Posts for aldelaro5


Post subject: New docs about abilities and how the sequence works
aldelaro5
He/Him
Experienced Forum User
Joined: 1/8/2014
Posts: 29
Location: Canada, Quebec
So, I tested yesterday and today 2 things, how the sequence is made and how the game treats input as I got into something very weird. First, I knew that it has been "said" that you can't skip the pillar, however, the only info I had was so fuzzy that I had to check. That is to show the more the documents, the better. Anyway, using the usefull peach warp in her first segment, I could check what happens to the pillar, they are basically all triggered and the doors are there, but I tested every pillar and they all would not allow you to activate them early EXCEPT the second one. That one, if you activate it before ending the ch1, it will play the cs of the pure hear (that you didn't get....computer logic :) ) and play the cs when the orange door appears. However, it spawns you back to where you were before (near when you would flip int he background) and not play the last cs right next to the pillar. The music also get screwed by simply not being there. In the end, the pillar actually never completed its activation because if you come back, you don't see the heart in the pillar and you can't activate it. The door actually never truly appeared so you would have softlocked. idk why this funky outcome happens,but it really seem that it just didn't completed the activation. As for other stuff in the sequence....it's more solid than TTYD. Not only you cannot skip any ch access, but unlike TTYD, there's only one way to revert the sequence backward asnd it's pretty much useless. See, when you seqyuence break, you go forward, but by doing this, just like TTYD, it would make other triggers like the Pixls chest already activated. What this means is if you would skip a Pixl, you would not be able to get it back. In TTYD, there was some triggers that doesn't get activated by this and they prety much required you to trigger them by yourself, but doing so after a sequence break would revert the whole sequence to that trigger thus, allowing you to get another chance for the the triggers after. This was good in TTYD, but spm only has one: the cat door of 3-4. For whatever reasons, almost nothing is activated in that ch except the locked door are unlocked.....not the cat one. The key chest are still closed so, you could pick them, open that cat door and it would revert the whole sequence to the Francis fight. In theory, this could be very usefull, but this means that to make use of this, you would skip 3-4 which is almost impossible because you need to end it to progress. With the only possibility dead, it's tighter than TTYD because we can only sequence break forward and we cannot undo any sequence break so it will be hard to do it without skipping too important stuff. However, I still maintain that the game has a lot of flaws elsewhere that can change a lot. Now, the input....I wanted to test the life shroom gltich, but I got very inconsistent with the jump even with frame advance. I actually got that several actions like the Pixl and jumping requires you to hold down the input during a certain amount of frames. This doesn't apply to any characters actions which are simply a one frame input EXCEPT any jumps which requires at least 3 frames of holding to be processed after. I think I recall that Malleoz did came to that conclusion, but I couldn't remember exactly. Anyway, the amount of frames are explained in detail in this google doc: https://docs.google.com/document/d/19HBFf3SuxsAjPzXSZFlPUSfwB_lpRsEWfsTiPQAya4c/edit?usp=sharing Amd I have to say, this is the weirdest and most confusing input management I've ever seen. At least I know how to count frames for testing the input queue :) So, that's all I had. The next part would be to test known glitch and try to draw some conclusions out of them.
Even if it's a sequel, lots of people have to give their all to make a game, but some people think the sequel process happens naturally." - Masahiro Sakurai
Post subject: little update on my testing (I'm back!)
aldelaro5
He/Him
Experienced Forum User
Joined: 1/8/2014
Posts: 29
Location: Canada, Quebec
So, just to be short, my laptop died so I had to suspend my testing for a bit, but I bought hardware and build a new desktop computer. The cpu (i5 4690) is just a beast for emulation and multitasking meaning I can finally start to be more productive :) As I write this message, nothing has been done since I made the google docs, I was supposed to check the pillars, but that's when it died.... Anyway, what is important is I'm back on testing.
CoolKirby wrote:
Hey aldelaro, nice to see another post from you. I went ahead and compiled your information, and the other useful information in this thread into a Game Resources page for this game. Feel free to edit it if you see anything that needs fixing.
I chnaged the 3d gauge offset because it was wrong. It should have been 0x004D043E. Btw, I might stream from time to time because now, I don;t have to fear of lag....I can definetely run both super paper mario and TTYD at 60 fps with lle :) Hope I will find stuff.
Even if it's a sequel, lots of people have to give their all to make a game, but some people think the sequel process happens naturally." - Masahiro Sakurai
Post subject: I made some research and found very usefull stuff
aldelaro5
He/Him
Experienced Forum User
Joined: 1/8/2014
Posts: 29
Location: Canada, Quebec
Hi! So, I decided 2 weeks ago to start glitch hunting SPM as to my surprise....not much was done and a lot remains. It feels like 2 years ago when I started on TTYD so....I have a feel of deja vu :) Anyway, I'm currently doing some general research and documentation. I made a pack of 134 saves (.bin wii saves) that you can get on the RTA wiki in the saves section. You could use these if you somehow want to test the game at some points, but I need these for my testing. Next, I found a lot of offsets for usefull values such as the coordinates or general stats like HP. I have both lists for ntsc-u and ntsc-j, but I need to make some vetter formatted documents. Eventually, the data would be on Malleoz site. The values you should be very interested are the coordinates, the speed and the RNG. For now, I'm just going to give them, but all offsets would be on the docs, it's jsut going to take me a little while and pasting every offsets here....is going to be a bit overkill. for ntsc-u: Coordinate X -> 4CEEB4 Coordinate Y -> 4CEEB8 Coordinate Z -> 4CEEBC RNG -> 56CB1C I won't put the speed now, because I'm really unsure of its reliability, but I guess I will find the eight one for both version soon and put them in the documentation. For ntsc-j: Coordinate X -> 4A27B4 Coordinate Y -> 4A27B8 Coordinate Z -> 4A27BC RNG -> 54243C Now, what I can tell about the speed is that it pretty much works the same way as TTYD. There's one value that is the speed for both the X and Z axis and this value would worth 3 when you hold a direction. The Y speed is a different value on its own. Now, this is when it becomes interesting: I found a lot about RNG. First, interstingly, the ntsc-u RNG has a different function than ntsc-j. Though, they both have the same template as TTYD, it's just the numbers that changes a bit. Here's the ntsc-u formula for rerolling the RNG: RNG(x+1) = ( ( 0xA230FAA1 * RNG(x) ) + 0x05A3BAE8)%0x100000000 and here's the ntsc-j one: RNG(x+1) = ( ( 0xC1685309 * RNG(x) ) + 0xE580BD92)%0x100000000 It's weird that they are different, but they both seem to behave the seem way. Which leads me to explain when it rerolls. Basically, the RNG is a 4 bytes (big endian of course) and it will get advanced with the operation I wrote everytime somethings needs to have random outcome generated. By default, it jsut doesn't change. However, this seems to be an exception when you enter new areas....I don't know why, but using doors causes some rerolls. What this implies is if you have any animations going on or some random voices, the rng will advance. However, if you just walk and jump as Mario, it would stay frozen. Interstingly, each time you see UI text on the screen, it would constantly advance each frames. My theory is that it would need to load the shake effect of the text cause random generation. Some map like the Underwhere which has random background animation will cause rerolls. This is true for other places like 3-2 and flipside. Finally,, some general animtion will reroll it such has: Seeing a damage star (so, taking OR receiving damage) flipping coins disappearing when you get them every time you heal, you will see some stars popping out, these cause rerolls. Switching Pxl causes 2 rerolls with their animation. Breaking a brick block Boomer explosion Some characters animation such has Bowser's fire Doing a stylish Some other random events like having a Floro Sapiens going out of a pipe in 5-4 (their colors are random) They seem to depend ont he animation. For example, Bowser fires is around 3-4 frames before rerolls (inconsistent) while having a long last shake is almost constant rerolls. And when I say seeing UI text, I mean when you have dialogues, the file selection screen, the auick and pause menu and any screen when you see text written in the game font. This means that if you stand still , this will cause the HUD that tells you how to pause or check the controls at the bottom. This will make the RNG reroll per frames which might serve if you want to reroll it. As for enemies drops, killing one will reroll it so I guess it's not predetermined and can be manipulated. Also, I made a list of all the items ID you could put in the arrays to hack items....always usefull for testing triggers. I just started, a lot needs to be done, but from what I got, I think this game has way more potential than TTYD. it just doesn't feel as solid as it is. Hope you look forward to more info :) EDIT: The documentations are done: The offsets: https://docs.google.com/spreadsheets/d/18JRw8_XHfbAUHCI4Ze4wlD0a0aZVazhd_o0q7ybumKg/edit?usp=sharing The ID: https://docs.google.com/spreadsheets/d/1r4N9HSJjyR4AFehS34rsG4fX_YzSThagv7oBmwa5k6c/edit?usp=sharing And the details about RNG: https://docs.google.com/document/d/11qd1ClGf3Fw3GyaOFFk6wBWZXy5hKeYEmznjoUTKe9c/edit?usp=sharing You should have most needed here :)
Even if it's a sequel, lots of people have to give their all to make a game, but some people think the sequel process happens naturally." - Masahiro Sakurai
aldelaro5
He/Him
Experienced Forum User
Joined: 1/8/2014
Posts: 29
Location: Canada, Quebec
darkszero wrote:
https://code.google.com/p/dolphin-emu/source/detail?r=cdfe58f7ede09ff11b13adf5a911da5a50927870 Is the issues finally fixed? :) EDIT: seems like a fix to the fix: https://code.google.com/p/dolphin-emu/source/detail?r=5f14f0e3ce407d32760491a12830f0dff8284b5e
yes they are http://i.imgur.com/FxVxyZX.png Water emulation is fixed http://i.imgur.com/ZSGFw7C.png Poster peeling effect is fixed http://i.imgur.com/14k2qua.png Staircase effect is fixed http://i.imgur.com/HllMwSy.png Grubba's office view from the hole in air duct is fixed you can see the whole room now Note: it doesn't fix the punies Tested on Dolphin 4.0-709 64bit linux but the update is from dolphin 4.0-707
Even if it's a sequel, lots of people have to give their all to make a game, but some people think the sequel process happens naturally." - Masahiro Sakurai