting part (second value). The third value seems to control certain text being displayed after the action is completed like telling the player that they gained some PA or whatever. Note that the third value doesn't control all text as some seems to be hard coded to certain actions.
Title:
Post by: Kamek on August 26, 2009, 03:02:43 pm
Yeah, it seems first unknown flag controls whether the AI will use the skill or not. It's turned off for skills like Gil Taking, Negotiate, Invitation, etc, so the AI will never steal gil or try to invite your units. It's also turned up for Scream (auto-battle Ramza will never use it), as well as Dragon Tame (adds Invite), Dragon Care, Dragon PowerUp, or Dragon LevelUp. Not sure why they left it off for those...

Assassin Seal is disabled too, but for a different reason (AI flags are set to "Cancel Status" instead of "Add Status", so it's never considered for use.)

I also expanded the observations on the ENTD unknown bytes (http://www.ffhacktics.com/wiki/ENTD#Unk ... servations (http://www.ffhacktics.com/wiki/ENTD#Unknown_section_observations)). I believe it's a flag-based field, though I'm not exactly sure how the flags work. For instance, a 48 in field 4 indicates to attack/defend a unit, while a 58 also indicates to attack/defend a unit, but may indicate a more aggressive strategy. The flag there is 10, and I tried setting that on a bunch of random battles, but didn't notice much of a change in the AI behavior... though I was able to get a critical Vampire to come and Blaster my guys at Mandalia Plains (instead of running away) even with other enemies still in good health.
Title:
Post by: Vanya on August 26, 2009, 08:33:11 pm
Cool. Thanks for the contribution. I added the info on "unknown 1st".
Title:
Post by: Xifanie on August 26, 2009, 08:46:59 pm
Statuses, bit 15 = Immortal immunity
Title:
Post by: Vanya on August 26, 2009, 08:52:45 pm
That would the last check box on the right hand column?
Also, by Immortal immunity I'm guessing you mean that the special status Immortal will block any status with this bit turned on?
Title:
Post by: Xifanie on August 26, 2009, 09:20:19 pm
not last, before last.
and yes.

I didn't know how to work it so I just went the simplest I could. :P
Title:
Post by: Vanya on August 26, 2009, 11:06:42 pm
That's ok. Data updated! ^_^
Title:
Post by: Xifanie on August 27, 2009, 08:11:24 am
not the 11th, the 15th. You're missing 4 bits in your list.
Title:
Post by: Vanya on August 27, 2009, 12:48:03 pm

[] Freeze CT     [] Can React
[]               [] Blank
[]               [] Ignore Attacks
[]               []
[]               []
[]               []
[]               [] <----- This one, right?
[] KO            []

Title:
Post by: Vanya on September 29, 2009, 02:10:34 pm
OK. I figured out the the blank bit in the attack type box for weapons in the Item tab.
It's the blank spot between 2 Hands and Forced 2 Hands.
It is required for weapons to show up in the Throw menus.
If there are any rare weapons you don't want to allow for throwing, then clear that check and vice-versa.
Title:
Post by: Xifanie on September 29, 2009, 02:40:10 pm
Whoa crap, I never saw that other post: Yes for that.

And cool, I always found the values weird, don't think I would've ever thought about it.

Oh, and I invite everyone to post modifications melonhead has to apply in this wiki page: http://www.ffhacktics.com/wiki/Melonhea ... o_do_notes (http://www.ffhacktics.com/wiki/Melonhead%27s_to_do_notes)
Title:
Post by: Vanya on September 29, 2009, 03:05:34 pm
Cool! I'll be heading over there to post some stuff! ^_^
Title:
Post by: Pickle Girl Fanboy on October 17, 2009, 04:55:54 pm
Anyone ever notice that monsters that you can Jump on top of to reach something else have Jump values over 128?
Title:
Post by: LastingDawn on October 17, 2009, 06:33:11 pm
Yes, it does appear that way, it's more then likely the one in front, or a specific byte value being called "special" by the game.
Title:
Post by: MiKeMiTchi on October 20, 2009, 11:35:38 pm
Is the second post up-to-date?
Title:
Post by: Pickle Girl Fanboy on October 21, 2009, 06:42:59 pm
Quote from: "LastingDawn"Yes, it does appear that way, it's more then likely the one in front, or a specific byte value being called "special" by the game.

Lol, I know what bit toggles are. :)
Title:
Post by: MiKeMiTchi on October 21, 2009, 08:19:15 pm
That's what monster sprites have that generic humans don't have.
Super Jump!

Like the Poison Nail animation..
If we give that skill to a non-mon sprite, it won't jump. Too bad. :(
Title:
Post by: Vanya on October 23, 2009, 08:09:34 am
Quote from: "MiKeMiTchi"Is the second post up-to-date?

AFAIK it is.
Title:
Post by: Pickle Girl Fanboy on October 26, 2009, 02:51:10 pm
If I set the Y Value of Scream to 129, Scream does exactly the same thing it normally does.  Nothing changes, which is surprising.
If I set the Y Value of Scream to 65, then Scream  gives a boost of 65 to PA, MA, and Speed.
So bit 7 for Y in ability data is unused.


If I set the X Value of Scream to 138, Scream does exactly the same thing it normally does.  Again, nothing changes.
If I set the X Value of Scream to 74, then Scream boosts Brave by 74.
Bit 7 is also unused.


Any MP Cost over 99 shows an infinty sign as it's MP Cost.
Just plain weird.

I know I had something else...

Got it.  You could give an item attribute +128 to Jump to turn any unit that uses that item into a Step-Up, like dragons, hydras, and worker 8.
Title:
Post by: MiKeMiTchi on October 26, 2009, 05:15:14 pm
QuoteAny MP Cost over 99 shows an infinty sign as it's MP Cost.
Just plain weird.

Cool!

Quoteinto a Step-Up, like dragons, hydras, and worker 8.

Step-up?
Title:
Post by: Aquablack on October 26, 2009, 05:52:51 pm
Step-up means they can hop on the monster's back to get up higher ledges.
Title:
Post by: Vanya on October 26, 2009, 07:04:56 pm
Quote from: "
Pickle Girl Fanboy"
If I set the Y Value of Scream to 129, Scream does exactly the same thing it normally does.  Nothing changes, which is surprising.
If I set the Y Value of Scream to 65, then Scream  gives a boost of 65 to PA, MA, and Speed.
So bit 7 for Y in ability data is unused.


If I set the X Value of Scream to 138, Scream does exactly the same thing it normally does.  Again, nothing changes.
If I set the X Value of Scream to 74, then Scream boosts Brave by 74.
Bit 7 is also unused.


Any MP Cost over 99 shows an infinty sign as it's MP Cost.
Just plain weird.

I know I had something else...

Got it.  You could give an item attribute +128 to Jump to turn any unit that uses that item into a Step-Up, like dragons, hydras, and worker 8.

Those are some far out discoveries.
Title:
Post by: LastingDawn on October 26, 2009, 08:00:53 pm
Haha, that is pretty interesting work Death is the Road to Awe, very nice discoveries, though I think the "infinite" sign wasn't a mistake. I feel it was intended originally for a spell or two to be above 100 MP, but didn't want to take the effort to change the two graphic to three.
Title:
Post by: Aquablack on October 26, 2009, 08:26:08 pm
So, could I make an accessory that, when equipped, allowed other units to use them as a step-up? That would be interesting.
Title:
Post by: Kokojo on October 27, 2009, 01:14:00 pm
Definitively using that in COP for some story missons...

And now that I think of it, Step-up Pillards that are really sprites...and can be destroyed. How awesome would that be ?
^ Dibs on that idea !
Title:
Post by: Zozma on October 27, 2009, 11:25:35 pm
it would be cool if it could be asm hacked so that a value of 100 mp setting the infinity sign just makes it so it uses all the casters mp.

i guess it would have to be a pretty good spell tho
Title:
Post by: Cheetah on October 27, 2009, 11:35:15 pm
Quote from: "Kokojo"Definitively using that in COP for some story missons...

And now that I think of it, Step-up Pillards that are really sprites...and can be destroyed. How awesome would that be ?
^ Dibs on that idea !

I wonder if there is a way to do this and your fence idea with EVTCHR stuff instead of using actual sprites. It seems like it should be possible. What happens if you have the sprite just use a frame of animation, not march, and have the "character" set up so their turn never comes up? I think it would just stay motionless in the animation. The only problem is if it took damage it would likely do the damaged animation...hhmmmm.
Title:
Post by: Zozma on October 27, 2009, 11:42:56 pm
isnt that why they actually made battle sprites for characters like "teta" because the event fraemes wouldnt stay or something?
Title:
Post by: LastingDawn on October 28, 2009, 12:53:03 am
I believe that's the case Zozma, but I can't 100% recall, but all things point to that.
Title:
Post by: Pickle Girl Fanboy on October 29, 2009, 02:20:28 pm
The X and Y values can be greater than 128 (0x80); it looks like it depends on the formula.
Raise [Formula 0x0D Heal_(Y)% Hit_F(MA+X)%] has an X of 180 and a Y of 50.
So this is good and bad news.
The good news is it should be less of a pain in the butt to find Scream's Formula and tell it to treat X and Y Values as positive if bit 7 is not set and negative if bit 7 is set.
The bad news is, if you want every formula to treat X and Y values as signed, you'll have to either modify each and every formula or rewrite the way the game handles formulas.
But I'm no programmer, so take my ideas with a dumptruck of rock salt.

Various other musings...

Scream's formula is 0x3B [+Brave_(X) +PA/MA/SP_(Y)NS]
For the X:
1. Bit 7 to toggle positive/negative values,
2. Bit 6 to toggle Brave, and
3. Bit 5 to toggle Faith.

0x80
0x40
0x20
0x10
^Would all be used, meaning you couldn't make any PA/MA/SP changes greater than 15.  Which isn't really a big deal, I think.
Why could this be important?  By consolidating as many formulas as possible, it frees up room for new formulas.  Unless, of course, these changes take up more room than all the formulas they would replace.  But still, this allows for quite a bit of customization...  

If you don't mind not being able to make changes to Brave or Faith greather than 31, then you could use in the Y:
1. You could use bit 7 as a toggle for positive or negative values,
2. Bit 6 could toggle PA,
3. Bit 5 could toggle MA, and
4. Bit 4 could toggle SP.


0x80
0x40
0x20
^So 0x00 to 0x1F are your limits.

Just this hack alone could replace formulas:
0x36
0x39
0x3a
0x3b
0x55
0x56
Along with half the Dance and Sing Skillsets.

So, why the heck didn't the game's programmers do this?  Typical crack-smoking Square programmers...

Later: Why not use the unused and unknown bit toggles in the "Learn with JP" Column?  Well, isn't that the freaking point of Vanya's little project here?  *Facepalm* I'm an idiot.

Still Later: Maybe the unused bits in the "Learn with JP" Column could be used for global, formula independent changes, and the unused bits in each formulas X and Y values could be used for smaller, more specific things?

EDIT:
More stuff.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on February 25, 2010, 11:39:43 pm
Bumping this with some guesses.

Status effects guesses derived by comparisons (I haven't tested these and I suspect 1, 5, and 6 can be switched until I know more):
Unknown 1: Unit will not return to squad?
Unknown 2: Ends on beginning of player's next AT
Unknown 3: Activates on end of player's AT
Unknown 4: Canceled by damage
Unknown 5: Cannot be reversed
Unknown 6: Panel can be stepped on?
Unknown 7:
Unknown 8:
Unknown 9: Affects (in the general sense) stats (accuracy, move, ability to act/cast, damage output)
Unknown 11:
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Vanya on February 26, 2010, 12:29:12 am
Interesting.
I think testing should begin with 2, 3 & 4 since they would be easy to verify.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on February 26, 2010, 05:21:05 pm
I'll have to retract 9.  Removing it from blind has no effect on accuracy or AI behavior (at least from what I can observe).
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on February 27, 2010, 04:10:54 am
Previous outdated comment:
I was FINALLY able to get the stupid AI to use a skill to absorb mana (like Spell Absorb) when he is actually out of mana (instead of using it as an offensive casting interrupt.  When he is in range of using a Normal Attack. Tested with a skillset that has a mana absorb skill, an attack skill that costs mana, and a healing skill that costs mana. I even made his job to give 1 PA multiplier and PA Growth = 200. With no weapon equipped... this guy still rather use a basic attack over his mana absorb skill, which lets him do a full screen attack that does huge amounts of damage, or heal himself to 100% HP.

Anyway, the trick is to have the first two flags after Learn on Hit? to be checked. The first flag we already know -- lets the AI actually use the skill. The second flag is usually reserved for special Self Target and Defensive abilities, like Golem, Moogle, Kiyomori, Murasame, etc. Looking in FFTPatcher, I see that Spell Absorb, Dark Sword, Ethers... etc do not have this checked. I have tested with all of these skills and the only thing he'll remotely use is Ether (usually to restore an ally's mana). Rarely uses Dark Sword or Spell Absorb when he has 0 mana. I wonder if having this flag checked will improve their performance in a mana-less situation.

As for exactly what does the flag stand for I'll never know. It's not very consistent and is almost never used outside of Summon and Samurai, and a few misc skills like Ramza's Scream. The only thing I know is that it probably gives the AI extra stuff to check for, mostly for defensive purposes. But at least I know now that this flag isn't exclusive to self-target skills.

Oh and I discovered another rediculously stupid feature of the AI. They never check for the target's PA or MA when they break stats. I haven't found a way to fix this yet. Tested with giving him only one skill that does nothing but break 20PA, 100% hit. He just spams this all day --on the same target-- rather than doing something actually useful.


edit: I believe the AI unknown flag that is immediately above Random Hits tells the AI to check cast time. Not fully confirmed but I'll try to test it out.


edit: There has to be something else. When the AI switches target he will use Dark Sword instead of the new mana absorb spell that I gave him, which is made far more superior to Dark Sword (Dark Sword gives him 22MP, the new spell gives him over 500MP). He chooses Dark Sword on certain targets... why? Both spells cost 0 MP to cast.

edit: Ok so I spent another couple of hours testing this.... and guess what? The stupid AI reverts back to his stupid --I don't care about MP-- mode. I have no idea why would this happen. It's not just about switching targets this time. It's a 1 v 1 and he absolutely refuses to use his MP Absorb spell once he's in the "Don't care about MP" loop. I have tested with the AI both in Critical and Non-Critical conditions as well. Same result. This means for a boss that relies on MP impossible to make correctly. He'll have to have spells that require 0 MP cost no matter what. I give up on this but if someone else wants to tackle this I'll be glad to help out.

(http://img203.imageshack.us/img203/8531/screenshot20100227at301.png)
I give up. You FAIL to steal MP, you @&$*@&$@. Even when you have only TWO skills to choose from.


Update: I got the AI to absorb MP again. Strange behavior. Here's the settings for the skill for future reference. Judging from the looks of these flags, it looks almost identical to Spell Absorb. I tested with the AI1 off (the flag below Undead Reverse), which is suppose to tell the AI to check the cast times, and the result becomes "don't care about MP". I am now suspecting that this might just be a random behavior, that's predetermined from the start of a battle.
(http://img46.imageshack.us/img46/564/screenshot20100227at525.png)
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on February 27, 2010, 02:57:28 pm
R999, from personal experience and from battle videos, it seems that AI priorities roughly correspond to something like:

Tier 1: Removing Golem
Tier 2: HP Healing
Tier 3: Incapacitating Status Effects (Charm, Weapon Break, Petrify, PA/MA Break) or HP Damage (including armor break, which counts since it reduces current HP)
Tier 4: MP Damage
Tier 5: MP Healing (since relatively few units use MP in the original)

As for that second flag, I've already said this in the post you made in the help section.  It's just the behavior flag that determines whether the AI will know that a given spell can't affect enemies.  Thus, you shouldn't give it to (an unmodified) spell absorb (or he'll have problems targeting enemies).
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on February 27, 2010, 03:56:46 pm
@formerdeathcorps,

Oops I must have missed that post you made. But I can confirm that that is not the case with regards to the second flag. The AI will still cast it on enemies even if checked (in fact, more so than unchecked).


In that list of priorities you have made, my "tester" job class only has access to two skills, one lets him do massive damage that costs MP, and one that steals a ton of MP off of a single target. He has almost nonexistent PA and does not carry a weapon. And he prioritizes attacking with bare fists over actually thinking about MP at all. I gave him Spell Absorb and Dark Sword as well (unmodified versions) along with the new spell absorb. He rather attack instead. Weird huh? In fact, I have seen him doing this consistently over and over and over. My tester job has high HP and innate Regen so he will never die from these 10 damage attacks. After watching them for 30 minutes they are both doing this (I put them in a 1v1 setting). Just attack and attack.


edit: I noticed something really cool today. You know the Holy Sword skillset in 1.3 are set to Evadeable right? Well, the way it works in 1.3 is, the AI will never use any of these Holy Sword skills unless they are out of range of their normal attacks, or if they can use an AoE. This is wrong by design. The AI should instead of closing in range, but try to stay as far away as possible and attack with Lightning Stab etc, just like how they used to in Vanilla FFT. I managed to get them to do just this, by removing all Shield/Mantle items off of their targets. There must be a corresponding flag that can change this behavior. Maybe there's an AI flag that controls whether to tell the AI to close in range or stay far away from their target. Or that there's a flag that tells the AI to check for evasion behavior differently.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on March 01, 2010, 04:27:19 am
Tested Unknown 4: Canceled by damage. Not true. Damnit. I thought it would be the case since only Confuse, Sleep and Charm share this attribute. I was trying to create a new Charm (cloned) status, but failed miserably. By the way, what does the Order do for the status effects? Is it the order applied? Like Applying Haste first and Slow second?

And, to my sincere disappointment, I think none of the flags do anything with the AI breaking Confuse/Sleep/Charm. I have tested applying the same flags from Confusion to Dark/Evil looking, the AI will always attack the targets with the status effect.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on March 01, 2010, 04:52:03 am
QuoteIs it the order applied? Like Applying Haste first and Slow second?
The negative status effects are always applied last.

QuoteI was trying to create a new Charm (cloned) status, but failed miserably.
I'm willing to bet these 16 flags are just the tip of the iceberg considering the behavior of each status effect.  For example, a unit with auto-berserk will not lose player control until he takes damage from the enemy.  Similarly, you retain control of your units when they start with invite (and they do not desert you; I'm willing to bet the same is true with charm).  It's likely that these effects only work if they are received (since an initially charmed unit in a 3 player battle has no default player to support).  From what I hear from other testers (I don't remember who or which topic it was), a unit who receives invite from enemies (after you change the flag in FFTPatcher to allow it) will desert you, but you still retain control of him and your battle will not end until the invited unit is defeated.  Furthermore, if you check the wiki, you'll find a section on status effects.  Some status effects have CT counters and others don't, but without ASM, you can't add change which status effects will have timers.  In other words, you can modify poison to last only 5 CT, 60 CT, or 0 CT (which means it never ends), but you can't constrain confuse or dark/evil looking to some fixed CT.  Try mixing some of these flags with the enemy actually casting these spells on you.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Skip Sandwich on March 01, 2010, 09:40:46 am
I wonder, what would happen if you had an ally unit that was set to AI control that was then invited by the enemy? Under normal circumstances you'd still control the unit, but you'd need to kill him to win the battle, but if the unit were under AI control, would they then turn on you? (as if charmed)

also, @R999

I think that particular flag doesn't control if a status CAN be canceled by hit, but just informs the AI that attacking that target will cancel status (not the AI is smart about doing that anyway, from how easy it is to get units with jump to kill their sleeping allies attempting to break sleep.)
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on March 02, 2010, 03:22:20 am
Thanks for the explanation guys. Today I just found an awesome way for the AI to break Charm/Confuse/Sleep. It's actually no secret though I am sure you'll know it too -- Give them Throw Stone. They love using this to break those status. I tested it they actually use it consistently. It might have to do with the conditions I gave them though (2 v 1 scenario, Barius Hill map).

As far as Status Effects flags go, I am thinking that it's probably best not to touch anything there. I wanted to add a new very short duration Don't Act status (uncurable by design). But at least it seems impossible for now. Makes me wonder if anyone was able to create any kind of new status effects at all (Assembly code hacking aside).
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on March 02, 2010, 01:18:16 pm
The AI will always use their lowest damage skill to break charm/sleep/confuse on their own units.  It's just that they won't target your units if your units have those status effects unless they're going for a KO.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on March 02, 2010, 08:26:24 pm
Quote from: "formerdeathcorps"The AI will always use their lowest damage skill to break charm/sleep/confuse on their own units.  It's just that they won't target your units if your units have those status effects unless they're going for a KO.

This makes sense to me, by design anyway. But if I actually remove Throw Stone, they are much less likely to not break the Confuse/Charm. I suppose that Death Sentence is still the biggest problem where the player can actually abuse.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on March 02, 2010, 08:39:04 pm
That's because there's probably code that predicts how much damage you're capable of.  If the AI calculates that throwing a shuriken on his unit for 100 damage + your units' attacks will finish that unit, the AI probably won't do it.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on March 12, 2010, 12:35:38 am
Going to bump this.

I figured out how range works.
If you have nothing checked, you must have greater than 3 range to attack and for each 2 height panels you are above your enemy, you will gain +1 range (just like a bow), but your damage is still limited to 1v3 unless you are equipping a gun or longbow.
Checking slashing sets maximum range to 1.
Checking lunging sets maximum range to 2 vertical.
Checking direct causes units in the way to get hurt, but you not to gain the range bonus.
Checking arc causes you to hit units 2 squares away if you aim at the third square behind them (projectile has arced trajectory).

EDIT:
The slots are fixed.  IN other words, if I modify leather helmet into a body slot item, even if I remove "head" to flag "body", I can still only equip the new item in the head slot.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on March 20, 2010, 12:33:13 am
Bumping this yet again.

I figured out Unknown 7th (Slot 12) under the status tab.  This determines whether or not an negative status effect is dispelled from a chocobo when you mount the bird (provided that you can mount the chocobo when it is inflicted with this status effect).
However, I suspect there's another function in the game that references these very same tabs because you can't mount berserked chocobos but berserk has this flag checked as well.  Removing this flag from poison causes poison not to be removed from chocobos, so I'm sure it's correct, but I don't think it's the intended purpose of this flag.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Xifanie on March 20, 2010, 08:07:45 am
why not? seems perfectly legit to me as the Immortal status protection is the 15th bit. I noticed a similar pattern for the surrounding bits and I didn't know what they could've been. Not all of them are used by the game though.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on March 21, 2010, 03:45:59 pm
From what I have tested today, Haste is always applied AFTER Slow (in a spell that gives Haste+Slow). Since they both can't stack on top of eachother, everyone will only have Haste (since Slow is cancelled by Haste). So the units affected will have Haste and not Slow. Now I am not sure if changing the "Order" in patcher will change this.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Vanya on March 21, 2010, 03:51:12 pm
Is there really any foreseeable circumstance where anyone would WANT to apply slow and haste?

BTW, is there confirmation for the 15th bit chocobo thing?
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on March 21, 2010, 06:01:11 pm
1. ^ It's a rare scenario, but I had this idea before: a spell that is intended to cast Haste on the caster and Slow on the others(by making caster immune to slow), making a dramatic difference in effective speed.

2. ^ yes formerdeathcorps has confirmed it.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on March 28, 2010, 03:38:06 am
A couple more discoveries:

1. Dead status kills the undead unless they are specifically immune to dead.
2. Cancel dead status has no effect on the undead, whether alive or fallen.
3. What DOES affect the undead and overrides the above are % HP healing, drain, or the wizard death formula (0E--no other % HP damage formula does this).  The effect of each is always reversed against the undead.  Thus, raise2 kills the undead not because it cancels dead status but because it heals for 100% HP, while death heals the undead for 100% HP because it deals 100% HP damage, not because it procs dead.
4. On its own, undead does allow anyone to absorb darkness elemental or allow anyone to become weak against holy.  That is a function of the undead race (skeletons and ghouls).
5. The trigger for counter is enemy is blade grasp flag + enemy is in range.  Aerostar was right in labeling attacks as "countergraspable".
6. Weapon formulas 00, 01, 02, and 05 function exactly the same except that 02 allows you to proc a spell.
7. Checking weapon strike means the if you use any of the above formulas and your weapon uses any of the above formulas, you will deal damage equal to the elemental damage of your weapon (if it is elemental).  Observe that guns use weapon formula 03 and 04 and thus should never be paired with any of the above weapon formulas unless weapon strike is unchecked.  Unchecking weapon strike will make the system assume you are dealing fist damage.
8. Every formula that references WP also references the elemental of your weapon unless the formula is NS (because it heals or drains blood).  I'm not sure if the ASM fix to holy sword also applies to formula 63 (SP*WP).  For these formulas, if you are barefisted, WP = 0.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on March 29, 2010, 04:04:21 am
Excellent findings.

I have 3 related questions and I assume you would have tested for these already,

1) Does the [weapon damage] formulas other than 2,3 and 4 inflict 25% status?  (not on my computer can't check now)

2) So it is then, impossible to have [weapon damage] + elemental UNLESS weapon strike is checked, correct? If weapon strike is unchecked, then there's 0 elemental bonus? So the only way to make a safe elemental weapon ranged skill would require the use of formula 3/4 + Weapon Strike, yes?

3) Now that we know what Countergrasp is, does that mean this can be applied to all abilities without restrictions to formulas? Like magic spells, status, etc,



edit:
(http://img406.imageshack.us/img406/7066/screenshot20100405at125.png)
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Pickle Girl Fanboy on April 05, 2010, 10:29:37 am
A few questions:
1.  What happens with formula 04 if you give a weapon fire, ice, and bolt elements?
2.  What happens with formula 04 when you put it on an ability and give the ability fire, bolt, and ice elements?
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Xifanie on April 05, 2010, 11:15:00 am
The last element has always priority:

In this case it's Ice

wind: ice1/2/3
nothing: ice1/2/3
fire: fire1/2/3
fire-lighting : bolt1/2/3
fire-ice : ice1/2/3
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on April 06, 2010, 06:25:18 am
Just a handy list mostly for convenience of the three remaining unknown bits on the right-hand side and all the corresponding status effects.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on April 09, 2010, 06:54:50 pm
The Unknown Flag D in the Learn JP section of the Abilities tab corresponds to the Miss animation. If it is flagged, and your attack is missed (tested for various formulas, mostly PA based, with damage and with no damage) the target will perform a step-aside animation. If not flagged, the target will simply stand still when the attempt is missed. This is all assuming that Evadeable (both the ability and the AI evadeable flags) are both unchecked. As far as I am concerned it has no influence on Weapon Guard/Shield Guard/Evasion animation. It'll be very apparent when you see one of your talk skills missed and the target steps-aside (that's how I discovered it at first).
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Vanya on April 09, 2010, 11:20:39 pm
Which one is D?
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on April 10, 2010, 01:40:57 am
The last one in the box containing Learn with JP, Action, AI Usable, etc.  Incidentally, this is one that's flagged for all physical attacks (which makes sense now) and not flagged for all magical attacks.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: R999 on April 10, 2010, 11:24:36 pm
(http://img696.imageshack.us/img696/3463/screenshot20100222at154.png)


(http://img59.imageshack.us/img59/1211/screenshot20100227at131.png)


from thread in Help section, which also contains updated descriptions for some of the flags.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Vanya on April 12, 2010, 12:02:14 am
Cool, I'll update my list on the first page with this info. Thanks.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: formerdeathcorps on May 05, 2010, 04:36:15 pm
I think I now know what normal attack does.  It makes it so if you are above the enemy you get +1 to vertical tolerance and if you are below, you get -1 to vertical tolerance.
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Vanya on May 05, 2010, 09:22:04 pm
Does that not happen with skills not set to "normal attack"?
Title: Re: FFTPatcher Blank & Unknown Research
Post by: Nirek on June 05, 2010, 08:15:47 pm
A B and C on that last chart appear to be:

Can be Used by AI
Hits only Allies
Hits only Enemies
The summons spell bear that out, though for some reason Fairy has the enemy only flag set, might be because it only targets dead units.