øAslickproductions.org/forum/index.php?PHPSESSID=5f0fck550j2m4m2fpbtkj2vkm1&action=printpage;topic=1980.0e:/My Web Sites/Slick Productions - FFIV Message Board/slickproductions.org/forum/indexcd5f.htmlslickproductions.org/forum/index.php?PHPSESSID=5f0fck550j2m4m2fpbtkj2vkm1&topic=1980.0e:/My Web Sites/Slick Productions - FFIV Message Board/slickproductions.org/forum/indexcd5f.html.zxÈ Print Page - FF5 Bugs & Glitches

Board of Slick

Library of the Ancients => Final Fantasy V Research & Development => Topic started by: Jorgur on March 07, 2011, 05:17:22 PM

Title: FF5 Bugs & Glitches
Post by: Jorgur on March 07, 2011, 05:17:22 PM
I thought the wiki would be the perfect place for this project so I created a page for it.
http://slickproductions.org/slickwiki/index.php/FF5_Bugs_and_Glitches

This page is free for everyone to edit, so please contribute if you can!
Title: Re: FF5 Bugs & Glitches
Post by: Djibriel on March 12, 2011, 06:05:04 PM
    I figured the best way to tackle this as a group is to have a quick list on the first page that lists all the bugs and their status. So I propose the following; I"ll keep this post updated insofar that it should contain all bugs discussed in this topic. All bugs are color-coded as far as their 'status' goes:

Unconfirmed
Cause unknown
Arguably a bug
Full extent of possible effects uncharted

All bugs/glitches that remain in black or obtain that much-coveted status as well-researched, easy-to-understand mistakes in the game could then be given a write-up by whoever feels like it. It sure sounds like an ambitious project, we'll see how far it goes :p

Last updated: 03/13/2011


Out of battle:
  • Desert of the Streaming Sands has no moving overworld tiles FFVA
  • Garula appears where he shouldn't (near Carwen)
  • Hidden out-of-battle timer in Tower of Magic does not stop when in the menu FFVA
  • Dash with the Pirate Ship and Flying Ship (maybe running helps you to better HOIST THE SAILS)
  • Versus Atomos, have Galuf as your lead party member. When Galuf exits your party to mourn Xezat, Galuf's sprite will still represent your party, so you'll have two Galuf's on-screen FFVSNES
  • After Exdeath confronts you for the first time in World 3, he zaps you away from Ghido's cave and after that you find yourself in the Library of the Ancients. Casting Teleport will take you back to the entrance of Ghido's Cave. You can now explore most of the world and its dungeons without Lenna. FFVSNES
  • In World 3, you can park your normal and black chocobo in the middle of the nine-squares forest that houses the Ancient Tree. You can walk out of the forest, and then you can't get back to your chocobo, trapping you and destroying your game forever. FFVSNES

Monster behavior:
  • Famed Mimic Gogo's AI script goes haywire upon aggro FFVA
  • Panthera's intended Stalker-like behavior instead causes it to self-terminate FFVA
  • Lots of bugs that can be caused by the fact that bits are set in-battle with a counter-attack, and these counter-attacks can be prevented by various means. Ramuh, Sekhmet, Archeoaevis, Exdeath all have troubles here. !Rapid Fire's last two hits don't provoke counters, the Monk's Counter ability doesn't provoke counters, casting less than all spells off of your own character doesn't provoke counters, weapon-induced MT magical attacks (Wind Slash katana, Gaia Hammer) don't provoke counters on all but the initial target of the physical attack, the list goes on. Also something with Chaos Cannon, I don't know the specifics.
  • Regen isn't set for Dragon Flower, Dragon Pod, Exdeath, Genji Shield Gilgamesh
  • Monster MP is a complete clusterfuck. Feed a monster with more than 9999 MP an Ether, and it resets to 9999. Give a monster with maximum 10 MP an Ether, and it'll get enough to cast Flash.
  • Ramuh, Sandworm, Catoblepas, Sekhmet, Exdeath (possibly others) have physicals that do not provoke Counters (okay) but also do not remove Sleep/Confuse/Singing (bug)

Attack behavior:
  • Vampire deals (maximum HP - current HP) damage rather than half that FFVA
  • Mix elemental properties (Resist Fire, etc.) and elemental up (Elemental Might) are removed upon equipment change
  • incapable of setting statuses from both 1 and 2 through special technique, only when statuses from 1 are protected against do status ailments from bit 2 appear. Examples: Skull Eater, Gil Turtle.
  • Casting Phoenix without the necessary MP (through !Dualcast, for instance), the firebird is stopped but the revival is not
  • Dualcast allows the caster to choose spells it has no MP for
  • death + encircle = alive with 0 HP. Omega can counter with consecutive Mustard Bomb and Encircle attacks. When a character is struck by both in succession, and if Mustard Bomb killed the character, the character will be alive after the battle with 0 HP.
  • The Chemist's !Recover doesn't cure Petrify, though it should
  • Pre-set Mute and in-battle cast Mute allow different things (Blue magic)
  • Negative Magic Power overflow (underflow?) bug: when giving Berserkers a Thornlet, the game will treat the Berserker as having a massive amount of Magic Power (99?). Fun with the Gaia Hammer and stuff. FFVSNES
  • Knives/Bows/Whips parameter bug. Now in orange because I'd like to re-evaluate the cause and effects of the bug, and the effects of the proposed fix. While the current formula seems wonky, the balance between the weapon groups seems fine to me. From what I understand, the proposed fix would about double their power on a character with equal Strength/Agility, making a Knife way stronger than a Sword. I'm not convinced that was really the intention, but I have no info at this point.
  • Power Drink adds to Goblin Punch, nothing more.
  • Odin's Zantetsuken appears to carve up Heavy targets when released from a Magic Lamp; most fun versus Neo Exdeath
  • !Analyze yields not all results intended
  • !Tame works on all BUT those with a Magic Beast nature, while the inverse is likely the intention FFVA[/color]
  • Kiss of Blessing sets Berserk to the Berserk-immune, shenanigans ensue.
  • !SwdSlap (or !Mineuchi) feels like it should be !Attack + Paralyze, but instead deals normal damage and keeps the target under Confuse/Sleep/Control.
  • Return spell does not reset in-battle timers such as those active in Karnak Castle or Odin FFVAUS
  • Moss Fungus and Gel Fish use the Banish spell that affects self upon !Release
  • Aqua Breath and Big Wave/Tsunamie (!Gaia attacks) are non-elemental rather than Water-elemental
  • You cannot learn Blue magic from attacks used by Released monsters
  • !Mix + !Dualcast = wonky targeting: both !Dualcast will target the second spell's target(s) (fun with Holy + Cura, for instance) FFVSNES
  • !Hide protects you long after you return FFVSNES
  • !Jump protects you long after you return FFVSNES
  • Kill yourself on the first turn of Quick may morph items held FFVSNES
  • Have a Ninja throw an empty slot (and some other stuff) to obtain 00-1 = 255 of a certain handheld item FFVSNES
  • Attack that set Zombie may strike a character who used !Hide FFVSNES
  • Zombie-setting attacks that strike dead Angel Ring characters revive instead
  • All-Berserker team can't run; Chicken Knife's !Flee will stop you from doing damage for the rest of the fight FFVSNES
  • !Controlling a target that is removed from battle (Mover) leaves the !Control uses catatonic

Item behavior:
  • Staff of Light/Judgment Staff deals double damage with !Mug , !Rapid Fire and !Jump
  • Mirage Vest renews Image upon equipment change.
  • Man-Eater deals double damage with !Jump
  • Gaia Bell is deals damage like a Long Reach Axe
  • Rune Chime works like other Rune Weapons
  • Rune Bow deals no damage, one-handed, randomly sets Silence...?
  • Brave Blade/Chicken Knife's Goblin Punch remains 255/0 even when listed Battle Power changes
  • Staff can be double-gripped
  • Multiple Wonder Wands use a single byte to keep track to which spells to use next

Event behavior:
  • Galuf's ghost in Sealed Castle; when you neglect to bring Galuf to the Sealed Castle, he'll talk when you waltz into the weapon chamber with Krile.
  • Lenna's calling spirit in Surgate Castle; go to Surgate when Lenna is in the Rift, and she'll still appear to get a line.
  • Gilgamesh may not appear versus Necrophobe even when all (known) criteria have been met (unconfirmed)

Miscellaneous:
  • Extra MP is lost after switching jobs and reequipping magic abilities FFVA
  • Blue Mage Bartz has messed up graphics when attacking with his left hand
  • Disrupt the Werewolves' ceremony by blocking their path; they'll go crazy
  • You can exit the slopes and walk on the air via a specific tile on Northern Mountain FFVSNES
  • Walk through walls in the Pyramid FFVSNES
  • !Hide from Reflective Zombie teammates, win all battles rather than run FFVSNES
[/list]
Title: Re: FF5 Bugs & Glitches
Post by: samurai goroh on March 12, 2011, 07:35:44 PM
- Garula appears where he shouldn't (near Carwen)
The reason is that world map battle formations are arranged in a 8x8 grid & the square touches the border of Carwen. It can be shown on my editor...

- Monster MP is a complete clusterfuck. Feed a monster with more than 9999 MP an Ether, and it resets to 9999. Give a monster with maximum 10 MP an Ether, and it'll get enough to cast Flash.
The reason is that EVERY monster has a fixed maximum Mp value of 9999.

- Casting Phoenix without the necessary MP (through !Dualcast, for instance), the firebird is stopped but the revival is not
IIRC, that particular magic is compounded of 2 spells, one of them costs 99 while the other cost 0 & that's why the revival is casted even if you don't have enough Mp...
Using Dualcast itself should be considered a bug since it allows you to choose a magic even when you only have enough magic for the first spell & not for the second. IIRC, when you mimic the Dualcast, you can then use both spells, even though you couldn't the first time. This is useful when playing a LLG :P

- Blue Mage Bartz has messed up graphics when attacking, best viewed with Barehanded.
To be more accurate, is when he uses his left hand to attack. His sprite is wrong...
Title: Re: FF5 Bugs & Glitches
Post by: Deathlike2 on March 12, 2011, 07:54:50 PM
- Observe

It was supposed to report MP IIRC, in addition to HP/weakness.

Quote
- MP bug

Switching classes while @ full MP would not retain full MP if the class increased the Magic Power of the character (like Knight->Summoner).

Quote
- timer?

2 different types:
1) There's a hidden timer at work when you must dual fight the bosses at the Tower of Power+Magic. It's supposed to be paused while in the menu, but in FF5A, the timer keeps going... so by the time you are ready, exiting the menu will cause game over.
2) Reset/Return doesn't reset the timer to when you started the battle. This is only applicable to FF5A US... FF5A Euro has it fixed. I assume the Jap version has two revisions.

Quote
Unfinished
- Swdslap

IIRC, it ignores the regular rules of a physical attack (like knocking a confused target to normal... kinda like FF4's Aim actually).

Out of battle:
- Desert of the Streaming Sands has no moving overworld tiles FFVA

It was mentioned somewhere it was swapped with other tiles... hence the issue.

Quote
- Garula appears where he shouldn't (near Carwen)

Seriously? Must be a tile battle location issue...

Quote
Monster behavior:
- Famed Mimic Gogo's AI script goes haywire upon aggro FFVA

I don't remember that happening... but that's probably because I do the standard way (it is interesting when you try to go after him anyways).


Quote
- Panthera's intended Stalker-like behavior instead causes it to self-terminate FFVA

I'm trying to recall which battle this is...

Quote
- Lots of bugs that can be caused by the fact that bits are set in-battle with a counter-attack, and these counter-attacks can be prevented by various means. Ramuh, Sekhmet, Archeoaevis, Exdeath all have troubles here. !Rapid Fire's last two hits don't provoke counters, the Monk's Counter ability doesn't provoke counters, casting less than all spells off of your own character doesn't provoke counters, weapon-induced MT magical attacks (Wind Slash katana, Gaia Hammer) don't provoke counters on all but the initial target of the physical attack, the list goes on. Also something with Chaos Cannon, I don't know the specifics.

I think FF6 addresses this better than FF5 (which was also an improvement when it comes to FF4) as counters are concerned. The evolution of the battle engine is notable. The counter issue with !Rapid Fire has to do with "poor enough" coding to simply deal with two-weapons (Ninja ability).


Quote
- Regen isn't set for Dragon Flower, Dragon Pod, Exdeath, Genji Shield Gilgamesh

I don't recall any monster having that attribute... so if it was meant to be there, it's an oversight.

Quote
- Monster MP is a complete clusterfuck. Feed a monster with more than 9999 MP an Ether, and it resets to 9999. Give a monster with maximum 10 MP an Ether, and it'll get enough to cast Flash.

I'd swear that the MP a monster had was almost always random. One time it would work.. other times it wouldn't... and from the same monster (the Forest of Mua monsters seemed to exhibit that the most).


Quote
Attack behavior:
- Vampire deals (maximum HP - current HP) damage rather than half that FFVA

I think it's a welcomed change... although it works against ???? so I guess that depends on how you feel about it.

Quote
- Mix elemental properties (Resist Fire, etc.) and elemental up (Elemental Might) are removed upon equipment change

I believe that was always some sort of issue when the equip via item screen in all the SNES FF versions... probably less so in FF6. A fun bug to exploit in FF4 was the Avenger Sword temporarily having the same properties of the previously equipped weapon (like the Blood Sword or the Ragnarok) when swapping weapons in battle.

Quote
- Casting Phoenix without the necessary MP (through !Dualcast, for instance), the firebird is stopped but the revival is not

IIRC, the best part about Phoenix was full maxMP on revival.

Quote
- The Chemist's !Recover doesn't cure Petrify, though it should

Really? Hmmm...

Quote
- Pre-set Mute and in-battle cast Mute allow different things (Blue magic)

Blue (Monster) Magic seemed to always follow a different set of rules in FF5 (like FF4) so I guess that has changed in FF6.

Quote
- Negative Magic Power overflow (underflow?) bug: when giving Berserkers a Thornlet, the game will treat the Berserker as having a massive amount of Magic Power (99?). Fun with the Gaia Hammer and stuff. FFVSNES

Nice. Titan would be proud.

Quote
Item behavior:
- Staff of Light/Judgment Staff deals double damage with !Mug

Is that against a Holy elemental enemy? Just checking.

Quote
- Mirage Vest renews Image upon equipment change.

Yea... explained earlier.

Quote
- A lot of weapons have strange side-effects because of lazy or strange coding. Man-Eater deals double damage with !Jump, two out of four Bells function like Long Reach Axes (Gaia Bell) or Front Row-only Rune weapons (Rune Bell).

Well... the Rune Bell is actually a very awesome weapon (Tinkerbell is better, like the Chicken Knife, but way more inconsistent). The critical hit is required to keep it on par with all the better swords+knives of the game... including the Chicken Knife+Ragnarok/Apocalypse.
Title: Re: FF5 Bugs & Glitches
Post by: Djibriel on March 13, 2011, 05:00:52 AM
This will be a long long post!

Garula walks near Carwen:
the World Map is divided in 64 squares, and since a bit of Carwen land occupies a square that's mostly near the Tower of Walse, it could not be stopped that he appears there. Cause known! I personally feel that the Carwen land there has no (narrative) purpose. They shouldn't have put it there if it caused NPC-ish monsters to transgress their boundaries like that. Does anybody feel it's not a bug? Im remember there being some discussion about Garula and if it was a single creatures or a race of Garulas; thos with the second theory might feel it's perfectly fine Garula walks near Carwen.

Monster's MP overflow:
It would have been fine if monsters had an MP cap of 9999, but according to their stat block they should have more. What happens if you use a Dark Ether versus, say, Omega on the first turn? Does it deal (75% of 60700) MP damage or (75% of 9999) MP damage?

Famed Mimic Gogo:
Almost all players will do one of two things; snooze until Gogo throws in the towel, or buff to da limit and then take down Gogo with one massive !Rapid Fire. This will not trigger the bug. You must do it wrong; attack him three times so that he starts his 'uncultured boors' speech, and he'll throw all kinds of weird in-battle text at you, eventually freezing the fight on most occassions.

Pantera:
From the Island Shrine chest, two foes may jump: Covert and Pantera. Pantera is a Bandersnatch upgrade that should pull a Stalker-like attack on you on its first turn. Two more Pantera's should appear, the text "Illusion!" should appear and then, you know, it's a guessing game. Pantera just dies when "Illusion!" is used, though.

Vampire:
It's an AWESOME spell now. It just heals you back to full and deals respectable damage in the process, damage that is non-elemental and barrier-piercing and everything. But a bug is a bug...

Staves + !Mug:
No. You'd think that the damage output would be erratic, or elemental properties would come into play, but it's not the case. I'm telling ya, those two staves deal double damage with !Mug. Other Staves (Sage's Staff, for instance) doesn't do this, nor do other magical attacks + !Mug (such as the Blood Sword).  :edit: They also deal double damage with !Jump.

Bells:
The Diamond Bell and Tinklebell deal damage based off of the caster's Agility and Magic Power, they're Back Row OK and don't function when Mute is in effect. Also, they're Song attacks, so they're listed as sound-based weapons. So that's what Bells are, you deal damage with them from a distance, because you're (and they're) MAGIC! The other two bells work off Strength, are not Back Row OK and work in Mute, so apparently you whack monsters with them, even though they're still Song weapons. The Rune Chime is a Rune weapon because all Rune-based weapons are the same; it's just lazy they didn't make a Rune Chime-unique routine. Why the Gaia Bell is a Long Reach Axe-type weapon, I don't know. Bells are attack type 39, LRA are 3A, maybe it was a simple typo?

Dunno why they felt the Rune Chime needed to be a MP-drived critical hit machine, its real value is the elemental boosts anyway.


Title: Re: FF5 Bugs & Glitches
Post by: Djibriel on March 13, 2011, 05:41:02 AM
All new bugs here listed also listed in my first post

- Dash with the Pirate Ship and Flying Ship (maybe running helps you to better HOIST THE SAILS)
- Moss Fungus and Gel Fish use the Banish spell that affects self upon !Release
- You cannot learn Blue magic from attacks used by Released monsters
- Disrupt the Werewolves' ceremony by blocking their path; they'll go crazy
- Versus Atomos, have Galuf as your lead party member. When Galuf exits your party to mourn Xezat, Galuf's sprite will still represent your party, so you'll have two Galuf's on-screen FFVSNES


Thanks to ebmid2's walkthrough on GameFAQs, the following list. I've not personally seen these bugs yet, so I can't comment on them too much.
http://db.gamefaqs.com/console/snes/file/final_fantasy_v_k.txt (http://db.gamefaqs.com/console/snes/file/final_fantasy_v_k.txt)

- !Mix + !Dualcast = wonky targeting: both !Dualcast will target the second spell's target(s) (fun with Holy + Cura, for instance) FFVSNES
- After Exdeath confronts you for the first time in World 3, he zaps you away from Ghido's cave and after that you find yourself in the Library of the Ancients. Casting Teleport will take you back to the entrance of Ghido's Cave. You can now explore most of the world and its dungeons without Lenna. FFVSNES
- In World 3, you can park your normal and black chocobo in the middle of the nine-squares forest that houses the Ancient Tree. You can walk out of the forest, and then you can't get back to your chocobo, trapping you and destroying your game forever. FFVSNES
- !Hide protects you long after you return FFVSNES
- !Jump protects you long after you return FFVSNES
- Kill yourself on the first turn of Quick may morph items held FFVSNES
- Have a Ninja throw an empty slot (and some other stuff) to obtain 00-1 = 255 of a certain handheld item FFVSNES
- Attack that set Zombie may strike a character who used !Hide FFVSNES
- Zombie-setting attacks that strike dead Angel Ring characters revive instead
- All-Berserker team can't run; Chicken Knife's !Flee will stop you from doing damage for the rest of the fight FFVSNES
- You can exit the slopes and walk on the air via a specific tile on Northern Mountain FFVSNES
- Walk through walls in the Pyramid FFVSNES
- Walk through walls in Istory Falls FFVSNES
Title: Re: FF5 Bugs & Glitches
Post by: Deathlike2 on March 13, 2011, 12:42:32 PM
    • Panthera's intended Stalker-like behavior instead causes it to self-terminate FFVA

    Does this happen to both FF5A versions? Chances are the Euro version would have corrected that...

    Quote
    • Knives/Bows/Whips parameter bug. Now in orange because I'd like to re-evaluate the cause and effects of the bug, and the effects of the proposed fix. While the current formula seems wonky, the balance between the weapon groups seems fine to me. From what I understand, the proposed fix would about double their power on a character with equal Strength/Agility, making a Knife way stronger than a Sword. I'm not convinced that was really the intention, but I have no info at this point.

    I get the feeling knives should be weaker than a sword if you're facing targets with low defense (like some sort of mage-type monster vs something like the Golems). Knives would then rule because of !Rapid Fire, which seems to make sense to me I guess.

    Quote
    • Odin's Zantetsuken appears to carve up Heavy targets when released from a Magic Lamp; most fun versus Neo Exdeath

    It might be running through some different code or taking the item's properties (ignore heavies since most of the summons should work outright) or some clear oversight.

    Quote
    • Aqua Breath and Big Wave/Tsunamie (!Gaia attacks) are non-elemental rather than Water-elemental

    Given that Aqua Breath works on a different formula (Desert monster killer), I can forgive that. Big Wave/Tsunami may be a holdover from FF4 in a sense... so Leviathan ends up being the natural water elemental... I think?

    Quote
    • !Hide protects you long after you return FFVSNES
    • !Jump protects you long after you return FFVSNES

    Sounds like some bad targeting coding... has happened a bit in variations (FF4 had a few when fighting Kain, FF6 has the jumping frogs and Super Ball I think).

    Quote
    • Kill yourself on the first turn of Quick may morph items held FFVSNES

    Never seen this, but sounds like fun. If we can get the Bloody Shield this way, it would be really cool.

    Quote
    • Have a Ninja throw an empty slot (and some other stuff) to obtain 00-1 = 255 of a certain handheld item FFVSNES

    I'm not sure how you'd reproduce that.

    Quote
    • Attack that set Zombie may strike a character who used !Hide FFVSNES
    • Zombie-setting attacks that strike dead Angel Ring characters revive instead

    The latter sounds like a feature, not a bug. ;)

    Quote
    • All-Berserker team can't run; Chicken Knife's !Flee will stop you from doing damage for the rest of the fight FFVSNES

    My guess is that you are never meant to run away if everyone has some sort of uncontrollable status (Charm, Sleep, Paralysis, Berserk), so the execution of !Flee must screw with that behavior.

    Quote
    • Rune Chime works like other Rune Weapons

    But... I like it!

    Quote
    • Rune Bow deals no damage, one-handed, randomly sets Silence...?

    I think it's intentional, but I'm trying to recall what circumstances you would want to use it... it smell of the Tower of Magic useful....

    Quote
    • Brave Blade/Chicken Knife's Goblin Punch remains 255/0 even when listed Battle Power changes

    I guess that's some oversight, but considering it is trying to reference the base number.. not the final number, it makes sense.

    Quote
    • Staff can be double-gripped

    One aspect of dbl gripping is using the Knight's STR stat (but why do that when you can go barehanded?) :P

    Quote
    • Multiple Wonder Wands use a single byte to keep track to which spells to use next

    Well, if one allowed many Cursed Shields in FF6, you wouldn't be wanting easy curse dispelling right? :P

    Quote
    • Gilgamesh may not appear versus Necrophobe even when all (known) criteria have been met (unconfirmed)

    My only guess is if you're tampering with the Berserk/counter code, you might must event data bits being set. That would explain that brother that haunts the innards of the Temple of Sands or whatever... that wouldn't go away until he's "officially" killed.

    Quote
    • Disrupt the Werewolves' ceremony by blocking their path; they'll go crazy

    Don't stop them from praying for you!
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on March 13, 2011, 07:45:10 PM
    Garula walks near Carwen:
    the World Map is divided in 64 squares, and since a bit of Carwen land occupies a square that's mostly near the Tower of Walse, it could not be stopped that he appears there. Cause known! I personally feel that the Carwen land there has no (narrative) purpose. They shouldn't have put it there if it caused NPC-ish monsters to transgress their boundaries like that. Does anybody feel it's not a bug? I remember there being some discussion about Garula and if it was a single creatures or a race of Garulas; thos with the second theory might feel it's perfectly fine Garula walks near Carwen.
    I kind of remember that an NPC mentioned he lived north of Walse Town. Carwen is surrounded by mountains, so you can argue that the Galuras shouldn't be there. You can use this on a LLG when you want to learn some abilities early on, so I do believe it was an oversight of the programmers, in other words, a bug.

    Monster's MP overflow:
    It would have been fine if monsters had an MP cap of 9999, but according to their stat block they should have more. What happens if you use a Dark Ether versus, say, Omega on the first turn? Does it deal (75% of 60700) MP damage or (75% of 9999) MP damage?
    Yea, the current MP of a monster can be greater than the max MP, just like when you use cheats yourself XD. I don't know about the Dark Ether, but using an Ether (or Elixir) when a monster has +9999 MP, it will set the MP to 9999. I think the max damage is capped to 9999, so I would believe that the Dark Ether would deal up to 9999 if the estimated damage gets over that...


    BTW, I remember a getting a bug that has never been mentioned before. It has been a lot of time since I did it, but it was when you're controlling a Mover & they change formation (instead of running away). IIRC, your character will be unable to anything, unless you make him loose control (e.g. attacking him) (he's still controlling a monster that is no longer there)

    Oh, not a bug per se, but more like an oversight is that the enemy Skull Eater can't be captured in order to release it... (Since it only has 1 Hp)
    Title: Re: FF5 Bugs & Glitches
    Post by: Djibriel on March 14, 2011, 07:05:42 AM
    - (Much) more info on all of ebmid's listed bugs can be found in his walkthrough I linked.

    - Dunno about Panthera in the EU release. I have no info on the Euro version whatsoever, that may be a fun game to pick up when the 1.1 battle strategy guide is done and the bugs 'nd glitches project is a little more advanced.

    - Since Odin normally won't appear unless all available targets are affected, the programmers were free to implement a full-screen effect to display the slicing and dicing. With the Magic Lamp however, you can summon Odin versus a screen full of Heavy targets (he won't appear) and a mix of Heavy and non-Heavy targets. An example is Neo Exdeath, whose bottom-left target is a lightweight. Odin comes to kill that target only, but because the animation calls for a full-screen effect, it will seem that all targets are killed even though they return a second later. The mistake here is one of combination. They should either have left Odin out of the Magic Lamp, have him perform Gungnir only or they should've coded a sprite-by-sprite animation for Zantetsuken.

    - Bah, I still say the Rune Chime needs its own formula, but lumping together of stuff can hardly be called a bug. It's just so widely illogical, I'm tempted to include a chapter "Batshit insane" in the Bugs 'nd Glitches document just to list those things we can't in good conscious call a bug, but want to anyway :p

    - I see your point about the Cursed Shld and the Wonder Wand, but there is an important difference. In a normal game, you can obtain one Cursed Shld maximum. But since the Wonder Wand can be stolen from Exdeath and equipped mid-battle, you can obtain two (in the SFamicom and PSX games) or 99 (in the GBA games). Either limit the Wonder Wand to a single weapon, or somehow keep track of the maximum of 8 Wonder Wands in use.

    - I seem to remember someone saying FFIII's terrain magic often was non-elemental despite elemental appearance, so through consistency, you could argue it's not a bug. I have no patience for NES games (except Battletoads) so I can't comment.

    - Added the !Control + run behavior in the list, unconfirmed as of yet

    - A very popular feature is the fact that !Hide + Zombie = win. When one character Hides and the rest is a Zombie, the fight will just go on. This can be used to your advantage via Wall Rings and weapons that randomly cast spells to win all battles. When one characters Hides and the other three are petrified or knocked out, the game will have you escape. I strongly feel that this should have been the case with Zombies as well, since Zombies count as 'dead' just like KO and Stone, and it's a filthy way of winning battles. Thoughts?

    -  I'm also including the fact that the quasi-physicals used by Ramuh, Sandworm, Exdeath, etc. do not remove Sleep/Confuse/Singing is a bug. My theory is that they wanted to create a physical attack that did not allow counters, and the rest was just unintended. !Sing is strategically superior versus Exdeath already since he can be so easily walled before he gets into the nasty stuff and the stat boosts carry over to the Neo Exdeath fight. Now, Exdeath is even incapable of smacking a Singing character back to reality like a measly Nut Eater does.
    Title: Re: FF5 Bugs & Glitches
    Post by: Deathlike2 on March 14, 2011, 11:43:46 AM
    - Dunno about Panthera in the EU release. I have no info on the Euro version whatsoever, that may be a fun game to pick up when the 1.1 battle strategy guide is done and the bugs 'nd glitches project is a little more advanced.

    I've looked into this... there's only a few possible explanations that I can gather.

    1) The formation used for this battle is not consistent with the original. Some formations are set up for a special purpose (generally hard-coded), so whatever the formation the monster is using is not from the original set. Of course... it's just possible that the current formation didn't have that special coding set...

    2) If I'm reading the algo doc correctly... it may or may not be a typo, but the code that is used to check if the original/real monster is dead is "different" than the one the Stalker uses... but that doesn't exactly explain the lack of use of "Escape" in the battle script.

    3) The most logical conclusion to me for that code being executed is that the monster is running the Stalker code bit check instead of its own. This could easily be overlooked in testing (since it is not a common formation and found once in a run). This is hinted by the monster's death (No Clue) which the original Stalker clones execute upon the real one's death. The problem is that the code is probably "simplistic" that the check is "if not real Stalker, kill self" and that the monsters (including the real Pantera) follows in that execution of that script.


    Quote
    - Bah, I still say the Rune Chime needs its own formula, but lumping together of stuff can hardly be called a bug. It's just so widely illogical, I'm tempted to include a chapter "Batshit insane" in the Bugs 'nd Glitches document just to list those things we can't in good conscious call a bug, but want to anyway :p

    I guess, but the Rune weaponry is underrated... but on the other hand, if the Rune Edge had the Swordblade compatibility... it would really really rock hard.

    Quote
    - I see your point about the Cursed Shld and the Wonder Wand, but there is an important difference. In a normal game, you can obtain one Cursed Shld maximum. But since the Wonder Wand can be stolen from Exdeath and equipped mid-battle, you can obtain two (in the SFamicom and PSX games) or 99 (in the GBA games). Either limit the Wonder Wand to a single weapon, or somehow keep track of the maximum of 8 Wonder Wands in use.

    I consider the Wonder Wand pretty niche and doesn't quite deserve that much attention.... this is the Wonder Wand after all.

    Quote
    - Added the !Control + run behavior in the list, unconfirmed as of yet

    I believe the Mover thing has happened to me once (I like Control, but the FF6 version is far more fun).

    Quote
    - A very popular feature is the fact that !Hide + Zombie = win. When one character Hides and the rest is a Zombie, the fight will just go on. This can be used to your advantage via Wall Rings and weapons that randomly cast spells to win all battles. When one characters Hides and the other three are petrified or knocked out, the game will have you escape. I strongly feel that this should have been the case with Zombies as well, since Zombies count as 'dead' just like KO and Stone, and it's a filthy way of winning battles. Thoughts?

    Zombies are technically "dead" for head counting purposes, but they are still targets on the field of play (for attacking themselves and others) so unless the cowardly one would like his/her brains being extracted... I think it's fine as is. Zombies want braaaaains!

    Quote
    -  I'm also including the fact that the quasi-physicals used by Ramuh, Sandworm, Exdeath, etc. do not remove Sleep/Confuse/Singing is a bug. My theory is that they wanted to create a physical attack that did not allow counters, and the rest was just unintended. !Sing is strategically superior versus Exdeath already since he can be so easily walled before he gets into the nasty stuff and the stat boosts carry over to the Neo Exdeath fight. Now, Exdeath is even incapable of smacking a Singing character back to reality like a measly Nut Eater does.

    It's a quick fix for something that was inherently broken in the original... but there are better ways to code the damned thing that didn't have to come to this... I would say.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on March 14, 2011, 01:08:12 PM
    I was wondering a few things with the bug list you have Djibriel:

    For the unconfirmed bugs, would only you mark them as confirmed?
    If that's the case, would reproducing the bug with the "movie" (like doing a TAS) be helpful? (I'd reproduced most of them before, so I know they do work)

    Are you planning to include bugs for the PS version?
    I've only tried a couple of the SNES ones & they are reproducible & the only bug I know is exclusive of the PS is the one with the Treasure %, where instead of displaying 100%, it'll display 099% (there are 3 digits instead of 1 or 2, so 99% <> 099%). I have some screenshots of this bug.

    Would you consider opening the menu, when you're in the world map, in order to avoid random battles a bug? I use it on the TAS I did & I think on the dual one too
    SNES Final Fantasy V (JPN) in 3:34:41.07 by Erick Guillen (samurai goroh). (http://tasvideos.org/942M.html)
    SNES Final Fantasy V & VI (JPN) in 3:52:15.97 by T.M. (neo_omegon). (http://tasvideos.org/1555M.html)

    There's no mention on the "classic" bug for the GBA version, the quick save that resets the seed of the RNG & lets you fight all types of formations relative quickly. (The part of resetting the steps for avoiding random battles may or may not be cataloged as a bug)
    Title: Re: FF5 Bugs & Glitches
    Post by: Deathlike2 on March 14, 2011, 01:10:50 PM
    If you can exploit a "feature", it's a bug.  :tongue:
    Title: Re: FF5 Bugs & Glitches
    Post by: Djibriel on March 14, 2011, 02:51:28 PM
    - I guess it's important to note that the Pantera battle works fine in the SFamicom game

    - I agree the Wonder Wand is an oft-overlooked feature, but since it's there it should work.

    - When one or more character(s) Hide(s) and the remaining party is a Zombie, the entire battlefield is unable to fight. It's not like !Jump where the character comes down regardless, a character can remain !Hidden for as long as he or she wants. In FFVI, there are no free run-aways; when one character is gone (via Sneeze/Snort or the like) and the rest is a Zombie, it's Game Over. So I still say bug.

    Whenever someone here says; I've seen this bug, let's roll, I take that as a confirmed. We're not in this business to lie. We're in this business for the ladiez. I mean the truth. If a video is found that'd be nice, but I'd still try to reproduce it myself.

    About the overworld in-and-out menu trick to keep away the random encounters: I wouldn't call it a bug, but it's certainly sloppy RNG coding that the player can exploit. It's not like FFVI's Slot command, where the player needs in-depth knowledge available only through direct code viewing to work the system; any fool can observe that Quicksaving leads to a fixed random monster encounter appearance.

    A similair thing is how monsters will never appear when you've just entered a room in a dungeon. I don't know how FFV works out when a battle is triggered, but it seems to be (partly) based on how many steps you've taken in a certain room. You could tackle 500 small rooms in a row and never get a random encounter just because you're entering and exiting all the time.

    Another form of sloppy coding that's not really a bug but still has major effects on the battlefield is that FFV loves to start at static ATB values. When you set Berserk or Confuse in-battle, the ATB is reset to that character's start value. This is also why you can fight such as fights as Archeoaevis and the Pages in the Library and never have them use any move; every time they call a new monster, that monster starts with a fresh ATB bar that has the entire way to go. Since you're already half-way with most characters, you'll always outspeed them and can destroy that form before it takes a move. The next creature befalls the same fate. Stuff like that isn't a bug, I think, when looking at the definition of the world, but it's certainly poorly executed.

    I'm a big fan of dropping the semantic nonsense about 'bug' and 'glitch' and just talk about intended behavior and unintended behavior. The Quicksave "feature" has no faulty coding and isn't a bug, but since it allows the player to seek out certain very rare monster formations every other fight, I'd say it's unintended behavior and thus worthy of being listed.

    As for the document itself, I feel a good example of how it (hopefully) can turn out is like Master ZED's Bugs 'nd Glitches guide for FFVI. It became a fairly big document, and the weigth of it eventually discouraged ZED from updating the thing when that faithful One-Hundred-Something Bugs topic was posted on mnrogar's forum back in the day.
    With that document I mind, I think it's best that we set up some sort of WIP document that Jorgur initiated in the first post of the forum (good idea!). Seeing the size of ZED's document however, I think it's suboptimal to embed it in the first post of this topic. I'd personally rather see a quick free wiki we all have access to, or a Google document or something. Since I've never done anything like that, any input or experience on the subject is welcome.

    Here's ZED's example:
    http://masterzed.cavesofnarshe.com/GameDocs/ff3bug.txt (http://masterzed.cavesofnarshe.com/GameDocs/ff3bug.txt)

    Jorgur asks about the scope of the document, let me post my thoughts:
    Bugs and stuff is fun for a tremendous amount of players. Some can be used for gain (Vampire bug, item duplication, stuff like that), others are better avoided (game freezing bugs, for instance) and most others deal with tremendously niche scenarios that are just fun to read about. While the majority of the posters on this forum are code monkeys, not everybody is. I think that showing were the code makes a mistake and listing possible patches or CodeBreaker fixes will alienate a large catagory of casual readers. Again, I think ZED's got a good vibe going on when he describes the cause in such a way my friends would understand the problem and its cause.
    Title: Re: FF5 Bugs & Glitches
    Post by: Deathlike2 on March 14, 2011, 03:16:17 PM
    - I guess it's important to note that the Pantera battle works fine in the SFamicom game

    Yea... I did the tests to make a proper comparison. Both US+Euro versions of FF5A exhibit the same bug. I'm just convinced the monster is using the Stalker code to execute... which is why that monster is a complete fail.

    Quote
    - When one or more character(s) Hide(s) and the remaining party is a Zombie, the entire battlefield is unable to fight. It's not like !Jump where the character comes down regardless, a character can remain !Hidden for as long as he or she wants. In FFVI, there are no free run-aways; when one character is gone (via Sneeze/Snort or the like) and the rest is a Zombie, it's Game Over. So I still say bug.

    Well... you have to remember that is a command that has been partly inherited from FF4... and Hide has many hidden bugs that would make one rage. For instance, you can cast Sylph in FF4A and a hidden Edward would receive healing... I mean.. there are states where someone has forgotten to test (there's also the Psycho Edward bug, but that's for another day).

    Quote
    Whenever someone here says; I've seen this bug, let's roll, I take that as a confirmed. We're not in this business to lie. We're in this business for the ladiez. I mean the truth. If a video is found that'd be nice, but I'd still try to reproduce it myself.

    Yea... although finding all the bugs can drive a person mad... mad I say!

    Quote
    Another form of sloppy coding that's not really a bug but still has major effects on the battlefield is that FFV loves to start at static ATB values. When you set Berserk or Confuse in-battle, the ATB is reset to that character's start value. This is also why you can fight such as fights as Archeoaevis and the Pages in the Library and never have them use any move; every time they call a new monster, that monster starts with a fresh ATB bar that has the entire way to go. Since you're already half-way with most characters, you'll always outspeed them and can destroy that form before it takes a move. The next creature befalls the same fate. Stuff like that isn't a bug, I think, when looking at the definition of the world, but it's certainly poorly executed.

    It's worth noticing these things in the progression of a battle engine between the series... and wondering WTF were they smoking sometimes. Of course partly some of these things were not totally finished....

    Quote
    As for the document itself, I feel a good example of how it (hopefully) can turn out is like Master ZED's Bugs 'nd Glitches guide for FFVI. It became a fairly big document, and the weigth of it eventually discouraged ZED from updating the thing when that faithful One-Hundred-Something Bugs topic was posted on mnrogar's forum back in the day.
    With that document I mind, I think it's best that we set up some sort of WIP document that Jorgur initiated in the first post of the forum (good idea!). Seeing the size of ZED's document however, I think it's suboptimal to embed it in the first post of this topic. I'd personally rather see a quick free wiki we all have access to, or a Google document or something. Since I've never done anything like that, any input or experience on the subject is welcome.

    Here's ZED's example:
    http://masterzed.cavesofnarshe.com/GameDocs/ff3bug.txt (http://masterzed.cavesofnarshe.com/GameDocs/ff3bug.txt)

    Jorgur asks about the scope of the document, let me post my thoughts:
    Bugs and stuff is fun for a tremendous amount of players. Some can be used for gain (Vampire bug, item duplication, stuff like that), others are better avoided (game freezing bugs, for instance) and most others deal with tremendously niche scenarios that are just fun to read about. While the majority of the posters on this forum are code monkeys, not everybody is. I think that showing were the code makes a mistake and listing possible patches or CodeBreaker fixes will alienate a large catagory of casual readers. Again, I think ZED's got a good vibe going on when he describes the cause in such a way my friends would understand the problem and its cause.

    Writing the document is always more painful than reading the intricacies. I'm sure someone will get around to writing it up in a doc. A wiki is only prone to the users that "think" they know what's going on.
    Title: Re: FF5 Bugs & Glitches
    Post by: Djibriel on March 16, 2011, 05:40:35 AM
    I pictured a somewhat closed wiki that only the active posters in this thread could access; if one of us goofed, the others could simply overrule him or her.

    In other news, Judgment Staff/Staff of Light also deals double damage with !Jump and gets no damage penalty from !Rapid Fire :p

    I fooled around with weapon-related oddities yesterday. I was most surprisd that the Twin Lance dealt double damage with !Jump it's a lance, it makes sense, but it's otherwise such a unique weapon I hadn't expected it to work LIKE a Lance with !Jump. It useless since !Jump removes it's double damage feature so you don't win any damage with !Jump, but still fun.

    The Blood Sword is such a strange weapon. It's a magic spell through the !Fight command; it's power isn't reduced by !Rapid Fire, nor is it increased by Sword Dance or !Focus. I half-expected the Blood Sword to share the Staves affinity for double-damage sources (since they are both magical damage routines), but 'tis not the case.
    Title: Re: FF5 Bugs & Glitches
    Post by: Deathlike2 on March 17, 2011, 12:04:19 AM
    I've compiled the posts into one post so you can feel free to clean up the duplicate.

    Original reply:
    In other news, Judgment Staff/Staff of Light also deals double damage with !Jump and gets no damage penalty from !Rapid Fire :p

    I'll have to look into it... is this FF5 or FF5A specific?


    Quote
    I fooled around with weapon-related oddities yesterday. I was most surprisd that the Twin Lance dealt double damage with !Jump it's a lance, it makes sense, but it's otherwise such a unique weapon I hadn't expected it to work LIKE a Lance with !Jump. It useless since !Jump removes it's double damage feature so you don't win any damage with !Jump, but still fun.

    I'll look into that... but my guess is that's it's someone's goof on that (blame TOSE for port fail).

    Quote
    The Blood Sword is such a strange weapon. It's a magic spell through the !Fight command; it's power isn't reduced by !Rapid Fire, nor is it increased by Sword Dance or !Focus. I half-expected the Blood Sword to share the Staves affinity for double-damage sources (since they are both magical damage routines), but 'tis not the case.

    The aforementioned Staffs run different code than the Blood Sword... the Blood Sword itself follows its own set of coding... so that's to be expected.

     :edit: - 2nd reply
    In other news, Judgment Staff/Staff of Light also deals double damage with !Jump and gets no damage penalty from !Rapid Fire :p

    Ok... I've confirmed it with FF5 and FF5A Euro (it'll apply to the US version probably as well). As I've said before... those two weapons use a different algo, so they share the same problem. The Staffs themselves deal less damage against Shell... I forget if other weapons follow that lead (I believe there are some).

    Quote
    I fooled around with weapon-related oddities yesterday. I was most surprisd that the Twin Lance dealt double damage with !Jump it's a lance, it makes sense, but it's otherwise such a unique weapon I hadn't expected it to work LIKE a Lance with !Jump. It useless since !Jump removes it's double damage feature so you don't win any damage with !Jump, but still fun.

    Twin Lance has the Spears property, so it's consistent in the 2x damage... while Double Lance itself is only Fight specific...

    Hey, if you want a fun related bug/feature... the Twin Lance doesn't cause the Necromancer to counter. It has a physical (Protect) and magic counter (Shell), but it doesn't trigger either.

     :edit: - 3rd reply
    Actually... all Rods, Judgment Staff+Light Staff, AND the Sage Staff ignore the attack reduction for Rapid Fire. It's simple to the fact in that the calculation for damage reduction is specifically geared towards weapons or specifically against Defense... not Magic Defense. This is also why the Blood Sword is consistent with this.

    Also... this wasn't actually obvious when digging... but the Dragoon Spear shares the same attribute as the Man Eater in the attack algo... this is why they have the same 2x Jump damage property. I think it has to do with lumping them in the racial damage component... they just forgot that the Man Eater isn't a spear...

    :edit: - 4th reply
    Sage Staff actually adheres to the 1/2 attack power rule ONLY against undead... seriously?

    Also... huge doc errata. You CAN dish critical hits via Jump (assuming the weapon you equip has a chance to dish a critical hit). The damage dealt is there, but there's no visual cue to show for it. Had fun doing this with the Tempest Katana and Yoichi's Bow.

     :edit:
    Crossed out the error.

     :edit: - 5th reply
    Sage Staff actually adheres to the 1/2 attack power rule ONLY against undead... seriously?

    That statement was tested on FF5. Apparently this behavior is "fixed" in FF5A (both US+Euro versions) as in it deals full damage instead of half damage.

     :edit:
    Jumping for critical hits was tested in FF5A and it works!

     :edit:
    Never mind... there is no bug with undead and the Sage Staff in FF5... :-/


     :edit: - 6th reply
    The Light Staff, Judgement Staff, AND the Blood Sword are all back row compatible by design. Since they use magic defense as part of the calculations and "magic" based attacks ignore row, those weapons are back row compatible. They do not necessarily need the back row bit set. FF5A does not inform you of this for the Blood Sword though. FF5 also doesn't tell you that the Light+Judgment Staff are back row compatible.

    Also, my guess to the primary reason why the Rune Bell isn't backward compatible in FF5 as stated is because it is following the Rune weapon type coding... which always takes back row into account. The weapon might have the bit set but the game won't take that into account.

     :edit:
    FF5A uses the same Rune weapon rules as FF5, so telling you that it is back row compatible is not entirely correct.

    A minor nitpick, but the Gaia Hammer in FF5A doesn't tell you the weapon is back row compatible.

     :edit: - First Post after the post merging
    With regards to the Regen status. Are we sure it's Regen that supposed to be applied and not some other status? In any case, the monsters you have listed (with the exception of the Genji Shield Gilgamesh) are the only ones with that attribute.

    Also, I'm unsure if the FF5A doc is correct about the Fairy Bow. 0x72 is for Creatures and this weapon doesn't deal with them. It is more likely to be 0x35. Even if that's not exactly the case, it does operate correctly as far as I can tell....

     :edit: - 2nd Edit after bleh
    According the the FF5A doc, the Mace of Zeus is stated to use 0x3A (meaning it uses the Long Reach Axes) algo. This is not correct. It's supposed to be 0x34, the Axe algo. Even in its description in FF5A does not mention it to be back row compatible.

    Also.... apparently the Fairy Bow attack property is correct, but it technically is not the one that should be used... which I don't think it matters too much but still noteworthy.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on March 17, 2011, 01:11:38 PM
    These are leftovers that may or may not be considered bugs, as you can't normally trigger them w/o using cheats or hacking the game...:

    Dummy Items: There are some items that can't be acquired in the game (perhaps maybe they can with the Quick morph bug, but never tried that), they include the Mythril Staff(#39), Sabre(#54), Cursed/Hematic Shield (#CE), Dummy (#EE), FF (#FF) and other unnamed/unused items...

    Dummy Terrain Backgrounds: There are 2 (IIRC) battle backgrounds (http://smg.photobucket.com/albums/v205/SamuraiGoroh/Final%20Fantasy%20V/SNES%20version/Terrains/) that are programmed but never used

    Dummy Enemies: There are some enemies that have name & sprite but cannot be fought. e.g. Big Boss(#FE) & NeoGoblin(#155)

    Dummy Commands: There are some commands leftovers, like Dummy1(#24) & Dummy2 (#27) (Check section 1.10 (http://www.erick.guillen.com.mx/Codes/SNES%20Final%20Fantasy%20V.txt) to know what I'm talking about :P)

    Dummy Rooms: There are some rooms that where planned, but didn't made it to see the light. e.g. Castle Tycoon - Dungeon(#56)


    :edit: I just remember another useless bug.

    When you see an NPC that likes to move, you can move yourself on the same place he's going to be (therefore having 2 characters on the same place). But, in order to work, you must be adjacent to him, you can't be in front & you only have 1 frame (1/60 sec) to pull it out. In other words, if he moves horizontally, you have to move vertically or vice versa.

    This is displayed on my TAS (which I've had liked on another post) when I'm on the Moogle Village or on the Great Forest of Mua...

    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on July 02, 2014, 01:00:00 PM
    When a character catches a monster, its name should be displayed on the "Stats" submenu, but only "blanks" are displayed instead in the RPGe v1.1 translation.

    To understand why is that happening, the monster names table was originally stored on D0/5C00 but now is stored on E0/0050.
    Moreover, the original size of monster names was 8 bytes long, but now is 10 bytes long.

    With that in mind I have found the code in charge of displaying these names:
    Code: [Select]
    C2/CA04: 68      pla ; Monster index
    C2/CA05: 0A      asl A ; .
    C2/CA06: 0A      asl A ; .
    C2/CA07: 0A      asl A ; *8 (must be *10)
    C2/CA08: 18      clc
    C2/CA09: 69005C  adc #$5C00 ; Monster name offset in bank D0 (must be 0050 in bank E0)
    C2/CA0C: A8      tay ; Y is offset to the monster
    C2/CA0D: A2C856  ldx #$56C8 ; X is pointer to the RAM tile to write the text (0x7E56C8)
    C2/CA10: A908D0  lda #$D008 ; aabb: aa = Bank, bb = #characters (must be -> E00A)
    C2/CA13: 209DE5  jsr $E59D ; Dump

    So, to fix the code, the addressing and the multiplication must be set to:
    Code: [Select]
    C2/CA05 22 00 5C D0 (jsr D05C00)
    C2/CA0A 50 00
    C2/CA11 0A E0

    [...]
    multiplyBy10:
    D0/5C00: 0A      asl
    D0/5C01: 48 pha ; Stack[1] is input*2
    D0/5C02: 0A      asl
    D0/5C03: 0A      asl ; A = input * 8
    D0/5C04: 18      clc
    D0/5C05: 6301 adc $01,S ; A = A + Stack[1] = input*8 + input*2 = input*10
    D0/5C07: FA      plx ; Free used stack
    D0/5C08: 18      clc
    D0/5C09: 6B      rtl

    The fix uses the addresses from D0/5C00 to D0/5C09 so it is not compatible with any patch using these addresses.
    I am attaching the patch as an IPS to apply over unheaded roms.

    I find strange that no one has noticed this before. Am I using an older or corrupt version of the translation?
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on July 05, 2014, 04:14:24 PM
    Good catch!

    I suspect the reason it hasn't been noticed is that it's because Catch/Release isn't something that used terribly often, and it's not difficult to remember which creature you've caught.
    Title: Re: FF5 Bugs & Glitches
    Post by: Grimoire LD on July 06, 2014, 08:21:34 PM
    Yes, it's my bet that the original translators, nor many people were even aware of your Captured Monster showing up in your stat screen. I certainly didn't know! Good fix though.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on July 10, 2014, 01:31:56 PM
    Here we go with another fix. This one is for the chicken berserk bug.
    This bug happens when the Chicken Knife is trying to run its special command (!Escape) and the team have only berserks.

    The code where a weapon check for its special command is here:
    Code: [Select]
    C2/4665: 20 A2 02    JSR $02A2   ; (0 to 99)
    C2/4668: A6 0E       LDX $0E
    C2/466A: DD 8F 40    CMP $408F,X
    C2/466D: B0 06       BCS $4675 ; Jump to not run the weapon special ability
    C2/466F: BD 90 40    LDA $4090,x ; A = Weapon special ability
    C2/4672: 4C D5 49    JMP $49D5   ; Force new command

    The fix then is overwriting the address C2/466F with a long jump to the patch:
    Code: [Select]
    C2/466F 5C XL XM XH JMP XXXXXX   ; jump to patch
    C2/4673 EA          NOP
    C2/4674 EA          NOP

    And the patch works as:
    1. If the weapon command is not !Escape, everything is right -> read weapon command and return
    2. If the weapon command is !Escape but the attacker is not a berserk, everything is right -> read !Escape and return
    3. If running away is forbidden in this combat, the command !Escape is not bugged -> read !Escape and return
    4. The bug can happen... -> Set the status 'Hidden' to every character, read !Hide and return

    When the 4 characters are hidden, the run away happens automatically avoiding the restriction of the berserk class.

    Code: [Select]
    0xBD, 0x90, 0x40,       /* LDA $4090,X *//* Read weapon special command */
    0xC9, 0x0A,             /* CMP #$0A    */
    0xD0, 0x0B,             /* BNE #$0B    *//* If (command != 'Escape') return command; */
    0xA6, 0x32,             /* LDX $32     */
    0xBD, 0x21, 0x20,       /* LDA $2021,X */
    0x29, 0x08,             /* AND #$08    */
    0xD0, 0x06,             /* BNE #$06    *//* If (!Attacker.Berserk) load A and return */
    0xA9, 0x0A,             /* LDA #$0A    *//* Run command !Escape */
    0x5C, 0xD5, 0x49, 0xC2, /* JMP $C249D5 */
    0xAD, 0xF0, 0x3E,       /* LDA $3EF0   *//* (Check if can run)   */
    0x30, 0xF5,             /* BMI #$F5    *//* Cannot run, so choose !Escape command */
    0xA9, 0x01,             /* LDA #$01    *//* Load 'Hidden' status */
    0x8D, 0x1D, 0x20,       /* STA $201D   *//* Set 'Hidden' to PC1  */
    0x8D, 0x9D, 0x20,       /* STA $209D   *//* Set 'Hidden' to PC2  */
    0x8D, 0x1D, 0x21,       /* STA $211D   *//* Set 'Hidden' to PC3  */
    0x8D, 0x9D, 0x21,       /* STA $219D   *//* Set 'Hidden' to PC4  */
    0xA9, 0x25,             /* LDA #$25    *//* Run command !Hide    */
    0x5C, 0xD5, 0x49, 0xC2, /* JMP $C249D5 */

    I am attaching the patch as an IPS to apply over unheaded roms.
    The fix patch uses the addresses from D0/2F05 to D0/2F30 so it is not compatible with any patch using these addresses. As long as I know, no patch in http://slickproductions.org/ff5patches.php share these addresses.
    Title: Re: FF5 Bugs & Glitches
    Post by: Digitsie on December 10, 2014, 02:54:55 PM
    noisecross's two patches aren't working.
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on December 10, 2014, 03:02:50 PM
    Quote from: Noisecross
    I am attaching the patch as an IPS to apply over unheaded roms.
    Is your ROM headered?
    Title: Re: FF5 Bugs & Glitches
    Post by: Digitsie on December 10, 2014, 03:33:34 PM
    Er, lemme rephrase that. His two downloads aren't working. :D

    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on December 10, 2014, 03:41:56 PM
    Crud, I don't seem to have them. I sent him a PM about it.

    For now, if anyone reading this has the files, please upload them here!
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on December 10, 2014, 07:33:22 PM
    Glad to see the forum is operational again!
    Here you are the patches again.
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on December 26, 2014, 02:04:24 PM
    I thought the wiki would be the perfect place for this project so I created a page for it.
    http://slickproductions.org/slickwiki/index.php/FF5_Bugs_and_Glitches
    Title: Re: FF5 Bugs & Glitches
    Post by: assassin on December 31, 2014, 01:58:20 PM
    damage types that call C2/8811 to finalize damage will multiply M by 2 if "Target Can't Evade vs. Attack type".  then if the attack is found to be "Magic Sword Drain", or if "Healing?" (elemental absorb, it seems) alongside a nonzero Variable $63, this higher M is used to recalculate damage or healing.

    damage types that call C2/8A05 to finalize damage do something similar.  M _can_ be multiplied by 2 if "Target Can't Evade vs. Attack type".  then if the attack is found to be "Healing?" (elemental absorb, it seems) alongside a nonzero Variable $63, this higher M is used to recalculate damage or healing.  the difference here is that M is only doubled IF it was found to be 0 and floored at 1.  so the "Target Can't Evade vs. Attack type" bonus is being applied VERY narrowly.  this makes me wonder whether the "BNE $8A26" at C2/8A0F shouldn't have been "BNE $8A13" instead.  if it's intentional, that raises the question why "REP #$20 / ASL $52 / SEP #$20" at C2/8A20 even needs the REP and SEP, if they know 16-bit $52 has to hold 0001h there.

    i don't expect the answer as to whether there's a bug to come quickly, as this M bonus doesn't seem well-known to begin with (i can't find it in the Algorithms/Stats FAQ).  if we determine exactly which C2/8811-using attacks it influences and how, maybe we'll be in a better position to judge what's going on in C2/8A05.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on January 30, 2015, 03:11:51 AM
    About the bug "!Tame works on all BUT those with a Magic Beast nature", I cannot reproduce it.
    The '!Tame' command set 'stop' status only in 'magic beast' monsters.
    The 'tame' whip effect set 'paralysis' status in every monster non inmune to that status.

    Am I missing something here?
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on January 30, 2015, 05:05:40 AM
    That's in the GBA version. All I know is that it doesn't work when it should.
    http://slickproductions.org/forum/index.php?topic=510.0

    I fixed it in the 2nd post.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 05, 2015, 05:26:06 AM
    I am attaching here a fix to:
    "You can exit the slopes and walk on the air via a specific tile on Northern Mountain FFVSNES"
    https://www.youtube.com/watch?v=1mVJpsWOP2s

    This patch only overwrites that scenario tilemap (CC/9396 to CC/96EB) so it is compatible with every other patch which doesn't edit maps. I will try to document in the wiki the little bit I have learned about maps.

    Have somebody documentation about that topic? I don't remember seeing anything.
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on February 06, 2015, 12:52:56 AM
    I sure haven't.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 23, 2015, 09:56:00 PM
    Made the Bugs & Glitches page (http://slickproductions.org/slickwiki/index.php/FF5_Bugs_and_Glitches) colorful to match what we have here on the first post (and updated some bugs to confirmed).

    BTW, there is a wikia page (http://finalfantasy.wikia.com/wiki/List_of_Bugs_and_Glitches) which has added a video (or a picture) of the bug in action. Any interest of doing the same here? I could create them, just would need an account to upload them, unless I use mine (https://www.youtube.com/user/SamuraiGoroh/) (which I have no problems doing so)...
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on February 24, 2015, 06:01:20 AM
    Is it really neccessary? Almost all of the bugs can be understood with text only, and there is really no space left to use.

    What if we make a sub-page for each bug/glitch? That would grant additional space where we can put technical details, images and/or videos.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 24, 2015, 09:32:09 AM
    By any means is not a must, but I think it would be helpful for people who are not familiar with them. (There are some that have vague information, like "Walk through walls in the Pyramid").

    I think we could name the bugs (just like in the wikia, replacing "description" to "name"), change the column for comments to link and then place the complete information for each bug in their corresponding sub-page. I guess we could name the pages as Category + #list (example: Miscellaneous_05 would link to the "Walk through walls in the Pyramid" bug)
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 24, 2015, 10:11:25 AM
    We have found a new bug in the RPGe code.
    After editing some texts, Digitsie noticed me a strange behaviour when the Abilities and Commands are displayed in the Abilities menu.
    The 'blank' slots after the learned abilites were not blank but the last displayed name repeated.

    The funny thing here is that issue is masked in the RPGe version of the game, so is never noticeable in-game. If the speech is edited otherwise there are probabilities to see it.

    The source of this thing is a non controlled index overflow. I am attaching here the explanation in a txt file.

    As long as we know, that will fix it:
    Code: [Select]
    E0/0F51 18 -> 38
    E0/0F57 71 -> 70
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 24, 2015, 10:14:11 PM
    Gilgamesh may not appear versus Necrophobe even when all (known) criteria have been met (unconfirmed)
    I was trying to reproduce the "Gilgamesh won't show at Necrofobia battle" and was able to identify how you trigger it.

    When you grab the (empty) treasure and trigger the battle against Gilgamesh at the end of the second world, there seem to be 3 cases:
    * Kill Gilgamesh w/o letting him transform
    * Kill Gilgamesh after he transform
    * Let Exdeath send Gilgamesh to the Void

    - For the first scenario, he won't appear at the entrance of the Void (nor against Necrophobia).

    - For the second scenario (which normally is hard to pull since you need to deal 55,000 damage) is what causes the bug.
    It seems that after he transform, a flag will register and he will appear later on at the Void. However, since it didn't triggered Exdeath's script a second flag won't be set (0x7E7C85 04). I suppose same scenario happens if you berserk him, though maybe I should try it just to confirm.

    - As mentioned above, after Exdeath sends Gilgamesh to the Void, another flag will activate, which is the one that is checked when you fight against Necrofobia.


    So, the problem relies on Gilgamesh appearing at the entrance of the Void when he is supposed to be death, which leads to the confusion as why he didn't appeared later on. (The other condition is that Necrofobia's HP must be below 9,999)

    I made a video of this bug (only the part where you fight him at the Void, not since the battle at world 2, because it would be too long)



    Edit: Youtube tags seems to display nothing, therefore rendering useless...
    Edit 2: Never mind, I had to use the short link, not the normal one...
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 25, 2015, 03:37:45 AM
    [...] the second scenario (which normally is hard to pull since you need to deal 55,000 damage) is what causes the bug. [...]

    So if I understand well, there are two ways to fix this issue:
      1- Forcing the first flag is not set in the transformation but at the same time than the second one (may cause bugs in that battle?)
      2- Forcing the flag checked to choose if Gilgamesh appears at the Void to be the second one (so the first one would be unused at all?)

    The second one seems easier, shorter and safer for me.

    How do you do to understand battle events? Just debugging with the emulator?
    I don't remember reading any documentation about them.
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on February 25, 2015, 07:01:35 AM
    I think we could name the bugs (just like in the wikia, replacing "description" to "name"), change the column for comments to link and then place the complete information for each bug in their corresponding sub-page. I guess we could name the pages as Category + #list (example: Miscellaneous_05 would link to the "Walk through walls in the Pyramid" bug)
    Yes, that's what I had in mind as well. I also think they should be named, but, if we give them numbers there will be complications (what to do when a bug/glitch is removed, moved to another category, etc.). I'm going to suggest that we simply give them names and nothing else, so the URLs become something like this:

    http://slickproductions.org/slickwiki/index.php/FF5_Bug:Walk_through_walls_in_the_Pyramid

    or if it is a glitch:
    http://slickproductions.org/slickwiki/index.php/FF5_Glitch:Walk_through_walls_in_the_Pyramid

    It's more convenient for us when we make changes to the bug/glitch list, and easier to manage in general. Of course we have to decide which entries are glitches and which ones are bugs.
    Thoughts?
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 25, 2015, 09:19:19 AM
    How do you do to understand battle events? Just debugging with the emulator?
    I don't remember reading any documentation about them.
    I have this addresses which I can monitor while playing in Snes9x

    7E7C85 XX          Records whether you've fought certain battles
           01 - ?
           02 - ?
           04 - Gilgamesh (Morphed - Exdeath Castle)
           08 - Sekmet
           10 - Golem
           20 - Golem
           40 - Shoat
           80 - Ramuh

    7E7C86 XX          Records whether you've fought certain battles
           01 - ?
           02 - ?
           04 - ?
           08 - ?
           10 - ?
           20 - ?
           40 - ?
           80 - Exdeath (solo battle vs. Galuf)

    7E7C85:04 is the flag checked at Necrofobia's battle. I also happen to have a mail from 2007 where RuneLancer (From Caves of Narshe) explained me the requirements for Gilgamesh to appear (which was around when the bug was reported)

    Quote
    First, here is the Necrophobia entity AI script's relevant parts.

    Code: [Select]
    NECROPHOBIA ACTION SCRIPTS
            If Var_02 = 0x01
                    Execute
                            Display Text(Necrophobia: Die!!)
                            Target Gilgamesh
                            Use Flare
                            Target Gilgamesh
                            Use (No-Target Magic 0xAC)

            If Var_01 = 0x01 AND Self is Alone
                    Use Destroy

            If Var_01 = 0x01
                    Target Gilgamesh
                    Use [Special, Fire 3, Bolt 3]
                    Target Gilgamesh
                    Use [Special, Ice 3, Critical]



    NECROPHOBIA REACTION SCRIPTS
            If Dead and Alone
                    Display Text(E... Ex...!!)
                    Idle

            If Damaged when (HP < 0x270F and Var_01 = 0x00) and Gilgamesh_Fought
                    Execute
                            Change Music to 0x22
                            Show Entity Gilgamesh
                            Display Text('Gil... Gilgamesh!!)
                            Display Text(Gilgamesh:  Phew!  I made it!)
                            Display Text('If I go home right now,)
                            Display Text('I'll remain in history)
                            Display Text('as a weakling!)
                            Display Text(Necrophobia:  Heh...)
                            Display Text('Blabbermouth...)
                            Display Text('I'll get rid of you first!)
                            Var_01 = 0x01
                            Idle

    When you attack Necrophobia, the first time he takes a hit while under 9999 HP Gilgamesh will appear. You have to have fought him and allowed him to leave the fight at the entrance of the Cleft of Dimensions. Note that because Necrophobia has a Death reaction script before the one that makes Gilgamesh appear, killing him will cause the Gilgamesh script to not run.

    During the fight, Necrophobia will target Gilgamesh solely. Once Gilgamesh is removed AFTER he has appeared (indicated by Var_01 if you haven't picked up on it yet) Necrophobia will remove himself from battle. At the end of the fight (when the Gilgamesh script sets Var_02 to 0x01) he will target Gilgamesh with Flare (note that this does not kill Gilgamesh) and use a "fake" attack to which Gilgamesh reacts, prompting him to use Exploder and to end the fight.

    In short, you have to hit him once his HP have been brought under 9999 without killing him after encountering Gilgamesh and letting him leave.

    As for a flag when Gilgamesh transforms, that's speculation right now because I don't have that address now, but I'm sure it's there somewhere...
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 25, 2015, 09:53:17 AM
    I think we could name the bugs (just like in the wikia, replacing "description" to "name"), change the column for comments to link and then place the complete information for each bug in their corresponding sub-page. I guess we could name the pages as Category + #list (example: Miscellaneous_05 would link to the "Walk through walls in the Pyramid" bug)
    Yes, that's what I had in mind as well. I also think they should be named, but, if we give them numbers there will be complications (what to do when a bug/glitch is removed, moved to another category, etc.). I'm going to suggest that we simply give them names and nothing else, so the URLs become something like this:

    http://slickproductions.org/slickwiki/index.php/FF5_Bug:Walk_through_walls_in_the_Pyramid

    or if it is a glitch:
    http://slickproductions.org/slickwiki/index.php/FF5_Glitch:Walk_through_walls_in_the_Pyramid

    It's more convenient for us when we make changes to the bug/glitch list, and easier to manage in general. Of course we have to decide which entries are glitches and which ones are bugs.
    Thoughts?
    The only problem I see with naming is that we should go for names that people have referred (so they know which bug were talking about). The problem lies that sometimes the name is not accurate to describe what it really is happening. For example: Skipping Linna glitch. While you're initially skipping her, the problem lies with the Time magic Return which allows you to continue your quest for a while triggering some other bugs in the process...

    As for bug vs glitch they are pretty much the same (programming errors). What I remember it was that glitches are more oriented to graphical/sound errors while bugs are oriented to bad programming missing cases (like the underflow bug that lets you have infinite items or 255 Mg power). I would say it's best to avoid technicalities and simple call all of them as bugs in the link. (Then inside the page, you can refer it as a glitch if needed)

    I haven't created a page yet on the wiki, so I don't know how easy is to rename them if needed (that's why I suggested to name them as "Category + #list", as links would hardly change unless you labeled wrong). And you can still refer them with the name given, it's just the page that would have such name.


    BTW, any feedback with the video?
    Like was the LUA script helpful or bothersome. Should I tweak it or remove it? etc...
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 25, 2015, 10:20:16 AM
    Thanks for the explanation.

    First, here is the Necrophobia entity AI script's relevant parts.

    That "AI script" looks pretty impressive!
    I remember reading about RuneLancer as well as about Silverlance when I was looking for FFV docs a couple of years ago. I think they are missing and their researches are not avaliable, am I right?

    By the way, talking about the "Walk through walls in the Pyramid", I have a fix for it using an experimental map editor, but I haven't test it yet.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 25, 2015, 10:39:17 AM
    RuneLancer & Silverlance are one and the same guy. IIRC, I once read he helped out instructrtrepe - J.L. Tseng - on the algorithm FAQ, so he's quite knowledgeable about the ROM. I have his e-mail although it has been a long time since I last contacted him (so I'm not sure if it's still active). I can PM it to you if you want it.

    He along with Caesar (both from Caves of Narshe) were making a hacking project  (http://www.cavesofnarshe.com/forums/ipb/index.php?showtopic=10524) for this game and had gathered plenty of information to do almost anything, but they never shared their investigation (he told me they had covered about 80% of the ROM cracked by that time I talked to him). He also told me that MAYBE he would if they released their project, but that never happened (and that was a big maybe, because people had used their information on other projects w/o giving credits).

    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 25, 2015, 11:11:41 AM
    It is a mess...
    I have named him because we are now "rediscovering" a lot of things which were documented some years ago.
    The location maps, in example, were researched and there was even an editor, but I think they focused highly on FFVI.

    The FFVI have a great locations editor as well as FFIV but I have not found any for FFV.
    Anyway, sorry for the :offtopic:
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on February 25, 2015, 11:17:57 AM
    The only problem I see with naming is that we should go for names that people have referred (so they know which bug were talking about). The problem lies that sometimes the name is not accurate to describe what it really is happening. For example: Skipping Linna glitch. While you're initially skipping her, the problem lies with the Time magic Return which allows you to continue your quest for a while triggering some other bugs in the process...
    I think we should go for accuracy. If a bug has an inappropriate name, we should just give it a new name, don't you think?

    As for bug vs glitch they are pretty much the same (programming errors). What I remember it was that glitches are more oriented to graphical/sound errors while bugs are oriented to bad programming missing cases (like the underflow bug that lets you have infinite items or 255 Mg power). I would say it's best to avoid technicalities and simple call all of them as bugs in the link. (Then inside the page, you can refer it as a glitch if needed)
    Ok, good point. I think that would be the best way to do it. So, should we rename the page from "FF5_Bugs_and_Glitches" to just "FF5_Bugs"?

    I haven't created a page yet on the wiki, so I don't know how easy is to rename them if needed (that's why I suggested to name them as "Category + #list", as links would hardly change unless you labeled wrong). And you can still refer them with the name given, it's just the page that would have such name.
    Renaming pages is easy. You simply use "Move" to give it a new name. Dead links is not an issue because the wiki automatically sets up a redirect so people get forwarded to the new page.

    As of now, we have 2 suggestions:
    http://slickproductions.org/slickwiki/index.php/FF5_Bug:Walk_through_walls_in_the_Pyramid
    http://slickproductions.org/slickwiki/index.php/FF5_Bug:Miscellaneous_05

    I prefer the first one because it looks nicer and is more descriptive. If you see that link somewhere you don't have to click the link to know which bug it links to. The second option will cause some problems if a bug is moved to another category.

    BTW, any feedback with the video?
    Like was the LUA script helpful or bothersome. Should I tweak it or remove it? etc...
    I'm not sure if that bug really needed a video. I watched for 5 minutes just to see that Gilgamesh did not appear.  :tongue:
    It was pretty well done though.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 25, 2015, 11:27:42 AM
    I prefer the first one [...]

    Me too for the same reasons.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 25, 2015, 05:11:27 PM
    With paging renaming not being an issue at all, then I vote for the first one:
    http://slickproductions.org/slickwiki/index.php/FF5_Bug:Walk_through_walls_in_the_Pyramid

    I'll try creating one and playing with it to see what information can be placed
    Title: Transforming Item bug
    Post by: samurai goroh on March 14, 2015, 01:08:52 AM
    I wanted to try and replicate the Transform Item bug (http://slickproductions.org/slickwiki/index.php/FF5_Bug:Transform_Item) (the main reason I wanted to do videos of the bugs). At first I had the impresion that the bug was hard to pull, but all that matters is the frame on when you kill the caster (And you have all the time you want to use after Quick). I made a quick script that replicated suiciding on the first attack just delaying the attack 1 frame each time and indeed it had about 25% chances of happening (in average every 2 of 8 frames, but not necessary in 1 out of 4 frames). I still have no clue what trigger this (as in why it works in some frames but not in others), but it seems is partially related to the RNG battle counters (7E003A, 7E003B), but that's just a coincidence.

    I tried equipping different items, but was only able to do it with the ones mentioned originally PLUS the Giant's Gloves (wonder why Ebmid missed this one), which also gives Vitality bonus (All but the White Robe gives +5, the robe gives +3). I also tried equipping the Bone Mail which takes away Vitality points (-5), but the bug could not be reproduced. I also tried switching 1 item at a time with no avail. So it seems only having 3 out of 5 possible items (Ribbon, White Robe / Angel Robe, Guard Ring / Giant's Gloves) is the key to reproduce the bug it seems, so it's a very limited bug. (The weapon equipped does not matter, just what you have equipped on the Left Hand which is going to be morphed matters)

    Here's the video in action:
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on March 30, 2015, 09:46:01 PM
    I was lurking in youtube and found a video with what it seems like a bug (unless is using cheats, but I doubt). It happens at 3:50

    Since is in Japanese, I can't tell exactly what happens, but from what I can tell is that Butz gets zombie status (Zombie Breath) and he's alive with 0 HP (he no longer has "near death" status). My guess is that it has equipped the Angel Ring and that's why the attack didn't killed him (Although if that's the case, he shouldn't be with 0 HP as far as I know, but could be wrong).

    Title: Re: FF5 Bugs & Glitches
    Post by: Djibriel on April 02, 2015, 01:49:35 PM
    Seems like Zombie is set. Zombie Breath sets Zombie if it kills. Butz has 0 HP and moves to attack his own party/himself. I see nothing wrong here (except for the Undead).

    If there is a bug here, it's the oversight that a full party, or at least all considered alive, can be Hidden in a boss fight simultaneously. Also, the 00 Doom counter on the Zombie should be removed. Or maybe it triggers as soon as you cure Zombie?

    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 02, 2015, 07:02:42 PM
    I'm sure Zombie should override other status, so the sprite should change accordingly. The algorithm FAQ says you're immune to a lot of status if you are a zombie, though it has no mention what if the other way around happens...

    So if a zombie frog has the look of a normal frog, that's a bad design IMO. Same goes if it happens with mini status.
    Title: Re: FF5 Bugs & Glitches
    Post by: Djibriel on April 04, 2015, 10:59:57 AM
    Zombies are immune to Darkness, Poison, Float, Mini and Toad, but characters with those status ailments are not immune to Zombie. Having Zombie remove those statuses would have been an understandable design choice, but they didn't make it.

    I seem to remember that eyes turn fully white when turned into a Zombie, so a Frog under the effects of Zombie is distinguishable from a 'normal' Frog. It's kind of hard to see in the video since there's no normal frog pose to compare it to.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 04, 2015, 08:00:51 PM
    I see, they have no pupils when they turn into frog zombies. Yea, hard to find if you're not familiar with it. Thanks Djibriel.

    So, when you are affected by zombie while being doomed, the countdown freezes instead of disappear once it reaches 0, huh. That doesn't seem right.
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on April 05, 2015, 06:12:54 PM
    Regarding the Item Transform Bug... I'm tooling around with it trying to find the cause.  I'm still in the setup phase, but I just had it happen to the item in the right hand - my Tempest became a Masamune (using two weapons).  Will post more as I find out more.

    EDIT: Looks like I figured it out.  I haven't written a fix, but it's easy enough to do.

    FF5's Main Battle Loop (MBL) can be found at C2/5872.  It runs every tick, and checks various conditions.  There are two variables we're interested in.  The first $7C51, which is a flag that notes if a character has the Quick status.  The other is $7C54, which is a flag that is set every four ticks, and for the purpose of this big handles HP Leak timing, among other things.

    Part of the MBL is a check for Quick status being nullfied (C2/627B) - say by killing yourself off.  This is where the bug is.  This clears the flag correctly, but then calls JSR $24E4.  The code is as follows:

    Code: [Select]
    C2/627B: AD 51 7C     LDA $7C51 If Quick Flag not set, skip this part
    C2/627E: F0 31        BEQ $62B1
    C2/6280: A5 0E        LDA $0E If Creature Number /= 7CB6, skip this
    C2/6282: CD B6 7C     CMP $7CB6
    C2/6285: D0 2A        BNE $62B1
    C2/6287: BD 1A 20     LDA $201A,X If Zombie, Petrify, Dead...
    C2/628A: 29 C2        AND #$C2
    C2/628C: D0 1D        BNE $62AB
    C2/628E: BD 1B 20     LDA $201B,X ...Berserk, Charm, Paralyze, Sleep...
    C2/6291: 29 78        AND #$78
    C2/6293: D0 16        BNE $62AB
    C2/6295: BD 1C 20     LDA $201C,X ...Stop...
    C2/6298: 29 10        AND #$10
    C2/629A: D0 0F        BNE $62AB
    C2/629C: BD 1D 20     LDA $201D,X ...Singing, Erased...
    C2/629F: 29 84        AND #$84
    C2/62A1: D0 08        BNE $62AB
    C2/62A3: BD 06 20     LDA $2006,X ...or HP = 0...
    C2/62A6: 1D 07 20     ORA $2007,X
    C2/62A9: D0 06        BNE $62B1
    C2/62AB: 9C 51 7C     STZ $7C51 ...Clear Quick Flag
    C2/62AE: 20 E4 24     JSR $24E4 ...and clear 3EAB array (#0C values)
    C2/24E4: 7B           TDC
    C2/24E5: AA           TAX
    C2/24E6: 9E AB 3E     STZ $3EAB,X
    C2/24E9: E8           INX
    C2/24EA: E0 0C 00     CPX #$000C
    C2/24ED: D0 F7        BNE $24E6
    C2/24EF: 60           RTS

    See the bug?

    24E4 overwrites the X register, which is supposed to be holding the character data offset.  It never gets saved/restored.  Oops.

    So X is now #000C, and the MBL continues as normal until we get to the HP Leak Check (C2/6301).
    Code: [Select]
    C2/6301: BD 1D 20     LDA $201D,X If HP Leak...
    C2/6304: 1D 73 20     ORA $2073,X
    C2/6307: 29 08        AND #$08
    C2/6309: F0 2C        BEQ $6337
    C2/630B: AD 54 7C     LDA $7C54 ...and Tick Flag set...
    C2/630E: F0 27        BEQ $6337
    C2/6310: BD 1E 20     LDA $201E,X ...Exit if Jumping...
    C2/6313: 29 10        AND #$10
    C2/6315: D0 20        BNE $6337
    C2/6317: C2 20        REP #$20
    C2/6319: BD 06 20     LDA $2006,X If HP /= 0, HP--
    C2/631C: F0 04        BEQ $6322
    C2/631E: 3A           DEC
    C2/631F: 9D 06 20     STA $2006,X
    C2/6322: 7B           TDC
    C2/6323: E2 20        SEP #$20
    C2/6325: BD 1D 20     LDA $201D,X If Hidden, min HP = 1
    C2/6328: 29 01        AND #$01
    C2/632A: F0 0B        BEQ $6337
    C2/632C: BD 06 20     LDA $2006,X
    C2/632F: 1D 07 20     ORA $2007,X

    Since X is never restored, it's now reading data off of the first character's data.  The first check (for HP Leak status) usually passes.  However, this code is only run every four ticks - hence the seemingly random nature of the bug.  Since it's now looking at weapon data, and since it's not usually going to be zero, it runs the DEX/STA, decrementing the weapon number. 

    The obvious fix is to make 24E4 PHX/PLX like it should have.
    Title: Re: FF5 Bugs & Glitches
    Post by: Grimoire LD on April 06, 2015, 01:56:30 PM
    Great find LightPhoenix! I'm surprised you were able to wade through the tick readings. I seem to have a bit of trouble with those.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 06, 2015, 03:58:14 PM
    Great work LightPhoenix.

    From the explanation I could understand why it happens and why it does in 1 out of 4 frames (ticks) in average.
    But do you know why did it only happened with the specific items equipped? (the items that boost vitality)
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on April 06, 2015, 09:54:21 PM
    From the explanation I could understand why it happens and why it does in 1 out of 4 frames (ticks) in average.
    But do you know why did it only happened with the specific items equipped? (the items that boost vitality)

    I'm not certain, but $201D,X, with an index of X = #000C points you at $2026, which is Vitality.  $207F is a variable that I'm not entirely sure about.  It's probably that equipping those items allows the initial comparison to be more likely to pass.  Perhaps #20 is often set on $207F, and would only be set on $201D with vitality boosting equipment?  Honestly, I didn't test that part very well.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 13, 2015, 07:22:02 PM
    I forgot about this bug until recently I wanted to reproduce a different one XD, I've named it the Shiny Dragon:
    https://youtu.be/zeGIFvjPerw

    The one I was trying is to get Boko stuck under the Fork Tower but I got it once and when I tried to optimize the video by leveling up the heroes, I could no longer reproduce it XD
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on June 27, 2015, 03:18:08 PM
    Watching some Fiesta streams I learned about a few bugs:

    First one is that when you are close to level up and are trying to learn a Lv spell, due to leveling up when you finish that battle, you won't be able to learn the magic. I have some vague memories that this happened to me on the Ancient Library when I got the GBA version long time ago. I would need to reproduce it, but I'm certain that this happens.

    Second one is a visual one, when enemies that have a Wall run out of magic while trying to reflect the spell back to your team, the wall animation will be triggering despite being unable to cast at all. An example would be the Wall Mage.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on July 01, 2015, 01:14:03 PM
    Also watching streams (and while playing) I noticed that sometimes when you revive a Berserker, it will immediately attack. It seems that after certain amount of time has passed since the Berserked died, the ATB will be set to 1 (Like it happens in this video (https://youtu.be/pNKKMWcAwK8), although here it's set immediate after getting zombied...)

    Also, I had long forgotten about skipping Kelb Town bug (https://youtu.be/QpejRH595mE). Thanks to EiriasValar for rediscovering it.
    Title: Re: FF5 Bugs & Glitches
    Post by: Grimoire LD on July 02, 2015, 01:32:45 AM
    Also watching streams (and while playing) I noticed that sometimes when you revive a Berserker, it will immediately attack. It seems that after certain amount of time has passed since the Berserked died, the ATB will be set to 1 (Like it happens in this video (https://youtu.be/pNKKMWcAwK8), although here it's set immediate after getting zombied...)

    Also, I had long forgotten about skipping Kelb Town bug (https://youtu.be/QpejRH595mE). Thanks to EiriasValar for rediscovering it.

    A similar bug exists in FFIV and allows speedrunners to skip right past Mist

    https://www.youtube.com/watch?v=cDeZ3tINq6Y
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on July 19, 2015, 05:23:47 PM
    I've been doing some bug-fixing research as a distraction from various projects.  Lately I've been looking at how swapping equipment removes elemental status granted by Mix.

    The bug itself is relatively straightforward to understand.  Elemental absorb/evade/immune/half/weak is stored in five bytes, each bit being one element.  In FF5, the elemental status granted by equipment is stored as an index to a table which holds the specific elemental values.  When you swap equipment (in battle or not) the game "rebuilds" the status by looping through the equipment, getting the associated elemental resistances, OR-ing them all together, and saving those five values to the character's elemental resistances.  The bug (in my opinion) is actually with Mix - it modifies those elemental resistances directly.  So when you swap items, the game doesn't know you're supposed to be under the effects of a Mixed potion, and thus you lose the buff.

    The fix, unfortunately, is much less straightforward.  One strategy is to catch when the game "unequips" the item, and before it does so XOR the old item's elemental resistances versus the current (to remove them) and then ORA instead of STA against the elemental resistance values.  This is less straightforward than at first blush, since you have to deal with a lot of edge cases (for example, what if you use Absorb Fire and then unequip a Flame Shield). 

    A better strategy would be to find an available byte in the RAM and use that as a "temp resistance" slot, holding an index to the table.  This, of course, has it's own host of issues, but ends up being more flexible.  Some decisions would need to be made regarding the implementation scheme.  You could just do one byte holding values for Elemental Absorb and simply OR it with 2030,X.  However, you could also make this byte an index to the aforementioned table of elemental resistances.  This lets you do some interesting stuff modding-wise: you can have a skill that makes a character weak to an element, for example.  Either way, the implementation is relatively easy - simply add whatever you're OR-ing to the function that recompiles the elemental resistance bytes (at C2/9C40).

    Anyway, I haven't coded any of this up.
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on July 19, 2015, 06:13:21 PM
    Separate reply, since it's on a different bug.  The Monster Special not setting Status 1 and 2 at the same time is a lot simpler (note: I believe this also affects setting HP Leak in Status 4).  The function (at C2/8C2F) calls the function to apply status 1 (at C2/8CAC).  I don't know exactly where, but the data pointer in $32 is getting overwritten during the 8CAC call with the target's pointer instead of the monster's pointer.  Luckily, this is a pretty simple fix - remove an LDX $32 and sub in a PHX/PLX.

    [EDIT] Now with patches!
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on July 21, 2015, 04:52:00 PM
    So, what was the intended function here in your opinion?

    The way I see it is that it was planned that since you can (only) change weapons and some have elemental affinity, it should update its elemental properties accordingly. The simple way for the programmers was to recalculate everything again (reuse same code). They likely realized that it affected the Mirage Vest and made it have only 1 illusion instead of 2, however mixing elemental affinities was a complete oversight so they missed it.

    So I would think that both the mix and the vest are bugs, one was worked-around while the other was missed, but both are bugs.


    BTW, what's the elemental evade? is it the "Can't Evade" certain type of magic or does it really exist a space for elements always hitting?
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on July 24, 2015, 09:59:07 PM
    Another bugfix update: I've mostly got the Death+Encircle bug figured out, but I don't have a good save to test it with so I haven't coded it.  The main battle loop has code for handling "dead" characters, including a check for handling dead and/or erased characters (at C2/6202).  The bug is that it first checks if the character is Erased, and if so it doesn't set the Dead status (whereas it does if you are not Erased).  This is because an Erased character is not necessarily dead.  This should check the character HP when erasing them to determine if they are dead.  The re-code is pretty simple and could be done in-line with a little consolidation.

    EDIT: I went ahead and coded it up, made patches, and zipped it up.  I have not tested this, and would appreciate it if someone with a readily available save could do so.

    Quote
    So, what was the intended function here in your opinion?

    The way I see it is that it was planned that since you can (only) change weapons and some have elemental affinity, it should update its elemental properties accordingly. The simple way for the programmers was to recalculate everything again (reuse same code). They likely realized that it affected the Mirage Vest and made it have only 1 illusion instead of 2, however mixing elemental affinities was a complete oversight so they missed it.

    So I would think that both the mix and the vest are bugs, one was worked-around while the other was missed, but both are bugs.

    My guess is that, in battle, it was meant for switching out shields.  You still need to check all the armors since you may not flat out lose an absorb from losing a shield (hence my edge case comment).  However, given the state of various bugs related to Mix commands, I think the Mix effect probably weren't tested all that well and this was an oversight.  You should probably be able to retain the benefit, but perhaps it should fade over time.  I say that with the acknowledgement that I almost never use Mix, so I don't know the specifics of how it works.

    Quote
    BTW, what's the elemental evade? is it the "Can't Evade" certain type of magic or does it really exist a space for elements always hitting?

    Elemental "evade" for lack of a better term isn't to my knowledge used, but I'd have to double check my notes to be 100% sure about.  The basic premise is that when you're immune, the attack still counts as a hit even if there's no damage.  The "evade" is basically the same as immune, but counts as a miss.  It's the byte between Absorb and Immunity (7E/2031).  In theory it would be used as a better form of immunity, where immunity only prevents damage and evade prevents damage and secondary effects since it's a miss.  However, I don't believe this ever ends up getting used.
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on July 25, 2015, 06:23:16 PM
    Another one bites the dust: I thought the Pray not healing Petrify fix was going to be simple.  Fixing it once I realized the bug was simple... finding it involved me derping a bit on how abilities work.  If I had remembered Revive works on all targets I would have found it quickly.   :isuck:

    For each effect in a spell/ability/etc, there's a flag at 7BEB,X (where X = the effect number) that indicates whether the effect should ignore statuses such as dead and petrified.  This is set for Revive and for Magic (ie Heal) but not for Pray.  Therefore Pray fails on a petrified target, even though it should heal them (Heal and Pray use the exact same attack formula/parameters).


    :edit: Can someone expand on the Mute bugs?  I'm looking into it since it's next on the list, but I'm not sure the exact issues.  Otherwise I can move on to the Magic Power Underflow bug.


    :edit: It wouldn't be bug-fixing if everything was straightforward.  I assumed the Magic Power Underflow bug was simply failing to account for setting something to zero.  Nope!  Close though.  All stat penalties are saved as negative numbers so that they can just be added, avoiding some extra branching.  The only problem is that the addition is 16-bit, and the high byte was being set to #00 instead of #FF.  Oops.  Luckily I was able to fix this inline by moving a REP instruction, due to some shenanigans that I'm not sure as to the purpose of.  The game was adding 1000 to each base stat before adding stuff in, and subtracting it once done, properly zeroing if there's an underflow.  Fuck if I know why it does that, but it's what made this possible.  Note that this probably affected all stat penalties, not just the Barbarian/Thornlet one.  I've uploaded the zip with asm/headered/unheadered patches.


    :edit: Another not-so-easy to fix situation.  Power Drink calls attack formula 51, which alters stats.  Unfortunately, the Attack Power it modifies (2044/2045) is for monsters, not players.  Relatively simple to fix, just changing some references.  Except now we run into the same problem as with the Elemental Stats - these get overwritten whenever equipment is swapped.  Ugh.  The best scenario, IMO, is to add 2044 to the attack power when it's calculated into $7A03; I'm not sure how extensive those changes are yet.  At least in this case we have bytes that aren't used elsewhere (2044 is only ever used for Goblin Punch, which is why it's the only skill affected).
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on July 29, 2015, 12:50:07 PM
    I'd tried to reproduce the Death + Circle bug before but I was to over-leveled for it to connect (due to how evasion works) or to weak to try to kill Omega in a quick manner to even try it... (I would need mixed levels, but I've been lazy to set it up :P)

    I heard about a similar bug, which I forgot to mention here. It would seem that you can die and be healed with regen at the same time, so you end up death with more than 0 HP. Since timing is crucial & can't remember if you needed to die in a specific way I hadn't mentioned before (probably easiest is to attack a partner)...

    As for the elemental evasion, there is a gap between absorb and immunity in the RAM that always bothered me, so it sounds reasonable it should be for that (despite is not used ever AFAIK).

    The mute bug is just present on the RPGe version IIRC. Normally there's a parameter in the enemy group formation that determines whether or not you can cast magic (like when fighting Magic Pot), which is labeled as "Casts Void at start" on my editor. What it does it to prevent both sides to cast magic. There's also the Time magic Mute/Void which ideally should silence both sides but it doesn't (There's also the white magic Mute, but that's single target and works fine...).
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on July 29, 2015, 11:31:53 PM
    I'm pretty confident the Death+Circle bug will be fixed with my patch, I'm just overly cautious.

    I'll add the regen/death bug to the list; it's probably in the same function.

    Neither of the Chemist Item Swap bugs (Elemental Status, Power Drink) are fixable inline.  I'm still debating what the best strategy will be here.  There's an unused attack formula at C2/8261 which I might be able to use, but I'm hesitant to do that.  I could do a long jump to another bank where there is free space, but that involves finding unused space (not as easy in FF5 as later games).  The other option is to work on optimizing some of the code - using JSRs to consolidate duplicate code.  That's not particularly ideal either.  Once I decide on a course of action, my plan is to use $2044,X to store Power Drink's bonus Attack Power and $2045,X to store Mix's added elemental effects.  I think I'm going to move on to some other bugs and come back to these with a clearer head.

    :edit: So in my mind, the actual bug here is the whole function at C2/9A6F that recalculates stats when equipment has been changed.  I've been working on this function to see if it can be optimized so I can free up space to make changes necessary for the Chemist Item Swap bugs.  As it turns out, there are a whole class of bugs that are essentially due to this function.  This covers both the Chemist Item Swap bugs, as well as various underflow and overflow bugs, most notably the Berserker Thornlet one.  It turns out that this function performs no apparent overflow check, so while I didn't see any listed, there are probably numerous instances of that.  This will be a lot of work, and while I'm still working on it, I've decided to move on to other bugs so I don't get burned out on it.



    The Odin-Magic Lamp-Zantetsuken bug is interesting not because it was hard to debug or is overly hard to fix, but that there are actually a few things here that contribute to the problem.  The first is simple - some bosses don't have the Heavy flag set, including Neo Exdeath Pt. 4 :eek:.  So part of this fix is to go through and set all of those.  That would (I believe) adequately fix the bug, since Zantetsuken's attack formula will "ignore" Heavy targets (they will animate being cleaved, but then come back).  The second problem is that Magic Lamp, unlike Summon, doesn't check for Heavy targets when casting Odin.  Therefore, Odin will never use Gungnir.  So, a simple fix is to make Odin use Gungnir instead of Zantetsuken - change D0/EEE9 from 6A to 6F.  This won't affect the actual summon.  A better fix is to have Magic Lamp check for Heavy targets... which is a 24 byte inclusion there is no room for.  Alternatively, Magic Lamp could just call Summon... but I'm not sure of the full implications of doing that.  A cursory inspection shows they are similar but not exactly the same.  So my plan is to audit the bosses, setting the Heavy flag that should be set on all of them.  That should fix the issue, but if it doesn't then I'll look into routing Magic Lamp into Summon to see what happens.



    So I think everyone knows the Scan Bug by now - C2/6C93 should be 10, not 08.  Simple fix, not going to actually release a patch on it.  What I will mention here for completeness is that Observe, Scan, and Analyze all work a little differently.  Observe only returns level and HP, Scan and Analyze both return everything but Analyze works on bosses and Scan does not.  That appears to be the intended behavior unlike some other bugs, so I won't be changing that.



    I'm still alive, but I took a couple weeks off from coding, but I'm back in the game now!

    SwdSlap, by all accounts, should have a chance to inflict paralysis on the target.  It does not.  All actions, whether commands or spells or whatever, have eight-byte data tables that roughly define how the ability works.  For commands, bytes three and four indicate action flags and damage modifiers that get set in 201E,X and 201F,X respectively.  For example, this is how Defend and Guard set their respective action flags.  SwdSlap sets the damage modifier flag (201F,X) to #02.  However, this is never used in the code; the function that parses these modifiers simply skips over it - there is no code to handle it.  Presumably this flag was to indicate that the attack should paralyze.

    One way to fix this would be to code handling for the flag.  However, the better way to do this is to emulate another command that combines a regular attack with a secondary effect - Mug/Capture.  In fact, what I think happened is that SwdSlap was meant to be converted to this method since it is more elegant, but got overlooked during production.  So this is what I'm working on, although it will require some work since C2 is rock solid as far as extra space.

    While I'm here, a word on Capture.  I've seen it mentioned on occasion that Mug/Capture is bugged since it doesn't use any other weapon effects.  This is intended behavior.  It deliberately calls a version of Fight (0816 versus 45FF) that ignores added abilities.  I believe the intent is that you either get added effects OR you get the special effect, and I believe this is why FF6 simply overwrites and/or ignores those effects as well.



    :edit: An easy one this time that I've actually made a patch for!  What is this madness?!  The Release-Banish bug was a simple one to identify and a simple one to fix.  Moss Fungus and Gel Fish were calling spell index E9, which was a version of XZone that for whatever reason only targeted the caster.  I pointed both of them to spell index 47, which is the Time Magic version of Banish/XZone other enemies use.  I've attached the zip of the (very simple) patch.

    Release doesn't parse the targeting flags completely, so targeting is limited to five options.  It only looks at the "Target All" flag (#40) and the "Default Enemy" flag (#08).  This produces four options: target all enemies (48), target all allies (40), target a random enemy (08), and target self (00).  You can also get target a random ally by setting any flag other than those two.  It might be an interesting project to expand this targeting so that there are more options for Release, but beyond the scope of this bug-fixing project.



    :edit: As always, bugs can end up being pretty interesting and not always straight forward to fix.  Fixing the various Water spells - Aqua Rake, Tsunami, and Big Tsunami - was still pretty straightforward, but Aqua Rake/Breath required a bit more ingenuity.  The Tsunami spells use attack formula 06 - Magic Attack - and simply didn't have the element flag (parameter 3) set.  It was a simple fix. 

    Aqua Breath (AB) on the other hand was a little more interesting.  AB doesn't always do Water damage throughout the series, so I debated whether it was intended behavior for it not to be water-elemental.  It uses a separate attack formula (6C) that does a non-elemental magic attack with bonus damage against certain creature types.  There are two possible solutions - one is to modify formula 6C to look at elemental damage, and the other is to modify formula 06 to look at creature type.  Both formulas do not utilize parameter 1, so either approach is possible.  Ultimately I decided on the second approach, since it involves only changing one other spell (Apollo Harp) and uses much less space.  So this patch changes formula 06 to check for creature type, and changes AB to use the basic magic attack formula.  This has a few consequences.  AB now suffers multi-target penalties, and can now be absorbed, blocked, and otherwise defended against like any normal magic attack.  However, AB now gains bonus damage from enemies weak against water in addition to damage against desert creatures.  I consider this a fair trade-off.
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on November 10, 2015, 11:37:20 PM
    Just a quick update on bugfixing:

    I have finally finished a large patch that fixes everything wrong with re-equiping items in battle.  I'm doing some final testing but it should be up by the end of the week.  It is not for the faint of heart.

    I also have a preliminary fix for the stupendously rare Quick + Mix + XMagic bug.  I'll go into more detail when I release it, but it has to do with the submenus and enemy selection.  In an amazing bit of luck, I was able to fix it inline - yay for crappy compiling!  I do have to test it, but I'm fairly certain this will fix it.

    Would it be appropriate for me to create my own thread at this point?
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on November 11, 2015, 02:28:53 AM
    Quote
    Would it be appropriate for me to create my own thread at this point?
    I also have a thread for my own patches, so I don't see why not.
    Title: Re: FF5 Bugs & Glitches
    Post by: C. V. Reynolds on December 10, 2015, 04:34:41 PM
    Hey, LightPhoenix. What's going on with that great-sounding patch you were talking about? Did something come up?

    Hope you don't think there's no interest. I love that you fix FFV's bugs. :)
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on December 13, 2015, 04:49:52 PM
    I've been busy with work (inspection + training new people) and the holidays, and frankly been a little burned out for the last month or so despite a small vacation over Thanksgiving.  What little coding time I've had I've been wrapped up in understanding and solving the three "casting" bugs (Quick/Mix/Dualcast Targeting, the Phoenix 0 MP bug, and the XMagic 0 MP bug).  Plus, honestly, I like the investigation part more than the actual coding/testing/collating part, so I tend to be lax on that end.

    I've tested the Equip rewrite, it's working well.  There are a few edge cases that still could use some testing.  It changes some stuff, so it's not strictly inline.

    The targeting bug was fixed inline (luckily), and was a bug in the menu code where a flag for Mix wasn't getting cleared for X-Magic if both were used in the same turn via Quick. 

    The Phoenix/0MP bug is (AFAIK) an edge case with Phoenix where a spell can have a "secondary" spell.  far as I'm aware, Phoenix is the only spell that uses this secondary spell method.  This uses a different mechanism than X-Magic, since you could in theory cast Phoenix twice.  When you cast a spell the basic flow is: the data is loaded, effects and animations loaded, and in another function altogether the MP check happens, and the effects and animations are altered after the fact.  There's no MP check in the actual casting function (although it *does* check Mute/Void/Toad).  The bug is that the MP Check doesn't account for Phoenix, so it never cancels the second part (which by this point is its own effect and animation).  I'm debating the best way to go about fixing it, but it won't be inline.

    The XMagic/0MP bug is simply that you can select two spells even if you don't have enough MP.  The code *does* catch this and prevents you from casting the second spell (unless it's Phoenix, see above).  However, this doesn't prevent Mimic from copying it, for example.  This is another bug in the menu code where a flag for if a spell can be cast is not recalculated between selections.  I think I have this solved (the recalc code is simple), but again it won't be inline so I have to figure out where to put it.

    Anyway, that's where I'm at.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on December 20, 2015, 03:04:56 PM
    Found a new bug while playing on the GBA. When you get to see the scene of the King Tycoon in Gohn and get the Airship, when you go back to that place usually an scene triggers, right? Well, you have 1 frame to land before triggering the scene and you can the watch again that scene with the King. The good thing is that the airship gets brought back, so you don't get stuck in there.

    It's most likely that it works on prior versions too, not sure about future ones (mobile).

    Forgot to mention it only works if you reach the town from the right like shown here:
    (http://img.photobucket.com/albums/v205/SamuraiGoroh/Final%20Fantasy%20V/GBA%20version/Gohn%20bug_1.png) (http://smg.photobucket.com/user/SamuraiGoroh/media/Final%20Fantasy%20V/GBA%20version/Gohn%20bug_1.png.html)
    Title: Re: FF5 Bugs & Glitches
    Post by: assassin on December 20, 2015, 05:55:11 PM
    regarding Capture: it's one thing to skip an add-on effect, like a spell.  it's something else to change the fundamental nature of the weapon strike itself just because it's being done at close range.  in FF6: why should Dice lose their custom damage formula, Drainer no longer drain, or Heal Rod suddenly become damaging, etc., just because you're trying to steal?  game difficulty isn't a suitable answer, as weapons' special effects do not necessarily make the strike more damaging to start with; for instance, the omission actually makes Atma Weapon stronger for much of the WoB.  that's in contrast with a bonus spell, which is almost always in your favor.

    throwing out weapon strikes' inherent characteristics is nonsense, and was easily fixed in FF6 by adding a second byte to temporarily hold an additional special effect ID, and making a second call.

    i don't know much FF5, but from what i've read on this matter before and what i'm reading in your post, the game is more measured and sensible than FF6 in what Capture throws out.  that is, it doesn't temporarily re-forge the damn weapons. :P  (it keeps their various damage formulas, correct?)

    pardon the lack of quoting..  the iPad's inability/refusal to highlight blocks of text does not mix with your giant post.  i'd be thrown off the forum for quoting the whole thing, so i'll just refer to it. :P
    Title: Re: FF5 Bugs & Glitches
    Post by: assassin on December 20, 2015, 08:08:33 PM
    samurai goroh: that's weird.. Does it matter how fast you walk into the town, or only how fast you land?

    I would've thought that the game would change the necessary event flags during the cutscene (preferably at the beginning).  Does the game somehow need you to leave a town to "finalize" any event bytes altered therein?  If so, maybe rapid screen changing interrupts that process..
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on December 21, 2015, 09:52:24 AM
    If you try to walk from above, there are 2 squares in the choke-point that will stop you. If you try to fly, most of the checkpoints are located as a square on the mountains, so you cannot land before the scene triggers.

    It only matters when you land because those checkpoints won't trigger once you're on the ground. The scene doesn't trigger because you did it to fast, same principle as when you want to skip steps in the world map in order to avoid monsters. You can hold the button and land w/o relying on been on time.

    Regarding the event flags, it would seem that the only thing that prevents it is not been able to enter the town (because it would disappear when you approach it). There are some other events there (Mid & Cid falling into the catapult & the boss fight) that will not trigger again.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 03, 2016, 06:26:18 AM
    I have found another little issue only happening in the RPGe translation at the very end of the credits.
    There is a scene where all characters perform an attack and after that is displayed the list of commands and abilities learned by every character.
    Is interesting, that no command or ability with more than 8 characters is displayed in that list. That is because the original japanese names are no longer than 8 bytes, but some of the RPGe names can be as long as 24 bytes.

    The routine which fill the buffer to display the names first reads and store the name in the buffer and after that fill the remaing buffer with blanks until the '8' bytes are filled.
    The problem is, if more than 8 bytes have been stored, the buffer is going to be filled by blanks in every position from the last byte to 255 (when an overflow happens) and after that from 0 to 8. In the process, every stored byte from 0 to 8 is blanked.

    You can check it. In example "MgcSwrdL4" is never displayed as adding "L4" to "MgcSwrd" sets 9 bytes.
    Some other abilities as "Learning spells" or "Equip swords" face the same behavior.

    To fix that, the easier way is to stop the blanks fill in 0x18 and not in 0x08. That is done by editing:
    Code: [Select]
    C1/31FE 08 => 18
    That is not enough because the texts doesn't fit as they are very shifted to the right side of the screen. To move the texts to the left the byte which marks the tile-X to display these lines is stored at:
    Code: [Select]
    D0/E164 15 => 11
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on February 03, 2016, 09:52:07 AM
    I never encounter that bug because I don't use RPGe's translation (was it DeJap's the other translation?). I saw it like a year ago in a speedrun where someone used I believe Chemist as Krile & they seemed to be familiar with it.

    On other news, it seems that there's still the infinite item steal bug in FFVA. The steps to reproduce them are slightly different, but the setup is the same (you need Throw and Steal). It was discovered by the Japanese community (no idea who did it) and brought into me by neo_omegon. I haven't played much with it and don't really understand why it works, but it does.

    You can read about it here. I'll try (eventually) to look at it and see why it works. There's also a second bug (also explained there) that lets you steal a different item, I don't get it quite why it steals such item even if claims it uses the enemy's special attack ID (which doesn't seem to me):
    http://tasvideos.org/forum/viewtopic.php?p=425968#425968
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on February 03, 2016, 10:43:39 AM
    I never encounter that bug because I don't use RPGe's translation (was it DeJap's the other translation?)

    Yes, the DeJap translation doesn't have ability names longer than 8 bytes (at least in that scene) so every ability is correctly displayed.
    On the other hand the "MgcSwrdLX" command text still disappears.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on May 05, 2016, 01:02:03 AM
    I haven't tried this one yet, but wanted to mention it in case someone knows about it.

    Magic usually has certain percentage to connect, unless the "Always Hit" flag is on. There's Remora's magic Latch On which has a 100% of hitting (which doesn't immediately translate to always hit). I find it funny that is not 99% nor 128 (always hit). The catch here is that in a Speedrun, you use Remora in order to paralyze Titan & while he's level 1 I've heard you can miss the attack.


    Code: [Select]
    6.2.2) HIT DETERMINATION FOR MAGICAL ATTACKS
    Hit% = Spell Hit% + Attacker Level - Target Level
    Evade% = Target Magic Evade%
    ...
    Let N1 = (0..99) 
        If N1 >= Hit%, the attack misses;  do not follow any more steps.
    Let N2 = (0..99) 
        If N2 < Evade%, the attack misses;  do not follow any more steps.

    Titan has level 1 with no magic evade (and obviously can be affected by paralyze). I think in a speedrun you should be around level 13 or so... Unless the Hit% drops to 99 (no 3 digit number) or N1 can get to 100, I can't see why it would fail at all.


    Note: Faen Fantom & Chocobo Kick also have a 100% instead of always hit.
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on May 05, 2016, 04:11:59 AM
    If Titan has Shell, then Hit% = Hit% / 2 (6.3.6)
    Also if somehow you manage to summon it as multi-target (it should not be possible ...) you get same effect (for hit%) (6.3.3)

    P.S. Out of curiosity, why will you use Paralyze on Heavy enemy?
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on May 05, 2016, 09:32:42 AM
    Since is a speedrun strategy, you paralyze Titan with Latch On so that in the moment he dies, he can't use his death script (Earth Shaker). Titan doesn't start with Shell and since it's a speedrun, you don't inflict it to begin with (hence why I posted it under the bugs topic).

    However, looking at what you mention (Heavy status), that might be the reason it "doesn't work". Is not that it fails (AFAIK) but rather that the duration of the status gets reduced from 240 to 30.
    Code: [Select]
    If the Target is Heavy type, Status Duration = 30;  do not follow any more steps.
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on May 05, 2016, 09:51:28 AM
    Yeah that was my point, Duration = 30 (vs heavy) which is VERY short. I tried before to use Stop on Omega ... it was terrible - you need 2 chars dedicated to Stop, you need haste, battle speed reduced to slowest speed (in menu) ... and still it was very, very hard to keep him in Stop :D
    I like to kill super-bosses like Omega in non standard way :D

    P.S How about Float to counter Earth Shaker?
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on May 05, 2016, 10:15:44 AM
    You would need to control/confuse a Gaelicat for that, which takes time to prepare for it. The other strat is to jump with the Dragoon's ability so that you evade it that way (conceal works too, but you need lots of ABP before you learn it, so not viable). Can't remember if there's a setting that allows you to withstand the damage with those levels, but you usually want the ABP for jobs anyways.

    I'll try to talk to the speedrunner (EiriasValar) to see what really happened. I had the impression that the attacked missed, but looking at the data it probably was that the effect just ran too fast to work.


    Edit: I can't remember if I'd mentioned or not, but the infinite item steal glitch does work for the GBA. Since today I have the day off, I want to look at it. For the steps, they are mentioned here:
    http://tasvideos.org/forum/viewtopic.php?p=425968#425968

    Edit 2: Here's the video EiriasValar made, around 25s is where the magic misses (no animation):
    https://youtu.be/Hj71ZCrmJq8
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on May 13, 2016, 07:38:01 PM
    Here's one to add, how about the fact that you can land the Airship or the Hiryuu on the entrance to Guido's Cave? If you do so, it will become permanently inaccessible. In the case of the Airship, it will become impossible to finish the game. Full details here (http://finalfantasy.wikia.com/wiki/Airship_and_Hiryuu_glitch).
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on May 14, 2016, 12:31:44 PM
    Not only the airship


    BTW, it's already mentioned in the wiki page
    http://slickproductions.org/slickwiki/index.php/FF5_Bugs_and_Glitches
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on May 14, 2016, 05:07:31 PM
    OK. I was fishing a bit; it just kind of annoys me that no one's made a patch to say "you can't land anything on the entrance to Guido's Cave". And I don't know how to edit that myself.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on May 17, 2016, 09:59:57 AM
    OK. I was fishing a bit; it just kind of annoys me that no one's made a patch to say "you can't land anything on the entrance to Guido's Cave". And I don't know how to edit that myself.

    Well... Samurai Goroh, before me, and a lot of people before us tried to understand how the locations and world map work but there is very little information about the topic.

    You can start reading the wiki (http://slickproductions.org/slickwiki/index.php/Noisecross:Final_Fantasy_V_Maps) and continue the research from the point where it is now.
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on May 17, 2016, 03:37:08 PM
    OK. I was fishing a bit; it just kind of annoys me that no one's made a patch to say "you can't land anything on the entrance to Guido's Cave". And I don't know how to edit that myself.
    Do you mean editing the wiki? It is quite easy, just click on the Edit link and change the code as desired.
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on May 23, 2016, 11:02:19 PM
    OK. I was fishing a bit; it just kind of annoys me that no one's made a patch to say "you can't land anything on the entrance to Guido's Cave". And I don't know how to edit that myself.
    Do you mean editing the wiki? It is quite easy, just click on the Edit link and change the code as desired.

    No, I mean there has to be some bit, or combination of bits, somewhere in the ROM that is/are wrongly configured, allowing the player to land stuff on Guido's Cave. The problem, as noisecross points out, is that no one knows where that is yet. That's what bugs me right now.

    I can help out to some degree, if it means finding the problem faster. I see that noisecross appears to have narrowed things down a tad.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on May 24, 2016, 06:24:42 AM
    It is configured that way because you can see an event with the Chocobo (where Butz mentions that is too dangerous to enter the cave, so he should stay back). The problem is that because of that, all vehicles are allowed to land on that spot, so is not like you can just change it to any other cave.

    Mmm, can't remember how the Pirate's Cave is handled now that I think of it. (Can't remember how it goes once you permanently get the Chocobo...)
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on May 24, 2016, 10:36:55 AM
    It is configured that way because you can see an event with the Chocobo (where Butz mentions that is too dangerous to enter the cave, so he should stay back). The problem is that because of that, all vehicles are allowed to land on that spot, so is not like you can just change it to any other cave.

    Correct me if I am wrong but I think the event you mention happens in one "location" map and not in the world map.
    The location id is 0x0012. I think the world map and the locations have a different behaviour so maybe it would be possible to block all vehicles excepting yellow chocobo to cross/land on that tile, right?

    By the way, what we do know about "events"?
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on May 24, 2016, 04:16:03 PM
    Events are triggered when you walk into an specific tile, be it in the world map or dungeons/rooms. They are placed in choke-points (or simply place several of them to create the same effect) in order for you to step on them.

    There are several tiles in the world map, like when you start a game and enter a narrow path to fight Goblins x2. The tile remains, but the event flag gets marked, hence why it only triggers the first time you walk into. This make what is called "Danger Steps counter" not to advance on that particular tile, which throws off-guard many speedrunners when they want to avoid enemy encounters...
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on May 25, 2016, 05:08:31 AM
    Sure. My question is, how are these events stored on the rom? Do we know how to read into or hack them?

    Also, how the edges and the passable/impassable properties are managed?
    When I researched about that, I was only able to get to what I wrote in the wiki (http://slickproductions.org/slickwiki/index.php/Noisecross:Final_Fantasy_V_Maps#Tile_properties).
    Have anybody more information about this topic?
    Title: Re: FF5 Bugs & Glitches
    Post by: x0_000 on May 25, 2016, 02:49:44 PM
    I did a bit of research into it for the GBA version when I was looking up how to give the Gladiator job early. I haven't had time to organize them into something easily understandable but here's the notes:

    Code: [Select]
    080DA9A0 - updates bitflags for jobs
    080DF4F8 - calls above
    08144c98 - job IDs?
    080dc3b0
    080dccb8 - eventer?
    083d6065 - add job event? (starts at 012f21/083d5a79, ID 4ec)
    080dfd2c - event updater
    08144b70 - size of event data
    08708508 - event table?

    083c08d4 - pointer to event array?
    083c2b58 - start of event data?
    0x8a0 events

    083C78AE - first event of the game

    080DFCC0 - get event pointer
    080DFA9C - raise event flag?
    080DA330 - talk events?
    08142C58 - talk event table
    associates events with NPC event data
    080B1A10 - NPC identification?
    080B1280 - start of function
    0200EE2C - NPC data?
    0200ED80 - saves NPC event data here
    080B1C6A - loads npc data?
    0200EACC - room ID
    0200e2ae - also room ID
    0200E3AD - NPC count

    0200e3a2 - writes event ID to this location

    08024f46 - updates room data?
    08136760 - room data table? (28 bytes per entry)
    0811DC34 - ??? (1 byte per room ID)
    0811de64 - ??? (1 byte per room ID) affects room layering

    080AC366 - locates room transition data
    080AC2EA - searches through room transition data.
    080AC35A - updates room data through transition data here.
    0813CA84 - room door list index
    0813CECC - room door table (12 bytes per entry)
    @0x0 - X position of exit
    @0x1 - Y position of exit
    @0x2 - room of exit
    @0x3 - ?
    @0x4 - ?
    @0x5 - ?
    @0x6 - direction to face when exiting
    @0x7 - X position of room
    @0x8 - Y position of room
    @0xa - ?
    @0xb - ?

    ffva npc setup
    NPC data is stored at:
    083B6514 - table starting points (NPCs/room)
    values are relative to this pointer
    083B695C - data points
    7 bytes per NPC
    @ 0x0 (2) - identifier
    @ 0x2 (1) - sprite ID
    @ 0x3 (1) - X position
    @ 0x4 (1) - Y position
    @ 0x5 (1) - Walking parameter (affects how the NPC moves)
    @ 0x6 (1) - extra parameters (1-3 affect palette, 0x10 determines layer, 0x20, 0x40, 0x60 determines direction)

    when the player interacts with an NPC, the game looks up the reaction data in:
    083BE268 - table starting points
    values are relative to this pointer
    083BEB68 - table data
    format: 0xf1-0xff is a control code
    0xf0 is the generic text ID
    0813C901 - amount of data to advance for each code
    0xf1 - 00
    0xf2 - 03
    0xf3 - 02
    0xf4 - 02
    0xf5 - 04
    0xf6 - 01
    0xf7 - 02
    0xf3 - 04
    0xf9 - 05
    0xfa - 04
    0xfb - 02
    run the next event if event ID wasn't raised
    0xfc - 02
    0xfd - 02
    0xfe - 02
    0xff - 03
    events to run?

    if an event is called, the game looks up the event ID using the 0xfb ID in:
    08142C58 - talk event table

    I also used this to rip some of the data from the game, for example this giant block of data is used in the game when the player gets the Wind Shrine jobs; each line is a specific command with parameters, and at the end I made a small comment on the event code that unlocks jobs:
    Code: [Select]
    .byte 0x31,0xfe
    .byte 0x13
    .byte 0x13
    .byte 0x13
    .byte 0x40,0x00
    .byte 0xb8,0x83
    .byte 0xb8,0x85
    .byte 0xb8,0x8e
    .byte 0xb8,0x84
    .byte 0x1a
    .byte 0x13
    .byte 0x64,0x16,0x00,0x1c,0x00
    .byte 0xa9,0x84
    .byte 0xaa,0x8e
    .byte 0xa8,0x83
    .byte 0xab,0x84
    .byte 0xab,0x8e
    .byte 0xab,0x83
    .byte 0x1e
    .byte 0x5c,0xf1,0x80
    .byte 0x21
    .byte 0x48,0x02
    .byte 0x5a,0xe0
    .byte 0xb2,0x8e
    .byte 0xae,0x83
    .byte 0xae,0x85
    .byte 0xb2,0x84
    .byte 0x20
    .byte 0x48,0x02
    .byte 0x5a,0xe0
    .byte 0x1e
    .byte 0x48,0x02
    .byte 0x5a,0xe0
    .byte 0xb0,0x8e
    .byte 0xb0,0x83
    .byte 0xb0,0x85
    .byte 0xb0,0x84
    .byte 0x20
    .byte 0x5c,0xf2,0x80
    .byte 0x20
    .byte 0x59,0x06
    .byte 0x21
    .byte 0x6a,0x00,0x20,0x00,0x68,0x4d,0xd8
    .byte 0x25
    .byte 0x58,0x06
    .byte 0x62,0x14,0x5d,0x00,0x5e,0x00
    .byte 0x16
    .byte 0x1e
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x23
    .byte 0x6a,0x91,0x00,0x00,0x15,0x14,0x00
    .byte 0x61,0xee,0x04
    .byte 0xb6,0x95
    .byte 0xb7,0x95
    .byte 0x4d,0xc0,0x03,0x01
    .byte 0x9d,0x95,0x15,0x14,0x00
    .byte 0x58,0x06
    .byte 0x22
    .byte 0xb8,0x95
    .byte 0x20
    .byte 0xc3,0x95
    .byte 0x62,0x09,0x88,0x00,0x8a,0x00
    .byte 0xa7,0x95
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x6a,0x1f,0x00,0x00,0x2d,0x15,0x00
    .byte 0x2a
    .byte 0x61,0xef,0x04
    .byte 0x9d,0x80,0x2d,0x1d,0x00
    .byte 0x9d,0x81,0x2d,0x1c,0x00
    .byte 0x9d,0x82,0x2d,0x1d,0x00
    .byte 0x58,0x06
    .byte 0x20
    .byte 0xb8,0x80
    .byte 0xa7,0x80
    .byte 0xb8,0x81
    .byte 0xb8,0x82
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2d,0x1b,0x00
    .byte 0x9d,0x82,0x2d,0x1c,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2d,0x1a,0x00
    .byte 0x9d,0x82,0x2d,0x19,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2d,0x19,0x00
    .byte 0x9d,0x82,0x2d,0x1a,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2d,0x18,0x00
    .byte 0x9d,0x82,0x2d,0x19,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2d,0x17,0x00
    .byte 0x9d,0x82,0x2d,0x18,0x00
    .byte 0xb9,0x80
    .byte 0x1d
    .byte 0x9d,0x82,0x2d,0x17,0x00
    .byte 0xb9,0x81
    .byte 0x1d
    .byte 0xb9,0x82
    .byte 0xea,0x84
    .byte 0x1e
    .byte 0x5a,0x20
    .byte 0x1d
    .byte 0x5a,0x20
    .byte 0x1d
    .byte 0x5a,0x20
    .byte 0x1e
    .byte 0x5c,0xf3,0x00
    .byte 0xb0,0x84
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x6a,0x00,0x20,0x00,0xb0,0x58,0xd8
    .byte 0x25
    .byte 0x58,0x06
    .byte 0x62,0x14,0x1f,0x01,0x20,0x01
    .byte 0x14
    .byte 0x1e
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x23
    .byte 0x6a,0x73,0x00,0x00,0x08,0x07,0x00
    .byte 0x61,0xee,0x04
    .byte 0xb6,0x90
    .byte 0xb7,0x90
    .byte 0x4d,0x60,0x03,0x01
    .byte 0x9d,0x90,0x08,0x07,0x00
    .byte 0xb9,0x84
    .byte 0xb9,0x85
    .byte 0x58,0x06
    .byte 0x22
    .byte 0xb8,0x90
    .byte 0x20
    .byte 0xc3,0x90
    .byte 0x62,0x09,0x4e,0x01,0x50,0x01
    .byte 0xa7,0x90
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x6a,0x1f,0x00,0x00,0x2d,0x15,0x00
    .byte 0x2a
    .byte 0x61,0xef,0x04
    .byte 0x9d,0x80,0x2e,0x1d,0x00
    .byte 0x9d,0x81,0x2e,0x1c,0x00
    .byte 0x9d,0x82,0x2e,0x1d,0x00
    .byte 0x58,0x06
    .byte 0x20
    .byte 0xb8,0x80
    .byte 0xa7,0x80
    .byte 0xb8,0x81
    .byte 0xb8,0x82
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2e,0x1b,0x00
    .byte 0x9d,0x82,0x2e,0x1c,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2e,0x1a,0x00
    .byte 0x9d,0x82,0x2e,0x19,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2e,0x19,0x00
    .byte 0x9d,0x82,0x2e,0x1a,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2e,0x18,0x00
    .byte 0x9d,0x82,0x2e,0x19,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2e,0x17,0x00
    .byte 0x9d,0x82,0x2e,0x18,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2e,0x16,0x00
    .byte 0x9d,0x82,0x2e,0x17,0x00
    .byte 0xb9,0x80
    .byte 0x1d
    .byte 0x9d,0x82,0x2e,0x16,0x00
    .byte 0xb9,0x81
    .byte 0x1d
    .byte 0xb9,0x82
    .byte 0xea,0x83
    .byte 0x1e
    .byte 0x5a,0x80
    .byte 0x1d
    .byte 0x5a,0x80
    .byte 0x1d
    .byte 0x5a,0x80
    .byte 0x1e
    .byte 0x5c,0xf4,0x00
    .byte 0xb0,0x83
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x6a,0x00,0x20,0x00,0x40,0x8e,0xd8
    .byte 0x25
    .byte 0x58,0x06
    .byte 0x62,0x14,0xf1,0x01,0xf2,0x01
    .byte 0x15
    .byte 0x1e
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x23
    .byte 0x6a,0xdc,0x00,0x00,0x2f,0x28,0x00
    .byte 0x61,0xee,0x04
    .byte 0xb6,0x88
    .byte 0xb7,0x88
    .byte 0x4d,0x80,0x03,0x01
    .byte 0x9d,0x88,0x2f,0x28,0x00
    .byte 0x58,0x06
    .byte 0x22
    .byte 0xb8,0x88
    .byte 0x20
    .byte 0xc3,0x88
    .byte 0x62,0x09,0x1c,0x02,0x1e,0x02
    .byte 0xa7,0x88
    .byte 0x59,0x06
    .byte 0x20
    .byte 0x6a,0x1f,0x00,0x00,0x2d,0x15,0x00
    .byte 0x2a
    .byte 0x61,0xef,0x04
    .byte 0x9d,0x80,0x2c,0x1d,0x00
    .byte 0x9d,0x81,0x2c,0x1c,0x00
    .byte 0x9d,0x82,0x2c,0x1d,0x00
    .byte 0x58,0x06
    .byte 0x20
    .byte 0xb8,0x80
    .byte 0xa7,0x80
    .byte 0xb8,0x81
    .byte 0xb8,0x82
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2c,0x1b,0x00
    .byte 0x9d,0x82,0x2c,0x1c,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2c,0x1a,0x00
    .byte 0x9d,0x82,0x2c,0x19,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2c,0x19,0x00
    .byte 0x9d,0x82,0x2c,0x1a,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2c,0x18,0x00
    .byte 0x9d,0x82,0x2c,0x19,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2c,0x17,0x00
    .byte 0x9d,0x82,0x2c,0x18,0x00
    .byte 0xa7,0x80
    .byte 0x9d,0x81,0x2c,0x16,0x00
    .byte 0x9d,0x82,0x2c,0x17,0x00
    .byte 0xb9,0x80
    .byte 0x1d
    .byte 0x9d,0x82,0x2c,0x16,0x00
    .byte 0xb9,0x81
    .byte 0x1d
    .byte 0xb9,0x82
    .byte 0xea,0x8e
    .byte 0x1e
    .byte 0x5a,0x60
    .byte 0x1d
    .byte 0x5a,0x60
    .byte 0x1d
    .byte 0x5a,0x60
    .byte 0x1e
    .byte 0x5c,0xf5,0x00
    .byte 0xb0,0x8e
    .byte 0x21
    .byte 0x9d,0x80,0x2d,0x11,0x00
    .byte 0x9d,0x81,0x2d,0x12,0x00
    .byte 0x9d,0x82,0x2d,0x11,0x00
    .byte 0xb8,0x80
    .byte 0x5a,0xe0
    .byte 0x20
    .byte 0xab,0x85
    .byte 0xab,0x83
    .byte 0xab,0x8e
    .byte 0xab,0x84
    .byte 0x20
    .byte 0xa9,0x80
    .byte 0xb8,0x81
    .byte 0xb8,0x82
    .byte 0xa9,0x80
    .byte 0x9d,0x81,0x2d,0x13,0x00
    .byte 0x9d,0x82,0x2d,0x12,0x00
    .byte 0xa9,0x80
    .byte 0x9d,0x81,0x2d,0x14,0x00
    .byte 0x9d,0x82,0x2d,0x13,0x00
    .byte 0xa9,0x80
    .byte 0x9d,0x81,0x2d,0x15,0x00
    .byte 0x9d,0x82,0x2d,0x14,0x00
    .byte 0xa9,0x80
    .byte 0x9d,0x81,0x2d,0x16,0x00
    .byte 0x9d,0x82,0x2d,0x15,0x00
    .byte 0xb9,0x80
    .byte 0x1d
    .byte 0x9d,0x82,0x2d,0x16,0x00
    .byte 0xb9,0x81
    .byte 0x1d
    .byte 0xb9,0x82
    .byte 0x61,0x78,0x07
    .byte 0x5c,0xf6,0x00
    .byte 0x1d
    .byte 0xb2,0x83
    .byte 0xae,0x8e
    .byte 0x1f
    .byte 0xb0,0x85
    .byte 0x20
    .byte 0x5c,0xf7,0x00
    .byte 0x42,0x10
    .byte 0x5c,0xf8,0x00
    .byte 0x1e
    .byte 0xab,0x83
    .byte 0xab,0x85
    .byte 0xab,0x84
    .byte 0xab,0x8e
    .byte 0x20
    .byte 0xc8,0x87
    .byte 0x62,0x1e,0x36,0x03,0x3e,0x03
    .byte 0xb9,0x87
    .byte 0x41,0x03
    .byte 0xb8,0x87
    .byte 0x41,0x03
    .byte 0x1e
    .byte 0x5c,0xf9,0x80
    .byte 0xa8,0x84
    .byte 0xa8,0x84
    .byte 0xa7,0x84
    .byte 0x21
    .byte 0xb0,0x87
    .byte 0x1e
    .byte 0x5c,0xfa,0x80
    .byte 0x64,0x54,0x03,0x57,0x03
    .byte 0xa7,0x83
    .byte 0x13
    .byte 0xaa,0x83
    .byte 0xab,0x83
    .byte 0x1e
    .byte 0x5c,0xfb,0x80
    .byte 0x5,0x80,0xe0
    .byte 0x1e
    .byte 0xc8,0x87
    .byte 0x20
    .byte 0xbb,0x87
    .byte 0xa7,0x87
    .byte 0x20
    .byte 0x48,0x93
    .byte 0x62,0x1e,0x73,0x03,0x7b,0x03
    .byte 0xb9,0x86
    .byte 0x41,0x03
    .byte 0xb8,0x86
    .byte 0x41,0x03
    .byte 0x64,0x80,0x03,0x85,0x03
    .byte 0x13
    .byte 0xa7,0x83
    .byte 0xa7,0x84
    .byte 0x64,0x8a,0x03,0x8f,0x03
    .byte 0x13
    .byte 0xa7,0x83
    .byte 0xaa,0x84
    .byte 0xaa,0x84
    .byte 0xab,0x84
    .byte 0x5c,0xfc,0x80
    .byte 0x1e
    .byte 0xb0,0x87
    .byte 0x48,0x2d
    .byte 0x9d,0x87,0x2d,0x10,0x00
    .byte 0x9d,0x86,0x2d,0x10,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x87,0x2d,0x11,0x00
    .byte 0x9d,0x86,0x2d,0x11,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x10,0x00
    .byte 0x9d,0x86,0x2d,0x10,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0f,0x00
    .byte 0x9d,0x86,0x2d,0x0f,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x87,0x2d,0x10,0x00
    .byte 0x9d,0x86,0x2d,0x10,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0f,0x00
    .byte 0x9d,0x86,0x2d,0x0f,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0e,0x00
    .byte 0x9d,0x86,0x2d,0x0e,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x87,0x2d,0x0f,0x00
    .byte 0x9d,0x86,0x2d,0x0f,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0e,0x00
    .byte 0x9d,0x86,0x2d,0x0e,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0d,0x00
    .byte 0x9d,0x86,0x2d,0x0d,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x87,0x2d,0x0e,0x00
    .byte 0x9d,0x86,0x2d,0x0e,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0d,0x00
    .byte 0x9d,0x86,0x2d,0x0d,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0c,0x00
    .byte 0x9d,0x86,0x2d,0x0c,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x87,0x2d,0x0d,0x00
    .byte 0x9d,0x86,0x2d,0x0d,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0c,0x00
    .byte 0x9d,0x86,0x2d,0x0c,0x00
    .byte 0x41,0x01
    .byte 0x9d,0x87,0x2d,0x0b,0x00
    .byte 0x9d,0x86,0x2d,0x0b,0x00
    .byte 0xb9,0x87
    .byte 0xb9,0x86
    .byte 0x1e
    .byte 0x5c,0xfd,0x00
    .byte 0x1e
    .byte 0xed,0x83
    .byte 0x1e
    .byte 0x5c,0xfe,0x00
    .byte 0x20
    .byte 0xd9,0x83
    .byte 0x20
    .byte 0xe0,0x83
    .byte 0x42,0x10
    .byte 0x9d,0x88,0x2a,0x11,0x00
    .byte 0x9d,0x89,0x29,0x13,0x00
    .byte 0x9d,0x8a,0x29,0x15,0x00
    .byte 0x9d,0x8b,0x30,0x16,0x00
    .byte 0x9d,0x8c,0x31,0x14,0x00
    .byte 0x9d,0x8d,0x30,0x11,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x88,0x2a,0x12,0x00
    .byte 0x9d,0x89,0x29,0x14,0x00
    .byte 0x9d,0x8a,0x2a,0x16,0x00
    .byte 0x9d,0x8b,0x2f,0x17,0x00
    .byte 0x9d,0x8c,0x31,0x15,0x00
    .byte 0x9d,0x8d,0x30,0x12,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x88,0x2a,0x13,0x00
    .byte 0x9d,0x89,0x29,0x15,0x00
    .byte 0x9d,0x8a,0x2b,0x17,0x00
    .byte 0x9d,0x8c,0x30,0x16,0x00
    .byte 0x9d,0x8d,0x30,0x13,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x88,0x2a,0x14,0x00
    .byte 0x9d,0x89,0x2a,0x16,0x00
    .byte 0x9d,0x8d,0x30,0x14,0x00
    .byte 0x41,0x04
    .byte 0x9d,0x88,0x2b,0x15,0x00
    .byte 0x9d,0x8d,0x2f,0x15,0x00
    .byte 0x1e
    .byte 0x48,0x02
    .byte 0x5a,0xe0
    .byte 0x1e
    .byte 0x64,0xf1,0x04,0xf6,0x04
    .byte 0x15
    .byte 0xa8,0x85
    .byte 0xa9,0x84
    .byte 0xb2,0x8e
    .byte 0x1e
    .byte 0xb0,0x83
    .byte 0x1e
    .byte 0xa9,0x83
    .byte 0xa9,0x83
    .byte 0x1d
    .byte 0xae,0x83
    .byte 0x1e
    .byte 0x5c,0xff,0x00
    .byte 0x1e
    .byte 0x64,0x0d,0x05,0x19,0x05
    .byte 0xa8,0x88
    .byte 0xa8,0x89
    .byte 0xa8,0x8a
    .byte 0xaa,0x8b
    .byte 0xaa,0x8c
    .byte 0xaa,0x8d
    .byte 0x1e
    .byte 0x64,0x1f,0x05,0x2b,0x05
    .byte 0xa8,0x88
    .byte 0xa8,0x89
    .byte 0xa7,0x8a
    .byte 0xa7,0x8b
    .byte 0xaa,0x8c
    .byte 0xaa,0x8d
    .byte 0x5a,0xe0
    .byte 0xb9,0x88
    .byte 0xb9,0x89
    .byte 0xb9,0x8a
    .byte 0xb9,0x8b
    .byte 0xb9,0x8c
    .byte 0xb9,0x8d
    .byte 0xb0,0x85
    .byte 0xb0,0x83
    .byte 0xb0,0x8e
    .byte 0xb0,0x84
    .byte 0x1e
    .byte 0x61,0x78,0x07
    .byte 0x5c,0x00,0x01
    .byte 0x43,0x29
    .byte 0x6c
    .byte 0x1e
    .byte 0xab,0x84
    .byte 0xb2,0x85
    .byte 0xae,0x8e
    .byte 0x20
    .byte 0x5c,0x01,0x01
    .byte 0x1e
    .byte 0xa9,0x8e
    .byte 0xa8,0x8e
    .byte 0xa8,0x8e
    .byte 0xab,0x8e
    .byte 0xb0,0x85
    .byte 0xb0,0x84
    .byte 0x1e
    .byte 0x5c,0x02,0x01
    .byte 0x1e
    .byte 0xdf,0x85
    .byte 0xdf,0x84
    .byte 0x1d
    .byte 0xb0,0x84
    .byte 0xb0,0x85
    .byte 0x20
    .byte 0xae,0x83
    .byte 0x41,0x04
    .byte 0xab,0x83
    .byte 0x1f
    .byte 0xa7,0x83
    .byte 0xa7,0x83
    .byte 0x1f
    .byte 0xab,0x84
    .byte 0xab,0x85
    .byte 0x1e
    .byte 0xe0,0x83
    .byte 0x22
    .byte 0xa7,0x84
    .byte 0xa7,0x84
    .byte 0x1e
    .byte 0x5c,0x03,0x01
    .byte 0x20
    .byte 0xab,0x83
    .byte 0x1e
    .byte 0xb0,0x83
    .byte 0x20
    .byte 0xdf,0x83
    .byte 0x1d
    .byte 0xb0,0x83
    .byte 0x1f
    .byte 0x63,0x02,0xa1,0x05,0xa6,0x05
    .byte 0x15
    .byte 0xa9,0x84
    .byte 0xa9,0x83
    .byte 0x64,0xab,0x05,0xaf,0x05
    .byte 0xa9,0x84
    .byte 0xa9,0x83
    .byte 0xab,0x84
    .byte 0xae,0x83
    .byte 0xb2,0x8e
    .byte 0xb2,0x85
    .byte 0x1e
    .byte 0x5c,0x04,0x01
    .byte 0x1e
    .byte 0xe0,0x84
    .byte 0xe1,0x85
    .byte 0xe1,0x8e
    .byte 0x1d
    .byte 0xab,0x84
    .byte 0xb2,0x8e
    .byte 0xb2,0x85
    .byte 0x1e
    .byte 0x64,0xcf,0x05,0xd6,0x05
    .byte 0x15
    .byte 0xa7,0x84
    .byte 0xaa,0x85
    .byte 0xa7,0x8e
    .byte 0xab,0x85
    .byte 0xb9,0x84
    .byte 0xb9,0x83
    .byte 0xaa,0x8e
    .byte 0xb9,0x8e
    .byte 0xb9,0x85
    .byte 0x67
    .byte 0x19
    .byte 0xe
    .byte 0x43,0x08
    .byte 0x5b,0x05 ; sets job flag
    .byte 0x5b,0x06
    .byte 0x5b,0x07
    .byte 0x5b,0x0a
    .byte 0x5b,0x0d
    .byte 0x5b,0x0e
    .byte 0x5b,0x18 ; unlocks gladiator early
    .byte 0x60,0x4b,0x00
    .byte 0x60,0x4c,0x00
    .byte 0x60,0x4d,0x00
    .byte 0x60,0x4e,0x00
    .byte 0x60,0x4f,0x00
    .byte 0x60,0x50,0x00
    .byte 0x2f,0x17
    .byte 0x5f,0x7e,0x00
    .byte 0x32,0xfe
    .byte 0x57,0x06
    .byte 0x77

    I was also able to deduce what some event codes do from that event:
    Code: [Select]
    0x13, 1 - move player up
    0x14, 1 - move player right
    0x15, 1 - move player down
    0x16, 1 - move player left

    0x31, 1 - set temporary event flag

    0x5c, halfword - show text (byte 1 = text ID, byte 2 = up/down?)

    0x72, 3 halfwords - yes/no prompt (text to show(?), yes event ID, no event ID)

    0xa7, 1 - move character down
    0xa8, 1 - move character right
    0xa9, 1 - move character down
    0xaa, 1 - move character left
    0xab, 1 - face upwards
    0xae, 1 - face left
    0xb0, 1 - face down
    0xb2, 1 - face right

    0xb8, 1 - spawn character?
    0x83 = lenna
    0x85 = bartz
    0x8e = galuf
    0x84 = faris
    0x81 = ?
    0x80 = ?
    0x82 = ?
    0x87 = ?
    Title: Damage overflow glitch
    Post by: samurai goroh on June 11, 2016, 12:35:01 PM
    I just heard about this in the 4JF, so I haven't tried it out. It seems that if you are able to do over 2 Bytes of damage (+65,535) then you overflow the damage check and do whatever the reminder is (so damage have to range between 65536-75534 in order to deal less that the expected 9,999).

    This can be possible if your levels are too high (Hero Song or Mix will help) and also deal a ridiculous amount of damage via Sword Dance (Sword Magic may help too). Guess weakening the enemy helps too.
    Title: Re: FF5 Bugs & Glitches
    Post by: x0_000 on June 11, 2016, 10:43:23 PM
    If you buff both your levels and your stats you can easily overflow with the chicken knife because of its multiplier calculations, I've done it during a FJF on the GBA version.
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on June 12, 2016, 03:58:04 AM
    Quote
    then you overflow the damage check and do whatever the reminder is
    Does that mean you 'skip' the sanity check of 9999 and you can actually do bigger number like 10000?

    I'm asking because in FF7 if you overflow the damage, you kill the mob:
    Hers is my 1 hit kills vs Emerald :D (https://www.youtube.com/watch?v=ittIOItA4G4)
    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on June 12, 2016, 04:41:29 AM
    Quote
    Does that mean you 'skip' the sanity check of 9999 and you can actually do bigger number like 10000?
    No. If I understand it correctly, it will overflow and start back at 0x0000 plus remainder before the 9999 check. The issue is that it might end up being less than 9999.
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on June 12, 2016, 10:38:58 AM
    If it works that way then it sux - most of times you will get less dmg :D
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on June 12, 2016, 11:02:58 AM
    As Jorgur says, you will end up dealing less than 9999, so you can do anything from 0 to 9998 if damage rolls between 65536 and 75534 (0x10000 to 0x1270E) because the MSB gets ignored as it only uses 2 Bytes.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on July 18, 2016, 12:16:44 PM
    We have some news here.

    It seems this edition fixes the Guido's cave thing:
    Code: [Select]
    CF/EBFC 2F => EF
    If I am understanding well, the tables in CF/EA00-CF/F0BF stores the properties of the tiles in the world map tilesets.
    Every tile have 3 bytes of properties. This are the results of my research so far:

    Code: [Select]
    | Forest    C.Green   Green     L.Mnt     H.mnt     Waterfal  River     Desert    Void      Sea      | Floor UW |
    | 4F CF A1  4F 2F 80  4F 2F 80  44 EF 10  40 EF 10  45 EF 20  4E FF 40  4F EF 82  40 F0 83  7C 70 83 | 91 60 00 |
    | ---------------------------------------------------------------------------------------------------|----------|
    | 0         0         0         0         0         0         0         0         0         0        | 1        | Submarine can emerge ????
    | 1         1         1         1         1         1         1         1         1         1        | 0        | Airship can fly over ????
    | 0         0         0         0         0         0         0         0         0         1        | 0        | Boat can navigate ????
    | 0         0         0         0         0         0         0         0         0         1        | 1        | Submarine can navigate ????
    | 1         1         1         0         0         0         1         1         0         1        | 0        | Hiryuu can fly over ????
    | 1         1         1         1         0         1         1         1         0         1        | 0        | Black chocobo can fly over ????
    | 1         1         1         0         0         0         1         1         0         0        | 0        | Yellow chocobo can walk ????
    | 1         1         1         0         0         1         0         1         0         0        | 1        | Main character can walk ????
    |                                                                                                    |          |
    | 1         0         0         1         1         1         1         1         1         0        | 0        | Airship cannot land
    | 1         0         0         1         1         1         1         1         1         1        | 1        | Hiryuu cannot land
    | 0         1         1         1         1         1         1         1         1         1        | 1        | Black chocobo cannot land
    | 0         0         0         0         0         0         1         0         1         1        | 0        | ????
    | 1111      1111      1111      1111      1111      1111      1111      1111      0000      0000     | 0000     | ????
    |                                                                                                    |          |
    | 1         1         1         0         0         0         0         1         1         1        | 0        | ????
    | 0         0         0         0         0         0         1         0         0         0        | 0        | ????
    | 1         0         0         0         0         1         0         0         0         0        | 0        | ????
    | 0         0         0         1         1         0         0         0         0         0        | 0        | ????
    |                                                                                                    |          |
    | 0001      0000      0000      0000      0000      0000      0000      0010      0011      0011     | 0000     | Enemy encounter
    | ---------------------------------------------------------------------------------------------------|----------|
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on July 23, 2016, 04:54:10 PM
    Sure. My question is, how are these events stored on the rom? Do we know how to read into or hack them?

    I'm a bit late to the party, but I have a little bit of data regarding event scripting (doing basically the same thing as x0_000).  I will have to look through my notes and find them.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on July 25, 2016, 06:24:30 AM
    I'm a bit late to the party, but I have a little bit of data regarding event scripting (doing basically the same thing as x0_000).  I will have to look through my notes and find them.

    Great! Sure it will helps.
    Title: Re: FF5 Bugs & Glitches
    Post by: LightPhoenix on August 06, 2016, 06:48:12 PM
    I just wanted to post and let you know I haven't forgotten about this.  I know I have some information on this stuff, but I'm struggling to find it.  The only thing I've found thus far is where event parsing is (C0/A248).  I also saw this website (http://translate.google.com/translate?hl=en&sl=ja&u=http://www9.atwiki.jp/ffbin/pages/34.html&ei=j-74TYeDCM62tgfZ9sWuCg&sa=X&oi=translate&ct=result&resnum=4&ved=0CDYQ7gEwAw&prev=/search%3Fq%3Dffbin%2Bff5%26hl%3Den%26biw%3D1384%26bih%3D749%26prmd%3Divns) (in JP) a while back.  However, I know I've looked into this more, because I know some of the stuff on that website is incorrect or incomplete.  I'm looking for that information, and I'll post on it once I find it.

    [EDIT]And in true fashion, right after I posted this I found it.  I zipped everything in the folder, but it's five years old so I'm vouching for anything.  Also enjoy some of the titles for events.
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on November 02, 2016, 12:12:30 AM
    I was playing for a Low Level Game when I stumbled with this graphical error. I don't recall triggering it before, so I believe is just an emulator issue (until I confirm it with a better one). Basically I was trying to capture all the monsters so that I could learn Time Slip, but when I caught a Magnetes and then released it, the background color messed up (It was the formation #194, just after the first save point).

    Title: Re: FF5 Bugs & Glitches
    Post by: Jorgur on November 02, 2016, 03:45:40 AM
    I have seen that happen. It may have been while using a summon spell in the steamship.
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on November 02, 2016, 09:53:57 AM
    I dont recall seeing that palette glitch on Steam version.

    When I played to get all all achievements I had an interesting battle Ghidra. I wanted to steal Killer Bow, so I decided first to drain its MP first and then Auto till steal connect. Guess my surprise when the MP reached 0 and it was still able to use Poison Breath ...
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on November 25, 2016, 06:44:16 PM
    In the Events list, the event code DB is called "Restore Player Status (Frog/Mini)". It looks like this has the effect of showing the lead character in the party on the map. What I want to know is, does it in fact also remove status ailments? Or does it just update the sprite to show stuff like Frog and Mini?

    :edit:
    Never mind, I think I figured it out. It does not appear to remove status ailments from any party members.

    LightPhoenix, is it OK if I put your Event List document on Google Drive and share it for others to edit?

    :edit:
    LightPhoenix has told me that, yes, I can put the Event List on Google Drive and share it. Here is a link (https://docs.google.com/document/d/1wX10LWyX3Y09d0LAxsY6sRGJ4kGQF9TpLufK3kLhraI/edit?usp=sharing) for anyone who wants to contribute.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on June 27, 2017, 05:15:29 AM
    Sharing new patch here.
    This one is compatible with both original and RPGe's versions of the game and should fix:
    • Ghido's Cave Hates Vehicles
    • NEW - Elder Tree Eats Chocobos
    • North Mountain (Walk On Air)
    • Pyramid of Moore (Walk Through Walls)
    • Istory Falls (Walk Through Walls)
    I'd be very pleased if somebody check this one  :lame:

    :edit:
    I have added to this message a new patch which is the sum of the Xardas' "Elder Tree fix" and mine.
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on June 28, 2017, 12:50:29 AM
    I'd be very pleased if somebody check this one  :lame:

    The four bugs are definitely fixed. It is a pretty big patch, though, which makes me wonder two things: one, how you made the patch, and two, what other maps might be affected.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on June 28, 2017, 02:39:43 AM
    It is a pretty big patch, though, which makes me wonder two things: one, how you made the patch, [...]

    If you check the ROM map:

    Code: [Select]
    CB/0000-CB/028F Tilemaps offsets (Location maps) [2bytes * 0x0148]
    CB/0290-CD/FA01 Tilemaps (Location maps) [Compressed unheaded type02] [64 * 64 bytes (after decompressing)]
    CD/FA02-CD/FA3F (Blank)

    To fix the maps I had to decompress some tilemaps, edit them, compress the edited (fixed) tilemap and injected it back into the ROM.
    As the compressed maps have a different size than the original ones, I had to reallocate A LOT of data bytes and pointers.
    On the other hand, everything was done progamatically using the "FF5e Map Editor" :wink: but you are right. It is indeed a big patch.

    [...], and two, what other maps might be affected.

    A lot of them as all the information has been reallocated. If you mean what locations have been 'directly' affected by the editions:
    • North Mountain Fix: Location id 03B. The bg0 tilemap id is 0AD: Only affects this location.
    • Pyramid of Moore Fix: Location id 178. The bg0 tilemap id is 11D and the bg1 tilemap id is 11E. Affects locations 178, 179, 17B, 17C, 17D, 17E and 17F. All of them are the Pyramid inner.
    • Istory Falls Fix: Location id 1A6. The bg0 tilemap id is 088: only affects this location.
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on June 28, 2017, 09:09:08 AM
    OK, as long as all the tilemaps were in fact able to save correctly. I had to rewrite some of my FF6 patches that affect maps in that game because I had previously used a map editor to edit the tilemaps and then some of them got broken because the program didn't compress as well as it could have, so it wasn't able to save all of the tilemaps. I had to go in and hex edit the tilemaps to prevent them from expanding to the point that they f$@#ed each other up.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on June 28, 2017, 09:39:39 AM
    I had to rewrite some of my FF6 patches that affect maps in that game because I had previously used a map editor to edit the tilemaps and then some of them got broken because the program didn't compress as well as it could have, so it wasn't able to save all of the tilemaps. I had to go in and hex edit the tilemaps to prevent them from expanding to the point that they f$@#ed each other up.

    The FF6 map editor is infamous for it risk to corrupt the ROM.

    Anyway, I wrote the compression routine here to be slow but optimal. The CD/FA02-CD/FA3F blank space is more than enough to increase the size of these tilemaps plus the "FF5e Map Editor" don't inject and reallocate tilemaps in a ROM if the total size of the sum of all tilemaps reach the limit set at CD/FA40.
    In this sense, it is as secure as it can be.

    Again, the patches which deal with compressed data are big.
    Title: Re: FF5 Bugs & Glitches
    Post by: Praetarius5018 on July 02, 2017, 05:53:16 AM
    I think that is new?
    Equip Weight doesn't update when changing items in battle, i.e. change shields.
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on November 19, 2017, 12:30:38 AM
    The obvious fix is to make 24E4 PHX/PLX like it should have.

    Re: Item Transform bug. Sorry for responding to an old topic, but I thought I should bring this up here.

    I have found an easier fix than adding PHX/PLX: just change C2/24E4 to loop on the Y register instead of the X register. C2/24E4 is only called from two places; the one LightPhoenix mentioned is immediately followed by an LDY and the other is surrounded by a PHY/PLY, so there'll be no downside. You can also change the STZ to an STA since A = 0 and there's no STZ addr,Y.
    Title: Inu's Sandbox
    Post by: Xardas on March 03, 2018, 05:28:30 PM
    Not sure if that's the right place to say this, since it's more about fixes than new bugs, but I didn't want to start a new topic because it's not about my work.

    But have you ever read about Inu (or Inu's Sandbox)? It's a person from Japan who worked on various patches and programs for the SNES FFs (ever heared about FF4/5/6 analyzer/decoder? This guy released those for FF4/5/6 Advance).
    A huge collection of ips patches for FFV is actually downloadable on romhacking.net since 2010 - yet I haven't seen this guy mentioned here.
    Although the description reads...
    Quote
    This is a collection of bugfix and improvement patches for the Japanese version of Final Fantasy V for the Super Famicom from Inu?s Sandbox.
    ... they still work for RPGe's translation (at least the ones I tried). Some of Inu's fixes have also been made by other people similarily, like Butz' Blue Mage sprite fix, or Power Drink fix. But there is also stuff I haven't seen elsewhere, like an improved Item sort system (It sorts with the following order: Items - Knives - Swords - Spears -  Axes&Hammers - Katanas - Rods - Saves&Maces - Bows - Harps - WHips - Bells - Exotic - Shields - Helmets - Armor - Accessory).
    Or another fix which deals which the teleportation bugs of Qelb (by disabling Teleport) and Ghido's Dwelling / Library of the Ancients (by Teleporting outside the Library).

    *Unfortunately, the Mineuchi/SwordSlap fix does not work - Mineuchi is unchanged while the normal "Fight" command of both characters and monsters inflict paralyze*

    Still, I think this patch collection needs more attention. Also, this might prevent people to work on patches/fixes which were already made.
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on March 08, 2018, 10:15:34 AM
    :edit:
    I have added to this message a new patch which is the sum of the Xardas' "Elder Tree fix" and mine.

    It seems we have now a patch which fix every bug related with maps.
    Are we missing something?
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 01, 2018, 12:47:35 PM
    It seems we have now a patch which fix every bug related with maps.
    Are we missing something?
    If you can list exactly what is fixed, then I could take a look and see if there's any other map-related bug or not.

    Anyway, this was just posted today in TASvideos:
    #5899: pirohiko's SNES Final Fantasy V "game end glitch" in 10:42.29 (http://tasvideos.org/forum/viewtopic.php?t=20064)
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on April 02, 2018, 10:03:46 AM
    If you can list exactly what is fixed, then I could take a look and see if there's any other map-related bug or not.

    Sure. The mentioned fix (and the related bugs) are explained here (http://slickproductions.org/slickwiki/index.php/FF5_Patch:Fix_Maps).
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 02, 2018, 11:14:34 AM
    There's also a Walk Through Walls in Sunken Walse Tower, that is displayed here:


    And there's the Skip Kelb Town glitch that might or might not be considered part of this:


    And this might be a bit of stretch, but using teleport on either Kelb (https://youtu.be/aMt_l7i1YdE) or Ancient Library (https://youtu.be/Z7dzMTbddsc) have some relation to the world map.
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on April 02, 2018, 05:34:30 PM
    I have seen the second video (or similar one).

    The Sunken Walse Tower trick is interesting ... so you can reach Gogo right from a top level?
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 02, 2018, 09:02:25 PM
    At the Sunken Walse Tower, the map is only the crystal room, so you cannot reach other rooms (other than using the usual stairs). However, there's a side effect & that is that the counter will remain. With that in mind (and if you remember how Odin's fight goes), you can win a battle at the cost of your life & what better use than to reach Exdeath!

    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on April 03, 2018, 02:40:20 AM
    WOAAH impressive!!!

    I'm not into the glitchy things, but I guess that is 'mana from heaven' to speedrunners  :laugh:

    BTW what is the mechanics of that timer? When it reach 0 you die, yet you won???
    Title: Re: FF5 Bugs & Glitches
    Post by: samurai goroh on April 03, 2018, 12:06:11 PM
    This is from memory, but here's what happens:

    When the timer runs out, the game will use the attack "Terminate" (which is what Odin and the Movers use to abort the battle). If you were to use this glitch with any other monster, then when you return from the battle you will see the animation of your character drowning (even if you're not in the water anymore) & you'll go back to the Title Screen (If you're in the world, then you will simply see the drowning animation & IIRC if you opened the menu, the counter will reach 0:00 but nothing will happen until you return to the world).

    However, in the case of Exdeath the credits will play before you actually die, as you're not supposed to get out of the battle unless you won (dying takes you to the title screen). But due to this, there's a side effect that happens when you are watching the stats & this is that since you're dead the "battle" there ends too (and is not shown on the video, but instead of seeing the 4 characters with "THE END", you will go back to the title screen as you would with a game over).

    Once you triggered this bug (escaping from the Sunken Walse Castle with the counter) there is only a way to fix this, find something that triggers a time event. Easiest is to go back to the sunken castle, but there are alternatives such as fighting Odin or going to Jacohl Cave's entrance... If you happen to save the game and cannot stop the timer, then you just soft-locked the game XD

    Note: Is not really useful for speedrunners considering the setup you need (having the submarine) and that it can only be effectively used on Exdeath (and you would have to clear your way or else you wouldn't be able to reach him on time). Perhaps it counts as a cheap victory if you're doing a 4 Job fiesta or something and are stuck...
    Title: Re: FF5 Bugs & Glitches
    Post by: Xardas on April 03, 2018, 01:11:31 PM
    And this might be a bit of stretch, but using teleport on either have some relation to the world map.
    I know not everyone here has the time or possibility to read every post or check the contents of every post. (I am referring to my own post about Inu's fixes)
    Speaking of, both glitches (Teleport in Kelb and Teleport in Library) have already been fixed by this very Inu, and I tested the fixes also on RPGe's translation. They work.
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on April 04, 2018, 02:37:27 AM
    Speaking of, both glitches (Teleport in Kelb and Teleport in Library) have already been fixed by this very Inu, and I tested the fixes also on RPGe's translation. They work.
    IF our goal is to make a patch that fixes all map-related bugs, or IF we want to make slick the best FF5 resource on the net, Xardas you might contact Inu and ask for permission for the code that fixes the bugs. If we have that we can compile whatever patch we want. The other option is to as for permission to upload his patch here. For me the best will be if he is willing to join us here :) It is never enough to have people with extended knowledge around  :cycle:

    What do you think, guys?
    Title: Re: FF5 Bugs & Glitches
    Post by: Xardas on April 04, 2018, 09:09:16 AM
    Xardas you might contact Inu and ask for permission for the code that fixes the bugs. If we have that we can compile whatever patch we want. The other option is to as for permission to upload his patch here.
    I thought of contacting him, but the last time he updated his homepage was 2010, and I haven't seen an account of him somewhere. I got his patches from RHDN, but he doesn't seem to have a forum account though. Still, his profile was updated just last year - maybe I'll give it a shot!
    Title: Re: FF5 Bugs & Glitches
    Post by: Squall on April 05, 2018, 03:07:16 AM
    I got his patches from RHDN, but he doesn't seem to have a forum account though

    OK, I thought you have found him in a Japanese site, that's why I asked you :D
    Title: Re: FF5 Bugs & Glitches
    Post by: noisecross on April 05, 2018, 08:43:43 AM
    There's also a Walk Through Walls in Sunken Walse Tower [...]

    Nice one! I can see Walse Tower map (0x137) have a bug. The exit tile (actually is an 'Event', not an 'Exit') which connect with the next room is in a wrong position.

    To fix this, two changes should be done:
    • First, moving the event in the location 0x137 to tile [0x08,0x11]. Currently, the position is [0x08,0x12]
    • Second, moving the destination of the exit in the location 0x138 to tile [0x08,0x10]. Currently, the destination is [0x08,0x11]
    I'll fix this in the next release of the patch. It can be done easily with the FF5e_Map_Editor.
    Do we know any other place where the PC can walk on air/through walls?


    About the "Skip Kelb Town" and the teleport bugs, unfortunately, those are out of the scope of this fix.
    Title: Re: FF5 Bugs & Glitches
    Post by: 13375K31C43R on April 08, 2018, 10:03:58 PM
    Do we know any other place where the PC can walk on air/through walls?

    Check out these links:
    http://finalfantasy.wikia.com/wiki/Walking_through_walls_glitch#Final_Fantasy_V (http://finalfantasy.wikia.com/wiki/Walking_through_walls_glitch#Final_Fantasy_V)
    http://finalfantasy.wikia.com/wiki/Walking_in_air_glitch#Final_Fantasy_V (http://finalfantasy.wikia.com/wiki/Walking_in_air_glitch#Final_Fantasy_V)

    I think the only occurrence of walking on air has been fixed, though.
    Title: Re: FF5 Bugs & Glitches
    Post by: assassin on April 09, 2018, 03:55:09 AM
    ^ joey scarbury does not like being called defective.
    Title: Re: FF5 Bugs & Glitches
    Post by: Lumiere77 on April 11, 2018, 12:19:55 PM
    I think this is unintended but, you make a weapon double grip only & set it to be double grip OK, the double grip ability will apply it's bonus. Unfortunately, this only works on weapons that are double grip only anyway.