Binary Hack Research & DevelopmentGot a well-founded knack with your binary Pokémon hacks? Love reverse-engineering them? For the traditional Pokémon ROM hacker, this is the spot for polling and gathering your ideas, and then implementing them! Share your hypothesis, get ideas from others, and collaborate to create!
I'm searching for a routine that opens the bag to select an item (like when you give some item to a pokemon from the pokemon menu) and it saves the item ID into a variable. Does anyone know something about it or how to do it?
I'm searching for a routine that opens the bag to select an item (like when you give some item to a pokemon from the pokemon menu) and it saves the item ID into a variable. Does anyone know something about it or how to do it?
Actually, I wouldn't recommend using JPAN's method: for some reason he wrote a completely custom decryptor, while there is an existing one in the rom already. It's way easier to use because changing a pokémon's species requires just a single callasm. (If anybody is interested, I made this in ancient times: )
Spoiler:
Code:
.text
.align 2
.thumb
.thumb_func
.global Indexnumber_changeand20bytes_recalc
main:
push {r0-r3, lr}
getpokeoffset:
ldr r0, varnumber_8004
ldr r3, vardecrypt
bl bxr3
ldrh r2, [r0]
mov r0, #100
mul r2, r0
ldr r0, partystart
add r3, r2, r0
push {r3}
getandsetnewindexnumber:
ldr r0, vartocontainnewindexnumber
ldr r3, vardecrypt
bl bxr3
mov r2, r0
mov r1, #0xB
ldr r0, [sp]
ldr r3, pokeencrypt
bl bxr3
recalc20bytes:
pop {r0}
ldr r3, effectivestatsrecalc
bl bxr3
pop {r0-r3, pc}
bxr3:
bx r3
.align 2
varnumber_8004: .word 0x8004
vardecrypt: .word 0x0806E454+1
partystart: .word 0x02024284
vartocontainnewindexnumber: .word 0x8005
pokeencrypt: .word 0x0804037C+1
effectivestatsrecalc: .word 0x0803e47c+1
/*Insert anywhere. then use special 0x9F to select a pokemon, waitstate, setvar 0x8005 to the desired index number, and callasm this routine+1*/
However, I believe the thing that's important about his request is playing the evolution animation, which isn't as easy to do as this.
My question regarding this how can I make it so that not every pokemon that is picked can change into the form?
Also, I noticed that when you change forms it changes the pokemon's level to be one less than it was before.
So let's say I wanted to make Deoxys "evolve" into another form of Deoxys when I talk to a meteor. How would I go about making the "evolve" Deoxys specifically whether it is in slot 1 or 3 or 5, etc?
#org @start
lock
faceplayer
setvar 0x8004 0x4 //the species number of pokemon you want it to evolve(here's chamander)
setvar 0x8005 0x2 //the species number of pokemon after evolution(here's ivysaur)
callasm 0x8ZZZZZZ+1 //check the slot number of the pokemon
compare 0x800D 0xFF //check if the pokemon exists in the player's party
if1 1 @no_that_pokemon
callasm 0x8YYYYYY+1 //evolution
waitstate
//setweather 3 //you can define other events after waitstate
//doweather
release
end
//---------------
#org @no_that_pokemon
msgbox @message_string 6 //"You don't have that pokemon. "
release
end
//---------
// Strings
//---------
#org @message_string
= You don't have that pokemon.
Note: Only the first pokemon in your party with the species number in 0x8004 will evolve. I think that's better.
Quick Preview:
Download: https://github.com/jiangzhengwenjz/EV-IV-Screen/tree/c1b303e17788f21e1a663204f9db59ffdcfc2a2d
Usage:
1. Make sure that you have Devkitarm and GNU make/cygwin installed.
2. Download this repo and create a new folder called "build".
3. Assign the free space you want to use in linker.lsc.
4. Open your cmd prompt and type in make. Then press enter.
5. Now main.bin should be generated. Paste it to the offset specified in step3.
6. Write a script to test (You should callasm to the offset of insertion + 0x79).
Code:
Example:
lock
faceplayer
callasm 0x8800079 //I inserted the code at 0x8800000
msgbox 0x8FD0013 0x2 //"It's nice, huh?"
releaseall
end
80x80 mugshots
Spoiler:
Intro:
In Ruby, many hackers uses the BG0 to make mugshots. However, the same way that simply writing tileset, tilemap and palette to the RAM won't work for FR & EM as it will be cleared by the MSGBOX.
However, there's still a way to do it, which requires the knowledge of "rboxes". I've figured out some simple usage of the related functions, so I will show you my work:
The most ****ing thing should be that only 1 BG palette for FR can be used, which is the palette for the money box (the box with thinner border). Therefore, it will only be able to launch 1 mugshot at a time and you can't use it with the money box!
How to use: The routines are provided at the bottom.
1. Change 0x8900000 in mugshot_make.asm to free space, which is ended in 0,4,8,c. It will be the location of our custom table.
2. Compile the 2 routines and insert them to some free space (Not the table location in step 1)
3. Get/draw some 80x80 mugshots, here's 2 examples:
They should be indexed to 16 colors.
4. Open your UNLZ-GBA and insert your images and palettes.
Note: the image and the palette should all be LZ77 compressed, which means that you should insert them by using "export image" and "export palette" in UNLZ-GBA. Aparrently they should all be inserted at some free space.
5. Build the table at the offset in step 1, whose format should be [image0 pointer][image0 pal pointer][image1 pointer][image1 pal pointer][image2 pointer][image2 pal pointer]...................
It can load up to 65536 images in algorithm, which means that it's enough.
6. Test time! use
setvar 0x8004 [image index in the table]
setvar 0x8005 [0 or 1] // 0 = left, 1 = right
callasm 0x8[mugshot_make.asm+1]
to launch the mugshot and
callasm 0x8[mugshot_del.asm+1] to delete it.
An example:
I insert the 2 images at 0xFA0000, 0xFA09CC and the 2 palettes at 0xFA09A4, 0xFA1484. Then the table should be:
I haven't tested much on this hack, so it may have some bugs but you can modify it on your own as I've released the source code at the bottom.
The new bytes are:
Code:
0xAA - move up backwards (very slow)
0xAB - move down backwards (very slow)
0xAC - move right backwards (very slow)
0xAD - move left backwards (very slow)
0xAE - move up backwards (slow)
0xAF - move down backwards (slow)
0xB0 - move right backwards (slow)
0xB1 - move left backwards (slow)
0xB2 - move up backwards (normal)
0xB3 - move down backwards (normal)
0xB4 - move right backwards (normal)
0xB5 - move left backwards (normal)
0xB6 - move up backwards (fast)
0xB7 - move down backwards (fast)
0xB8 - move right backwards (fast)
0xB9 - move left backwards (fast)
0xBA - jump up backwards (2 squares)
0xBB - jump down backwards (2 squares)
0xBC - jump right backwards (2 squares)
0xBD - jump left backwards (2 squares)
0xBE - jump up backwards (1 square)
0xBF - jump down backwards (1 square)
0xC0 - jump right backwards (1 square)
0xC1 - jump left backwards (1 square)
0xC2 - face down and move left-down (normal)
0xC3 - face up and move left-up (normal)
0xC4 - face left and move left-down (normal)
0xC5 - face right and move right-down (normal)
0xC6 - face down and move right-down (normal)
0xC7 - face up and move right-up (normal)
0xC8 - face left and move left-up (normal)
0xC9 - face right and move right-up (normal)
Here's a picture including all the bytes above in a script:
The diagonal ones (0xc2~0xc9) can be used in some kind of stairs, as [MENTION=26643]Invert[/MENTION] asked of me about that kind of stairs.
Some people may be interested in the source code, so I will put it here:
How to insert if you don't like the offset 0xF80000 and don't want to use the patch:
1. Repoint the table at 0x3A65BC (3 pointers totally)
2. Change the 0x8F802A8 at the first line to the insert offset of the table + 0x2a8.
3. Compile it and insert it at the insert offset of the table + 0x2a8.
4. Write a script to test.
Credits:
Back~, Invert, Knizz
Hey would you by chance have a variant of that code that skips the evolution cutscene? I'm trying to do it for form changes, and the form change ASM routines here base it off of slot number and don't really give the option to check if you have a certain pokemon or not. For example. One will let you choose any pokemon to evolve into what you want and the other bases it on what slot number the pokemon is in. I need a code that will change a specific pokemon's index number or even making it "evolve" silently.
Hey would you by chance have a variant of that code that skips the evolution cutscene? I'm trying to do it for form changes, and the form change ASM routines here base it off of slot number and don't really give the option to check if you have a certain pokemon or not. For example. One will let you choose any pokemon to evolve into what you want and the other bases it on what slot number the pokemon is in. I need a code that will change a specific pokemon's index number or even making it "evolve" silently.
For silent evo you'd better take a look at FBI's routine which is also posted under this thread.
If you still want to use my routine:
Spoiler:
You need to modify the routine because the logic will be completely different and I currently don't have the desire for polishing my old trash routines. In addition, the way to implement it is strongly dependent of the author's requirement.
It will be easy if you want to edit it on your own. Basically it should be:
(From the script's view) call special XX (the pokemon selection screen with OW callback) -> get the slot number from var -> decrypt its species with a built-in function -> find its evo form in evolution table -> ... ???
Yes, here comes the question: if a pokemon has several different evo method, which should it be taken?
There will be several apparent options:
1. ban those pokemons in this event
2. always take the first form
3. ask the player (how? should we generate a dynamic option list? )
4. randomly pick one (pointless for players abusing S/L)
5. build a custom table to control everything
......
That's why I think you'd better implement it yourself. The code I wrote should be easy to understand but I still want to explain the addresses used here:
Code:
0x300537C: RAM location to store the callback to be called post evolution
0x80568E0: the callback function for returning to OW
0x20370C0: var 0x8004
0x80CDDA8: the evolution function (*)
0x803FBE8: The built-in pokemon data decrypter
*: actually it adds a task but you don't need to know its inner logic if you don't have much experience with ASM hacking.
(I posted detailed explanation because several people asked me about that routine. )
.text
.align 2
.thumb
.thumb_func
main:
push {r0-r7, lr}
ldr r0, .SLOT @calculate address of Pokemon we want to evolve
ldrh r0, [r0]
mov r1, #0x64
mul r0, r0, r1
ldr r1, =(0x2024284)
add r0, r0, r1
mov r7, r0
mov r1, #0xB
ldr r3, =(0x803FBE8 +1) @get current species, prior to evolving
bl linker
mov r5, r0
bl getEvolvedForm @check what species it evolves into
mov r6, r0
mov r0, r7
mov r1, #0x38
ldr r3, =(0x803FBE8 +1) @save level before evolution
bl linker
mov r4, r0
SetEvolve:
mov r0, r7 @evolve it
ldr r2, .VAR
mov r1, r6
strh r1, [r2]
mov r1, #0xB
ldr r3, =(0x804037C +1)
bl linker
mov r0, r7
ldr r3, =(0x803E47C +1) @recalculate stats
bl linker
AddToDex:
mov r0, r6
ldr r3, =(0x8043298 +1) @get dex index of species
bl linker
mov r0, r6
mov r1, #0x2
ldr r3, =(0x8088E74 +1) @normal dex
bl linker
mov r0, r6
mov r1, #0x3
ldr r3, =(0x8088E74 +1) @national
bl linker
FinishCalc:
mov r0, r7 @check current level (after evolving)
mov r1, #0x38
ldr r3, =(0x803FBE8 +1)
bl linker
cmp r0, r4 @exp curves skewed level
bne AdjustLvl
b Nickname
getEvolvedForm:
ldr r0, =(0x20370BC)
ldrh r0, [r0]
bx lr
Nickname:
mov r0, r7
ldr r1, =(0x2021CD0) @store current nickname here
mov r2, r1
mov r1, #0x2
ldr r3, =(0x803FBE8 +1)
bl linker
ldr r0, =(0x8245EE0)@calculate species default name
mov r1, #0xB
mul r1, r1, r5
add r1, r1, r0
mov r5, r1
ldr r1, =(0x2021CD0)
mov r4, r1
mov r0, #0x0
checkEqual:
cmp r0, #0xA @While nickname has char
beq setName
add r5, r5, #0x1 @get next char of species name
ldrb r1, [r5]
add r4, r4, #0x1 @get next char of nickname
ldrb r2, [r4]
cmp r2, r1 @if x != y ==> has nickname
bne end
cmp r2, #0xFF @strings are 0xFF terminated
beq checkOne
b contLoop
checkOne:
cmp r1, #0xFF
beq setName
b end
contLoop:
add r0, r0, #0x1 @loop counter ++
b checkEqual
setName:
ldr r2, =(0x8245EE0) @set nickname as species name
mov r0, #0xB
mul r6, r6, r0
add r6, r6, r2
mov r0, r7
mov r2, r6
mov r1, #0x2
ldr r3, =(0x804037C +1)
bl linker
b end
AdjustLvl:
cmp r0, r4
beq Nickname
mov r2, r0
mov r0, r7
mov r1, #0x19
ldr r3, =(0x803FBE8 +1) @x = currentExp * 1.1
bl linker
mov r1, r0
lsr r0, #0x4
cmp r2, r4 @if current lvl < old lvl; currentExp -= x
bls subtract
add r1, r1, r0
b updateExp
subtract:
sub r1, r1, r0 @else currentExp += x
updateExp:
ldr r2, .VAR
str r1, [r2]
mov r1, #0x19
mov r0, r7
ldr r3, =(0x804037C +1)
bl linker
mov r0, r7
ldr r3, =(0x803E47C +1)
bl linker
mov r0, r7
mov r1, #0x38
ldr r3, =(0x803FBE8 +1)
bl linker
b AdjustLvl
end:
pop {r0-r7, pc}
linker:
bx r3
.align 2
.VAR:
.word 0x20270B8 + (0x8000 *2) @change to any 2 byte free space in RAM
.SLOT:
.word 0x20270B8 + (0x8001 *2) @slot of Pokemon to evolve
That one uses the variable 0x8002 to determine it's evolution. You can use it to evolve into a specific evolution that way. Or be silly and evolve Rattata into a Venasaur :3
Mini update:
Remember how the TMs were showing quantities? Well I fixed that now. I'll edit my original post with the details as well. Here it is:
Where XX XX XX is the pointer to where you assembled the routine +1
Now navigate to 0x131EA5 and change the byte to E0
That's all.
So I'm having an issue using both codes when using them. The first code only works on pokemon that are not part of my expanded dex, as well as it doesn't work on custom evolutions for some reason. For example, I use it on deoxys to try to evolve it to some random pokemon and it just freezes the game upon execution.
The second code works for my expanded mons, but for some reason, when I use it on expanded pokemon it changes the pokemon's name. For example I evolved clefairy into Ribombee and its name turned into SHADOW PU lol what is going on here?
Hi, I'm new here, don't know if this has been said already, but I've been working on my own hack for a while, I've been stuck at this one thing, i made the evolution stones like water stone, etc, cost like 20,000 pokedollars, but when i go to buy them, i see they cost ?000...when i select it, it shows it's 20,000, i just would like to increase the length of the prices to show 5 digits, i followed the advice of a person named Touched, he said to change 04 to 05 at some offset i cant remeber right now...i changed it to 05 but the game froze, changed it back to 04 and the game went back to normal. Please help me, anyone, I'm doing fire red
Hi, I'm new here, don't know if this has been said already, but I've been working on my own hack for a while, I've been stuck at this one thing, i made the evolution stones like water stone, etc, cost like 20,000 pokedollars, but when i go to buy them, i see they cost ?000...when i select it, it shows it's 20,000, i just would like to increase the length of the prices to show 5 digits, i followed the advice of a person named Touched, he said to change 04 to 05 at some offset i cant remeber right now...i changed it to 05 but the game froze, changed it back to 04 and the game went back to normal. Please help me, anyone, I'm doing fire red
This isn't really the place to ask these questions. This thread is about developing ASM routines to add/modify functions to the game. The Quick Research & Development Thread is what you're looking for.
Your answer seems to be there, too! Try following this post and see if it fixes your problems.
Thanks Ritcher! I thought I could ask here because I thought I needed an ASM code or something, but now I see I just needed to change 3 different value not just two. Thanks I did get the result I needed. And thanks to all you guys being so great at all this ASM coding and being so helpful with your great skill of programming!
Thanks Ritcher! I thought I could ask here because I thought I needed an ASM code or something, but now I see I just needed to change 3 different value not just two. Thanks I did get the result I needed. And thanks to all you guys being so great at all this ASM coding and being so helpful with your great skill of programming!
No problem, glad to be of help!
Anyways, I've got my own idea. I don't know how much anyone cares to implement something like this, or if it's even possible. It's a pretty crazy thing to begin with. I'll still ask nonetheless.
Would it be possible to modify the GBA games to have two dedicated save files for two different people? The games already use all the space for a second save file to store a backup save. If possible, why not make a more practical use for it?
I know that this would cause problems with functions that don't require you to load up a save first, such as New Game, Options, Mystery Gift, GameCube to Game Boy Advance linking, migrating Pokémon to the Gen IV games, and save editing. My idea would be to make it so the most recently used save is the "active" save file. This way these functions interface with the active save, the Options menu reflects the settings of the active user, and New Game will save over the active save. Or, alternatively, change the New Game saving situation to be like the later games where you have to delete a save first. Then you can change the active save by loading up the "secondary" save file, which would now push the other save into secondary status.
I've always loved the idea of having multiple save files on one Pokémon cartridge. I know that it'll never be a feature in the official games, early on because of technical reasons (the first couple of generations couldn't handle it) and later on strictly for monetary reasons (buy more copies of the games!). But it'd still be a watershed moment in my eyes for it to become a reality.
Anyways, I've got my own idea. I don't know how much anyone cares to implement something like this, or if it's even possible. It's a pretty crazy thing to begin with. I'll still ask nonetheless.
Would it be possible to modify the GBA games to have two dedicated save files for two different people? The games already use all the space for a second save file to store a backup save. If possible, why not make a more practical use for it?
I know that this would cause problems with functions that don't require you to load up a save first, such as New Game, Options, Mystery Gift, GameCube to Game Boy Advance linking, migrating Pokémon to the Gen IV games, and save editing. My idea would be to make it so the most recently used save is the "active" save file. This way these functions interface with the active save, the Options menu reflects the settings of the active user, and New Game will save over the active save. Or, alternatively, change the New Game saving situation to be like the later games where you have to delete a save first. Then you can change the active save by loading up the "secondary" save file, which would now push the other save into secondary status.
I've always loved the idea of having multiple save files on one Pokémon cartridge. I know that it'll never be a feature in the official games, early on because of technical reasons (the first couple of generations couldn't handle it) and later on strictly for monetary reasons (buy more copies of the games!). But it'd still be a watershed moment in my eyes for it to become a reality.
yes richter, I agree with you and your idea as well. My daughter is always playing my savefiles, luckily for the switch has different profiles so she plays her own pokemon go pikachu file while she doesnt mess with mine. It's a great feature!
Doing things that are completely unnecessary since conception.
Join Date: Jul 2019
Location: Petalburg City
Age: 20
Gender:
Male
Posts: 20
I'm going to go ahead and feel free to request something. So a while back i wanted to edit emerald to force only super effective moves to deal damage, and i figured the best way to do this was to edit the type chart. After doing that and playing the game, i quickly realized that editing the type chart also applies to status moves (e.g. Toxic, Growl, etc.) which is undesirable. I could have also edited the abilities for every Pokémon to have Wonder guard instead, but i feel like that's taking a lazy approach, and in that situation Slaking not having truant would make it even more over powered. So now that i have free time again i decided to come back to this project, and see if i can finally make this come to fruition. I would do this myself, but i don't understand ASM at all, and through my searching i haven't come up with any info about how i would even go about adding this rule in (I'm guessing it needs to be added to the damage calculation?).
TL;DR Wonder Guard at all times while keeping abilities and status moves in tact.
__________________
Feel free to correct my grammar and spelling.
Quote:
Originally Posted by Luigi
Quote:
Originally Posted by Mario
Nice of the princess to invite us over for a picnic, eh Luigi?
i have same issues in pokemon stadium i cant add new types due to next addresses being for moves animations , accuracy and evasion i dont know how to reallocate the chart
Can someone please help me with the Sitrus Berry Updates for Pokemon Emerald? I'm a bit new to ASM and not sure if I should run an overworld script with callasm command in game somewhere to change the effect the Sitrus Berry has or if I insert the pointer somewhere else in the Rom to get the desired effect out of the ASM in the First Post.
Diluting the process down to Hex and Free Space finding would be appreciated too!
Right now I've got an ASM folder with the thumb compiler. What all would I then need to copy/paste save and convert to HEX? With such a HEX code I'd insert it into freespace ending in 0 and then where do I go from here?
Black and White 2 introduced a feature where you can move items directly between two Pokémon in your party from within the party menu, rather than having to use the bag as middle man. This ports that feature to FR ROMs.
@ CHANGE THESE:
.equ offset, 0xB7DCB0 @ ROM address of 0x400 bytes of free space
.equ OW_MALE_DIVE, 148 @ male player's diving overworld sprite number
.equ OW_FEMALE_DIVE, 149 @ female player's diving overworld sprite number
.equ FLAG_ALLOW_DIVE, 0x827 @ flag which allows the use of Dive
@ YOU MUST CHANGE THE ABOVE VALUES
@ INSTALL:
@ 1. Change the .equ values above, labelled CHANGE THESE
@ 2. If you wish to disable badge checking, follow the DISABLE BADGE CHECKING instructions below
@ 3. Save this file
@ 4. Assemble this file using the following command:
@ thumb firered_dive.s firered_dive.gba
@ 5. Create a .IPS patch using Lunar IPS:
@ Use the included base.gba file as the base ROM
@ Use the newly generated firered_dive.gba file as the modified ROM
@ 6. Patch your FireRed v1.0 ROM with the generated .IPS patch
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ DISABLE BADGE CHECKS:
@ Find and remove the appropriately labelled lines of code below
@ Ctrl + F: REMOVE THE FOLLOWING 5 LINES TO DISABLE BADGE CHECK
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ MAKE DIVE POSSIBLE IN MAPS:
@ Set the second behaviour byte of Dive-able tiles to the value of DIVE_TILE_BYTE below (0x23 by default)
@ Set their second behaviour byte of Emerge-able tiles to the value of EMERGE_TILE_BYTE below (0x24 by default)
@ Set underwater maps' light level (the Type: field in the Header view on AdvanceMap) to Underwater (0x5)
@ Create Emerge map connections to underwater maps leading to their surface maps
@ Create Dive map connections to surface maps leading to their underwater maps
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ Changing these is optional (they are the default FireRed currently)
.equ OW_MALE_WALK, 0 @ male player's normal, walking overworld sprite number
.equ OW_MALE_CYCLE, 1 @ male player's cycling overworld sprite number
.equ OW_MALE_SURF, 2 @ male player's surfing/"sat down" overworld sprite number
.equ OW_MALE_ITEM_0, 3 @ male player's first item usage overworld sprite number
.equ OW_MALE_FISH, 4 @ male player's fishing overworld sprite number
.equ OW_MALE_ITEM_1, 5 @ male player's second item usage overworld sprite number
.equ OW_FEMALE_WALK, 7 @ female player's normal, walking overworld sprite number
.equ OW_FEMALE_CYCLE, 8 @ female player's cycling overworld sprite number
.equ OW_FEMALE_SURF, 9 @ female player's surfing/"sat down" overworld sprite number
.equ OW_FEMALE_ITEM_0, 10 @ female player's first item usage overworld sprite number
.equ OW_FEMALE_FISH, 11 @ female player's fishing overworld sprite number
.equ OW_FEMALE_ITEM_1, 12 @ female player's second item usage overworld sprite number
.equ DIVE_TILE_BYTE, 0x23 @ second behaviour byte value for diving
.equ EMERGE_TILE_BYTE, 0x24 @ second behaviour byte value for emerging
.equ FLAG_ALLOW_RUN, 0x82F @ flag which allows the use of the Running Shoes
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ DO NOT CHANGE THESE VALUES:
.equ rom, 0x08000000
@ DO NOT CHANGE THE ABOVE VALUES
@ Hooks and branches
.org 0x5BA30, 0xFF
ldr r0, =(disable_running_underwater + rom + 1)
bx r0
.pool
.org 0x5C7F4, 0xFF
.word player_overworld_list + rom
I had asked about this in Rom Hacking Help with no luck, since this is an ASM specific thread maybe this is a better place to ask. Basically, it functions but has problems with female players. Firstly, the Dive sprite uses the Dive overworld sprites but with the palette of the normal walking overworld sprites. It does this for both genders, but upon entering and exiting a menu or battle, the palette is corrected to the dive sprite palette. This in itself can be a non-issue if both sprites share palettes. For male players, this is the only issue.
The main problem is with female players upon entering a menu/battle, the dive sprite is reverted to the normal walking sprite. This includes behavior as well, no longer bobbing and is now able to run. There's an additional oddity if the dive and walk sprites are different sizes. The reverted sprite appears glitched, but upon another menu/battle, fully reverts to a normal looking walk state.
So what's causing female players to undo the dive state like that, and not males? One other thing I noticed is that the underwater fog still appears but the bubbles only appear after entering and exiting a menu. I tried implementing this in both my ongoing project and in vanilla Firered, no change in results. Then I checked diving in Pokemon Gaia, and found that the sprites function perfectly well for both genders, no palette issue, no reversion issue. But that the bubbles not appearing until after a menu thing is there. My main concern is just the female players breaking. The palette and bubbles issue isn't important, I can live with that, but I can't have players walking underwater.
__________________
The more you learn about something, the more you realize just how much you don't know. I've shelved my more ambitious ideas in favor of smaller, more feasible projects and contributions. While I still have an ongoing project, and still intend to experiment with improving Gen 3's multiplayer, it's been much less stressful doing smaller things, like entries for Anthroyd's MAGM contests.
The routines are very similar so I ended up just combining the two routines into one. Basically given a slot number of a Pokemon in 0x8004, the routines will return the EVs or IVs, respectively, into the vars 0x8005-0x800A with 0x8003 as the IV/EV switch.
How to insert:
Compile and insert the following routine into free space:
Usage:
setvar 0x8003 0x[anything except 0 = IV, 0 = EV]
setvar 0x8004 0x[slot number 0 to 5]
callasm 0x[this routine +1]
The variables will be:
0x8005: HP IV/EV
0x8006: Atk IV/EV
0x8007: Def IV/EV
0x8008: Spd IV/EV
0x8009: S.atk IV/EV
0x800A: S.def IV/EV
So for about the last 10 hours I've been trying to figure something out with this code.
It works completely fine, other than the fact that when I run it, it sets all 6 variables to the IV of the pokemon I specified in the scripts HP IV
Spoiler:
Code:
.text
.align 2
.thumb
.thumb_func
main:
push {r0-r5, lr} @loads registers r0-r5
mov r4, #0x0 @copies 0(aka 0x0) to register r4
loop:
cmp r4, #0x6 @finds the difference between the value in register r4 and the number 6(aka 0x6)
bhi end @I believe this is roughly equivalent to if(r4==6) goto end
ldr r0, =(0x20370C0) @var 0x8004 = slot number
ldrh r0, [r0] @loads register r0 at the offset stored in r0 (set right before this to 0x20370C0) ONLY A HALFWORD IS LOADED
mov r1, #0x64 @copies 100(aka 0x64) to register r1
mul r1, r1, r0 @multiplies r0(the value at 0x20370C0/var 0x8004) by r1(100 aka 0x64) and stores the result in r1
ldr r0, =(0x2024284) @loads register r0 at offset 0x2024284, upon inspection of this memory location, it is where the pokemon in your party are stored
add r0, r0, r1 @adds the value in r1 to the offset r0 is loaded at, getting the party member location
ldr r3, =(0x20370BE) @loads register r3 at 0x20370BE, the location of var 0x8003
ldrh r3, [r3] @reads the halfword at 0x20370BE and writes it to r3, aka sets r3 equal to var 8003
cmp r3, #0x0 @checks if r3(at this point effectively var 0x8003) is 0
beq EVs @goes to evs if var 0x8003 is 0
add r1, r1, #0x27 @IV r1 at this point is equal to the offset at whatever pokemon party number you want, adding 0x27 sets the offset to the ivs section of the pokemon's data
b continue @goes to continue
EVs:
add r1, r1, #0x1A
continue:
ldr r2, =(0x803FBE8 +1) @I think this sets register r2 equal to the offset 0x803FBE9
bl linker @branches to linker and stores the current offset in the link register(r14)
mov r1, r0 @copies r0 to r1
@get var
ldr r2, =(0x20370C2) @using vars 0x8005-0x800A
lsl r0, r4, #0x1 @I think shifts r0 to the left(subtracts 1 from r0)
add r2, r2, r0
strh r1, [r2] @store IV in var
add r4, r4, #0x1 @adds 1 to the value of r4, allowing the assembly to terminate once it runs 6 times
b loop @goes to loop
linker:
bx r2 @branches to the offset stored in r2
end:
pop {r0-r5, pc}
.align 2
I've taken the code and put in comments on each line summarizing what I understand each line to do. Am I mistaken? Am I completely wrong?
Spoiler:
Code:
//---------------
#org 0x2514D0
lock
faceplayer
msgbox 0x82515D0 MSG_YESNO //"Would you like to know the\npotent..."
compare LASTRESULT 0x0
if 0x1 goto 0x8251570
setvar 0x8003 0x1
setvar 0x8004 0x0
callasm 0x8251C51
buffernumber 0x0 0x8005
buffernumber 0x1 0x8006
msgbox 0x8251610 MSG_NORMAL //"HP: [buffer1] IVs\nAttack: [buffer..."
buffernumber 0x0 0x8007
buffernumber 0x1 0x8008
msgbox 0x8251630 MSG_NORMAL //"Defence: [buffer1] IVs\nSpeed: [bu..."
buffernumber 0x0 0x8009
buffernumber 0x1 0x800A
msgbox 0x8251650 MSG_NORMAL //"Special Attack: [buffer1] IVs\nSpe..."
end
//---------------
#org 0x251570
//---------
// Strings
//---------
#org 0x2515D0
= Would you like to know the\npotential of your Pokémon?
#org 0x251610
= HP: [buffer1] IVs\nAttack: [buffer2] IVs
#org 0x251630
= Defence: [buffer1] IVs\nSpeed: [buffer2] IVs
#org 0x251650
= Special Attack: [buffer1] IVs\nSpecial Defence: [buffer2] IVs.
and just to make sure it wasn't a statistical fluke, I ran the same script with 5 or so more randomly encountered pokemon and ran them through an IV calculator. Each time, the HP IV shown by the ASM was correct, but then copied it to every other variable set in the script.
The routines are very similar so I ended up just combining the two routines into one. Basically given a slot number of a Pokemon in 0x8004, the routines will return the EVs or IVs, respectively, into the vars 0x8005-0x800A with 0x8003 as the IV/EV switch.
How to insert:
Compile and insert the following routine into free space:
Usage:
setvar 0x8003 0x[anything except 0 = IV, 0 = EV]
setvar 0x8004 0x[slot number 0 to 5]
callasm 0x[this routine +1]
The variables will be:
0x8005: HP IV/EV
0x8006: Atk IV/EV
0x8007: Def IV/EV
0x8008: Spd IV/EV
0x8009: S.atk IV/EV
0x800A: S.def IV/EV
So for about the last 10 hours I've been trying to figure something out with this code.
It works completely fine, other than the fact that when I run it, it sets all 6 variables to the IV of the pokemon I specified in the scripts HP IV
Spoiler:
Code:
.text
.align 2
.thumb
.thumb_func
main:
push {r0-r5, lr} @loads registers r0-r5
mov r4, #0x0 @copies 0(aka 0x0) to register r4
loop:
cmp r4, #0x6 @finds the difference between the value in register r4 and the number 6(aka 0x6)
bhi end @I believe this is roughly equivalent to if(r4==6) goto end
ldr r0, =(0x20370C0) @var 0x8004 = slot number
ldrh r0, [r0] @loads register r0 at the offset stored in r0 (set right before this to 0x20370C0) ONLY A HALFWORD IS LOADED
mov r1, #0x64 @copies 100(aka 0x64) to register r1
mul r1, r1, r0 @multiplies r0(the value at 0x20370C0/var 0x8004) by r1(100 aka 0x64) and stores the result in r1
ldr r0, =(0x2024284) @loads register r0 at offset 0x2024284, upon inspection of this memory location, it is where the pokemon in your party are stored
add r0, r0, r1 @adds the value in r1 to the offset r0 is loaded at, getting the party member location
ldr r3, =(0x20370BE) @loads register r3 at 0x20370BE, the location of var 0x8003
ldrh r3, [r3] @reads the halfword at 0x20370BE and writes it to r3, aka sets r3 equal to var 8003
cmp r3, #0x0 @checks if r3(at this point effectively var 0x8003) is 0
beq EVs @goes to evs if var 0x8003 is 0
add r1, r1, #0x27 @IV r1 at this point is equal to the offset at whatever pokemon party number you want, adding 0x27 sets the offset to the ivs section of the pokemon's data
b continue @goes to continue
EVs:
add r1, r1, #0x1A
continue:
ldr r2, =(0x803FBE8 +1) @I think this sets register r2 equal to the offset 0x803FBE9
bl linker @branches to linker and stores the current offset in the link register(r14)
mov r1, r0 @copies r0 to r1
@get var
ldr r2, =(0x20370C2) @using vars 0x8005-0x800A
lsl r0, r4, #0x1 @I think shifts r0 to the left(subtracts 1 from r0)
add r2, r2, r0
strh r1, [r2] @store IV in var
add r4, r4, #0x1 @adds 1 to the value of r4, allowing the assembly to terminate once it runs 6 times
b loop @goes to loop
linker:
bx r2 @branches to the offset stored in r2
end:
pop {r0-r5, pc}
.align 2
and just to make sure it wasn't a statistical fluke, I ran the same script with 5 or so more randomly encountered pokemon and ran them through an IV calculator. Each time, the HP IV shown by the ASM was correct, but then copied it to every other variable set in the script.
Hi, first of all I'd like to apologize for the mistake. I haven't tested that code, and it's in assembly so a semantic error here and there tend to be around for some of them. A quick skim through it, it does seem like there is a minor error.
Quote:
add r1, r1, #0x27 @IV
This line of code I assume gets the HP IV, however, it will always be 0x27, and therefore not get any subsequent IVs.
That line should be replaced by:
Quote:
add r1, r4, #0x27
If the assembler says the above line is invalid, try:
Hi, first of all I'd like to apologize for the mistake. I haven't tested that code, and it's in assembly so a semantic error here and there tend to be around for some of them. A quick skim through it, it does seem like there is a minor error.
Code:
add r1, r1, #0x27 @IV
This line of code I assume gets the HP IV, however, it will always be 0x27, and therefore not get any subsequent IVs.
That line should be replaced by:
Code:
add r1, r4, #0x27
If the assembler says the above line is invalid, try:
Code:
add r1, r1, r4
add r1, r1, #0x27
Hopefully that does the trick.
Hey FBI, thanks for the reply, and I'm happy to say that
Code:
add r1, r1, r4
add r1, r1, #0x27
did the trick perfectly! It seems to be displaying the correct IVs for each stat now and isn't causing any issues. Again, thank you so much for the help.