Request Missed L-cancel doubles landing lag

[Quillion]

Fans did it for Project M, can it be done for Melee too?

 

[flieskiller]

Failing a L-cancel doubles landing lag (1.02) [flieskiller]
0408d69c 60000000
C208D6A4 00000003
7C050000 4080000C
EC010024 48000008
EC010032 00000000

This removes the normal check from the game (the 04), and check when dividing by 2, and replace by multiplying by 2 if the player don't L-cancel.

 

[Quillion]

Fantastic.

 

[Vegerot]

Failing a L-cancel doubles landing lag (1.02) [flieskiller]
0408d69c 60000000
C208D6A4 00000003
7C050000 4080000C
EC010024 48000008
EC010032 00000000

This removes the normal check from the game (the 04), and check when dividing by 2, and replace by multiplying by 2 if the player don't L-cancel.

This doesn't work in 20xx.

In 20xx 3.02, nothing happens.

In 20xx 4.05, the game crashes. I wanted to be punished more for missing an L-Cancel, but making my game crash is too much.

 

[flieskiller]

This doesn't work in 20xx.

In 20xx 3.02, nothing happens.

In 20xx 4.05, the game crashes. I wanted to be punished more for missing an L-Cancel, but making my game crash is too much.

It's because the same region is used by the codes "turn red when not L-cancelling", "N64 L-cancel version", and maybe others.

 

[Vegerot]

It's because the same region is used by the codes "turn red when not L-cancelling", "N64 L-cancel version", and maybe others.

Is it possible to make it work in 20xx?

 

[flieskiller]

Is it possible to make it work in 20xx?

I'll check on that this weekend, but I can't promise a success out of it.

 

[Vegerot]

Fantastic.

Did this work for you?

 

[Vegerot]

I'll check on that this weekend, but I can't promise a success out of it.

My apologies. I think the fault may not be on 20xx, but I don't know why. I decided to try this out on vanilla 1.02 Melee and it didn't work there either.

 

[flieskiller]

Vegerot I just checked, and it's supposed to work flawlessly on 20XX 4.05. Try with a bunch of other codes from the Gecko codes thread to see if it's you doing something wrong instead of the code being wrong.

 

[Vegerot]

Codes confirmed to work in Vanilla 1.02 and 20xx 4.05:

1. Leaving Debug Menu Loads CSS (only tested in 20xx)
2. Captain Falcon - Raptor Boost enters "fall" action state...
3. Solo Popo Up-B Gives Increased...
4. Peach - Infinite Float...
5. Samus - Always have Extended Grapple [donny2122]

I spent the last few hours trying these in maybe 20 different orders, and they all work.


Confirmed to NOT work in Vanilla 1.02 and 20xx 4.05:

1. Failing an L-cancel... (your code)
2. L Cancel % in Results Screen [Wooggle] @Wooggle
3. Falco/Fox - Hold Z While *Aerial* Laser Emits to Fire at Half ... [Achilles] Achilles1515



I've tried playing with/without Memory cards, enabling/disabling Memory card emu, (can't seem to get any codes to work by running directly from Nintendont instead of from USBLoaderGx,) trying to run from a USB stick instead of an external hard drive, and some more things.

Please give me some other tests to help you out. Any specific codes you'd like me to try?

Is there a way to get logs when I load nintendont? I'm running the latest versions of Nintendont and USBLoaderGX



Edit: I don't know much about IOSs, but here I go:

I have things installed in IOS 202, 222, 223, 236, 249, 250, and maybe more.

202: [Something cut off] wiirk's cIOS with USB 2.0 support...
222: Hermes cIOS or STUB
223: Hermes cIOS or STUB
236: Custom IOS based on IOS 36 with patches for full home-brew support.
249: [something cut off] anikoko's cIOS or STUB v. v21010
250: [something cut off] anikoko's cIOS or STUB v. v65535

Which IOSs would you recommend I reinstall for debugging purposes?

Edit 2: Reinstalled IOS 249 and 250 using this guide. Still doesn't work

Edit 3: Additionally, the codes that don't work now used to work when I had 20xx 3.02 in File System Format using Dios Mios. Does Dios Mios handle cheats differently?

 

[flieskiller]

I just found that info: https://gbatemp.net/threads/nintendont.349258/

"Note2: Nintendont is working in Wii mode, with shifted and custom memory addresses.
Most (if not all?) existing cheat codes you will find online have been created for GameCube mode (gamecube console, Dolphin emulator, Wii in gamecube mode using MIOS/DM/DML).
Therefore the addresses are not the same anymore and a lot of cheatcodes will NOT work."

"This is especially true for the cheats related to controller/Button to be pressed as the controllers are not using original functions and memory location."

 

[Achilles1515]

I don't think C2 injection codes work in Nintendont.

 

[Vegerot]

I don't think C2 injection codes work in Nintendont.

I just found that info: https://gbatemp.net/threads/nintendont.349258/

"Note2: Nintendont is working in Wii mode, with shifted and custom memory addresses.
Most (if not all?) existing cheat codes you will find online have been created for GameCube mode (gamecube console, Dolphin emulator, Wii in gamecube mode using MIOS/DM/DML).
Therefore the addresses are not the same anymore and a lot of cheatcodes will NOT work."

"This is especially true for the cheats related to controller/Button to be pressed as the controllers are not using original functions and memory location."

Thank you both so much! I can't believe I'm the only person that's experienced this frustration. Maybe this could be put in the Gecko thread? I tried searching for my problem but couldn't find my answer. Putting it in the official thread would have helped me out a lot.

Is there any way to fix this problem? If it has to do with the controller, would messing with the "Native controller" option help at all?

Additionally, the codes that don't work now used to work when I had 20xx 3.02 in File System Format using Dios Mios. Does Dios Mios handle cheats differently?

Vegerot I just checked, and it's supposed to work flawlessly on 20XX 4.05. Try with a bunch of other codes from the Gecko codes thread to see if it's you doing something wrong instead of the code being wrong.

How did you check with 20XX? How are you running 20xx not on a Wii?

 

[flieskiller]

How did you check with 20XX? How are you running 20xx not on a Wii?

I checked directly in the 20XX dol to see how it was made, and there is no interaction with my code and Achilles' codes from his mod.

 

[Vegerot]

I don't think C2 injection codes work in Nintendont.

In what might it work? Dios Mios?

 

[Achilles1515]

In what might it work? Dios Mios?

They work in everything but Nintendont, which is what that quote that Flieskiller posted is saying. Unless you are a Nintendont developer, I doubt you can fix this.

 

[Vegerot]

They work in everything but Nintendont, which is what that quote that Flieskiller posted is saying. Unless you are a Nintendont developer, I doubt you can fix this.

I understand, thank you. While I have your attention, I think punishing a missed L-cacnel by doubling landing lag is a great idea, and would be nice to be included in 20xx.

You're great at quickly responding to peoples' questions. For one of the most popular members on Smashboards, it really shows your dedication to the community. Thank you.

 

[Vegerot]

Confirmed: Switching to Dios Mios fixes the issue.

However, something I believe is an actual issue with flieskiller 's code: trying to use this code with the "L-Cancel percent on results screen" code doesn't work. There is no extra landing lag, and the results screen always says 0%.

 

[Vegerot]

They work in everything but Nintendont, which is what that quote that Flieskiller posted is saying. Unless you are a Nintendont developer, I doubt you can fix this.

Achilles1515

I checked directly in the 20XX dol to see how it was made, and there is no interaction with my code and Achilles' codes from his mod.

I've been trying for the last hour to insert your code into the start.dol file using this tutorial, when I realized I can't because one of the tools doesn't work on OSX. If it wouldn't be too difficult, could you insert your code into this .dol? It would be greatly appreciated, as a few of my friends expressed interest in having this kind of thing.

At least I learned a lot about how the GameCube and Melee works, and this type of memory hacking is something I'd like to learn more about soon.

 

[flieskiller]

Achilles1515


I've been trying for the last hour to insert your code into the start.dol file using this tutorial, when I realized I can't because one of the tools doesn't work on OSX. If it wouldn't be too difficult, could you insert your code into this .dol? It would be greatly appreciated, as a few of my friends expressed interest in having this kind of thing.

At least I learned a lot about how the GameCube and Melee works, and this type of memory hacking is something I'd like to learn more about soon.

Do these modifications to your DOL, and it should work
0x8a27C
60000000

0x8a280
4BF768d0

0xF70
C00600E8 7C050000
4080000C EC010024
48000008 EC010032
48089720

 

[Vegerot]

Failing a L-cancel doubles landing lag (1.02) [flieskiller]
0408d69c 60000000
C208D6A4 00000003
7C050000 4080000C
EC010024 48000008
EC010032 00000000

This removes the normal check from the game (the 04), and check when dividing by 2, and replace by multiplying by 2 if the player don't L-cancel.

Do these modifications to your DOL, and it should work
0x8a27C
60000000

0x8a284
4BF768d4

0xF78
7C050000 4080000C
EC010024 48000008
EC010032 4808971c

Interesting. I guess the last three lines of the original code is where the magic happens, because the other are different.

Why are the first two lines different? Wouldn't "the normal check from the game (the 04)" not be removed? Or did Achilles1515 already remove it? Why did you add to the last line?

Edit: I guess I had some good suspicions: The code doesn't double landing lag, it halves it. There is no difference between L-cancelling and not.

Is there anything wrong here? Here are the relevant locations; tell me if anything's incorrect. Alternatively, this is my .dol. I gave everyone write access, so you can open it up in a Drive-connected hex editor if you want.

 

[Vegerot]

flieskiller Why doesn't it work?

 

[flieskiller]

flieskiller Why doesn't it work?

I accidentally did my branches wrong, doing an infinite loop. I edited the code's hexes and it should work now

 

[Vegerot]

I accidentally did my branches wrong, doing an infinite loop. I edited the code's hexes and it should work now

Thank you! But are you sure you saved the edits back to my start.dol? There's no difference! It can actually be kinda confusing to do that right. You could just upload the start.dol to your own Drive/Dropbox and send me the link

 

[flieskiller]

Vegerot here it is, in the .zip attached to this post. There is a few probability that Achilles used the part I used (that was all 0) for in-game switches, like in the debug menu, but it should be fine.

 

[Vegerot]

Vegerot here it is, in the .zip attached to this post. There is a few probability that Achilles used the part I used (that was all 0) for in-game switches, like in the debug menu, but it should be fine.

Maybe Achilles1515 can chime in, but it STILL doesn't work. Interestingly is that it still gets rid of failed L-cancelling lag . It's not like it does nothing, it changes how the game behaves when I miss an l-cancel. However, when I do l-cancel I flash white, and when I miss I flash red (like normal).

Edit: I think I may have found a reason for it not working. These are the steps to my problem.

1. Download your start.dol, extract, and rename (to make this easier to read) flies.dol
1. Use GCR to import your flies.dol to my start.dol (so now flies.dol is start.dol)
2. Use GCR to export start.dol
3. diff start.dol and flies.dol diff tells me that they differ.

Is this supposed to happen? Does GCR modify files when you import/export them, or is there a problem with GCR that won't let me import files?

 

[flieskiller]

Vegerot Ok, I've made sure that the code was right in the .dol, so it's not supposed to break at all. Something weird is the branch to the custom code, the line wasn't registering the editing at all, maybe that's why it didn't work at all. Attached in this post.

 

[Vegerot]

Vegerot Ok, I've made sure that the code was right in the .dol, so it's not supposed to break at all. Something weird is the branch to the custom code, the line wasn't registering the editing at all, maybe that's why it didn't work at all. Attached in this post.

Thank you! It worked, this is wonderful. It'll really help me get my l-cancel rate up!

Just out of curiosity, why didn't it work all those other times, and what did you need to change?

 

[flieskiller]

Thank you! It worked, this is wonderful. It'll really help me get my l-cancel rate up!

Just out of curiosity, why didn't it work all those other times, and what did you need to change?

Sometimes it's the code not being properly written in the DOL, or some branches to go the the custom code that was wrong (it must be at the exact adress, or it doesn't work or doesn't do what it wants. The last time, I changed the injection point of 1 hex behind because the one I tried at first wasn't changed at all, so I moved it and it worked fine then.

 

[Vegerot]

Sometimes it's the code not being properly written in the DOL, or some branches to go the the custom code that was wrong (it must be at the exact adress, or it doesn't work or doesn't do what it wants. The last time, I changed the injection point of 1 hex behind because the one I tried at first wasn't changed at all, so I moved it and it worked fine then.

If you used the wrong address, it would make sense if your code did nothing at all, or caused the game to entirely break. Why did it remove normal missed l-cancel lag?

 

[flieskiller]

There are 2 parts to the code: skip the condition that see if the player L-cancelled, and applying the "if the player did it or not, then multiply or divide by 2". The first part always worked because it's just applying a "nop" to the branch, but the injection branch didn't work, so only the "always do the L cancel action" was working.