[RELEASE] Golden Sun: The Balance Age

[Salanewt]

Easier way: Just checking for that now. I know that many other animations have a sort of "flag" that is set when a graphic is loaded, which decides whether the animation (or at least a part of it) will use the palette for said graphic or instead default to elemental palettes. What is interesting about the buffs so far though is that they seem to have "djinni identifiers" which decide the initial particle palette effect and holographic djinni (there are at least 8 effects that I can find?). Which I am not at all used to and, while interesting, also wonder what this means for finding an easier fix. Plus for the fact that both palette and holographic are selected at the same time; this means the RAM solution may be necessary for this code unless I am overlooking something.

I have to study now, but:

#02030714 - Ability ID here...
#02030718 - Element is stored here for the entirety of an animation.
#02030720 - Animation ID (argument is also stored here)


Since I don't really feel like hunting for an "easier way" right now, I figure that these should help you with your planned method. :P

[Caledor]

Got it, thanks again!

EDIT:



The formula:


@Squirtle: Sorry to bother you again but there's something that makes absolutely no sense for me. When i call the function in the pic above via branch-link from the impact function, everything works perfectly. If I do the same thing from the dull function, at 0815054C, the game ignores it. I'm positive the game doesn't even enter inside the function at 0814FA76 because i tried placing breakpoints inside it and the game carried on as normal, plus r0 holds the same value when checked at 0815054A and at 0815054E (basically right before and right after the call). You can patch a TLA rom with the ips attached and see for yourself.

[Salanewt]

Nice work!

Dull: Hm. I'm still studying (taking a short pre-dinner break); could you please tell me where you tried to set up the function call (maybe with a screenshot)? If it's not much trouble for you that is.

If I had to guess based entirely on what you are saying, it sounds like you put the function call somewhere that is only executed under certain conditions.

Edit: Oh wait, you said where it is in your message. Oops!

Edit 2: Okay. The code immediately before your function call checks for two debuffs; Dull (x03) and purple Dull (x00). Meaning that it should only call the function for those two animations. I haven't actually tested the code yet, but is that what you were going for?

[Caledor]

Well... it replaces the very same instruction both in the impact function and the dull function: the LDR r0. so...

anyway, this is from the original ROM: http://imgur.com/X02dMae
this is from the patched one: http://imgur.com/3DskO7f

BTW, there's no hurry, you can check it when you're free obviously :D
And thanks!

[Caledor]

Exactly. It's the same thing i did with impact: replacing the LDR that sets the palette with the function that sets it depending on the element of the ability.
Also FYI, Purple Dull is never used cause there's no preliminary function (the ones you access from the animation pointer) that sets the argument to x00. But it's not like i care... once this thing works for either x03 or x00 the other becomes completely redundant.

[Salanewt]

I know, lol. I did the overhaul patch a while back and made it more accessible. :P

Going to very briefly test something, but it will be quick; will do more testing later if this doesn't work out.

Edit: Oh, um. So I got something weird when I changed one of the checks to Impair instead. I figure I should actually test with Dull though; could you please tell me which class(es) can add it to my spell list?

[Caledor]

just use the attached sgm file in vba.
If you can't, Sheba's base class (but with at least 4 jupiter djinn), Piers in the jupiter dual class and Ivan, Sheba and Mia's White Mage have Dull and Blunt.

[Salanewt]

Didn't work, but thanks anyway; managed to do some light testing. Same as when I changed the checks, this happens:



I haven't really looked through the code to figure out why this happens yet, but this doesn't look too great. A very tiny part of me thinks that there may be an issue with the full function itself, but I doubt it.

Will check later on.

Edit: Just for clarification, this also tends to happen when an invalid palette is present. So there could be an issue with the palette somewhere. Pretty sure you also said that r0 (the palette) doesn't actually change for some reason, so that could be it as well...

[Caledor]

Thanks a lot. it's the same thing that happens to me. r0 remains set at x1a40 instead of going through the function and changing.

Also, v1.30 for TBS and v1.30b for TLA are UP. Looks like this impact thing is gonna take a while so i went and released those two as they were yesterday after i gave Raise the mars animation.

@Varden: I was checking again the thing you asked me this morning. Actually it's 7 Djinn. In the monoelementals though it appears as if it were 8 cause they lack the 7-djinn tier. So... basically the second to last monoelemental tiers are the only ones that lost the upgraded EPA. Because in the vanilla game, the tri-elementals with 3 tiers like Paladin had Ragnarok/Cutting Edge etc at the first stage, which could be reached with 6 djinn.

@Squirtle:
I made screenshots of a step by step excecution of the BL in both Impact and Dull
Impact 1: http://i.imgur.com/No1jrwf.png
Impact 2: http://i.imgur.com/0DeYFrG.png
Impact 3: http://i.imgur.com/Q3pHXnZ.png

Dull 1: http://i.imgur.com/gcExUsX.png
Dull 2: http://i.imgur.com/uPgNJrQ.png
Dull 3: http://i.imgur.com/lYpNKNg.png

As you can see, 1 and 2 are identical for both and everything goes smoothly. BL is also preliminarily executed somehow since the LR changes. But while impact 3 actually shows that control goes to the formula i made, dull 3 shows the screw up of the LR and how the execution carries on with the following instruction as if nothing happened.

BTW, the same, identical issue happened when i tried using the BL in GS1 to fix impact there as well.

Also, I just remembered that this whole issue with branch link goes a looooong way back... Since the implementation of the PP multiplier. This is me at page THREE of this thread:

You guys did it! Thanks to your efforts the multiplier is practically DONE!
The only thing that somehow bugs me is that the division function didn't do anything but I don't really care cause I just replaced it with a 10-bit lsr. Now on to the balance of summons and djinns!

I think the same thing happened back then. my bl to 08002054, which is the division function, didn't do a thing (i clearly remember r0 not changing) so i went with the simpler and flawless 10-bit lsr. I'm actually starting to wonder why it worked for impact.

[Salanewt]

Okay! I just noticed (I never actually looked at the hex before, please don't judge me :P); you have the BL instruction set up wrong. The second part of the instruction has to be xF800 or higher. Should be a super easy fix if you change it to xFA93.

[Caledor]

OMG It's true! Thanks a lot man!

Also... Why?

[Salanewt]

Can't remember why, lol. Probably the way the assembly format was designed; could be so it can set the actual branching address? Not sure.

Happy to help!

[Caledor]

Just wanted to say that every animation that displays Attack, Defense or Resistance changes (both increase and reduction) now is elemental-dependant in TLA.
This fixes, among others: Impact, Demon Spear, Dull, Acid, Sour and a few unleashes like Storm Brand's Hurricane.

Iron is the most notable exception. Probably i'll never figure out why it turns white halfway.

Now... I'll try to replicate it into GS1 as well.

EDIT: GS1 done too... it was a lot easier since I had to fix only impact.
However, i don't feel like releasing only this... So i'll wait until there's more content.

[Salanewt]

I haven't investigated Iron yet, but I get the feeling that it and normal Steam have their own conditional branches that allow them to blend to other palettes. In fact, it probably works a lot like the 101 unleashes.

Also, nice work!

[Caledor]

Thanks. ;)

Also, I thought the same thing about iron... IMO it's just a matter of skipping a certain portion of the code. I even thought of going through the whole thing and trying to figure it out via trial-error... but I wanted to take a little break from hacking after fixing impact. I feel like i've never stopped since i started this thread...

Yesterday I noticed the "more stat" button on the forum home page and looking at the numbers there made me think "maybe I overdid it..." LOL

[Chronotakular]

I'm about to head to Lemuria!

Here are some things I've noticed that are pretty low priority, since they really don't have a huge effect on the game as a whole, like the game is still very playable.

• The event with Feizhi at Champa didn't occur this time, even though I used the same password that I did last time (though, that was a while ago..), and it happened last time. It's possible this is some weird one time freak glitch, I'll have to type in the password for my data again and see if I can get the event to trigger. Though I never really used the Golden Ring so I don't care too much. The Golden Boots event occurred fine, and I know the Colosso event will occur because of the NPC that warns you about it earlier in the game, so I'm not sure what happened here. Edit: Updated. No change. Someone else should probably confirm this though. It's possible that password I generated a year ago (I have screenshots, not the battery file) may have some last minute running around for events, and I never updated the screenshots.
• Sometimes, Felix tends to leave a battle as he entered the previous room. ie. I enter a room from falling. After a battle in the same room, Felix will fall in place. This also very rarely happens when entering and exiting the next room, but not as frequently.
• Not a bug per-say, but I know you've changed text in the past(ie. Susa talking about the Murakimo.)(it's possible that the game does this by itself, but I'm assuming otherwise.)There is a typo in an event in Champa after you defeat the Salamander Boss, Chucha says "they paid back everthing", or something along those lines. Might be more work than it's worth to fix, however.

Anyways, I'll keep on trucking; slowly but surely.

[Caledor]

There's not much I can say about the first 2. The first only needs to be tested (but i need to find the save file for it), about the second... I really have no idea. It could be even like that from the original game... either way, I'm clueless.

About the third: you guessed right, I had to manually change Cloud Brand to Murakumo. Also, fixing those typos is a total joke since the text editor has a search function. I'm fixing it as I write this post.

BTW, please keep reporting any dialogue-related mistakes... not just typos... even a sentence that just doesn't make sense to you for whatever reason is worth reporting (I remember the dialogue between Garet and his little brother - when the party returns back to Vale to find Dora's sick in GS1 - made no sense whatsoever in the Italian ROM).
I won't fix those by myself cause i'll never play the english ROM, but if you spot something and tell me, i will ;)

[Rolina]

?  I thought the cloud brand was the Kusanagi?  Is it really Murakumo in japan?  If so, that does leave it open in GS4 for it to have a power up of sorts...

[Caledor]

Actually, Ama no Murakumo and Kusanagi no Tsurugi are one and the same. The japanese name for the Cloud Brand is "amanokumo", so i went with Murakumo. Thinking about it, "kumo" usually means "cloud" so i can kinda see where the english name comes from. I still think the translator(s) failed pretty hard on this one BTW.

[Aile~♥]

• Sometimes, Felix tends to leave a battle as he entered the previous room. ie. I enter a room from falling. After a battle in the same room, Felix will fall in place. This also very rarely happens when entering and exiting the next room, but not as frequently.

This glitch existed in the original game, I believe. It's most noticeable in Gaia Rock.[/list]