Rather late, but I believe I figured out where overworld NPCs addresses are.
Also, it seems for the ones that move, they remain at their own location if offscreen far enough, so that should make manipulating certain Denjuu walk cycles somewhat easier.
Edit:
Language: lua
memory.usememorydomain("IWRAM")
--[[
SystemID GBA
3570 w u 0 IWRAM wtf?
3572 w u 0 IWRAM x cam
3574 w u 0 IWRAM y cam
3576 b h 0 IWRAM sprite related
357C d 3 0 IWRAM x
3580 d 3 0 IWRAM y
3584 d h 0 IWRAM z?
3588 d 3 0 IWRAM movement
358E b h 0 IWRAM nth sprite in area
358F b h 0 IWRAM state
]]--
local NPC = {
x,y,xcam,ycam,state
}
while true do
local num = 0
for i = 0x34D0, 0x3790, 0x20 do
if memory.readbyte(i) ~= 0 then
NPC.xcam = memory.read_u16_le(i+0x2)
NPC.ycam = memory.read_u16_le(i+0x4)
NPC.x = string.format('%.1f',memory.read_u32_le(i+0xC)/65536.0)
NPC.y = string.format('%.1f',memory.read_u32_le(i+0x10)/65536.0)
NPC.state = memory.readbyte(i+0x1F)
num = math.floor((i-0x34D0)/0x20)+1 --So it would be 1 offset
--gui.drawText(xcam,ycam,"X"..x.." Y"..y.."i"..num.."s"..state,null,null,10,null,null)
gui.drawText(NPC.xcam,NPC.ycam,num,null,null,10,null,null)
end
end
gui.text(0,70,num)
emu.frameadvance()
end
Edit: At the very first cave, this already influences the 2 Denjuu to the right of the cave when you go grab the first key; it's possible for them to wander far enough such that they "stop" at different locations, thus changing where they wander once they appear on screen again.
Also, moving diagonally for the "stairs" in certain areas such as the first cave is faster than moving straight through it.
memory.usememorydomain("IWRAM") --Change this to whatever Majoras Mask uses
while true do
local Address = 0x --Insert your address to toggle here
local value = 0x40CB --Insert the value you want here
local toggle = 0
if toggle == 0 then
if (joypad.getimmediate().L == true) then
toggle = 1
end
else
if (joypad.getimmediate().L == true) then
toggle = 0
end
memory.write_u32_le(Address,value) --Change 32 to whatever byte size it is
end
emu.frameadvance()
end
Or something like this. I haven't tested it, but it's something around that.
I think it's just better to stop "1-upping" ourselves, because let's face it, it's not going to keep happening. At some point in time, we're just not going to be able to keep doing it. How long do you intend to wait to find something new? A year? Two? Three? At some point we have to face the fact that we're just not going to be able to do it. dwangoAC & Co has admitted as much. So let's just break the ice and stop doing it. I honestly think we're beyond the point where we can keep doing it. Takeovers just aren't as cool as they once were. So let's get it out of the way and just keep delivering fun content, regardless of whether we "1-up" ourselves or not. Skipping a GDQ would probably make a large majority unhappy anyway.
I agree. Also it seems it becomes less and less "speedrun/superplay" and more "finding security exploits" as it gets more and more elaborate.
People mentioned the aweful game idea, which sounds nice, but not sure how much entertainment can be squeezed out of them.
Does anyone else have the problem such that whenever you open the RAM Watch, then using that to open the hex editor and messing around with hotkeys in hex editor causes hotkeys (control P for instance) to never work on Ram Watch again until the emulator is restarted? This happens on the latest version and interim.
Hey, sorry for the silence. I'm having some trouble at the moment, since it seems:
1. For this particular section, if the NPCs (or Harry) hits 0, it does not appear to be able to make it critical
2. I level up after the turtle, so I get fully healed. The mandrake hits ~25, and ~50 for normal hit and crit respectively. To make it such that the next time Harry encounters an NPC kills him, I need the following:
*Mandrake hits normal hit + crit
*Harry lands 1 normal hit + 2 crits
Point 1, along with rather terrible luck makes this section tricky. And by terrible luck I mean waiting 300+ frames for each crit.
Might have to check how long would a somewhat "ideal" non-deathwarp route would take then see how much waiting is possible before it ends up slower.
Sorry about that.
Thanks for the invite, but no thanks. I'm struggling with wiki activities, 3 different side projects, and 4 classes for the next 4 months. Sorry about that.
Alright. I'm also curious if other languages may be faster, but I have no idea how much dialogue would be encountered in the case new skips would be found.
1. If you entered Hagrid's hut, then left, it becomes day.
2. It seems the cutscene right after the Willow fight can be triggered (by cheats), allowing the story to progress without fighting said boss.
3. Unlike the first game, you're severely low level, so I might need to test if deathwarping would help, since in empty classrooms, things like the knights kill you in 1 hit.
I don't think DMC channel state at bootup is consistent. Try playing Blades of Steel, and see how the demo changes every time you play it.
I tested it for a bit using different DMC timer start states, and it looks like you are probably correct.
That sucks :(
I thought we had a good chance of console verifying this run, but if that is true it seems unlikely.
I have no idea what DMC timer means, but would it be possible to set this to some known, "legit" value to make things consistent on console?
So, did reverting the change fix the problem? feos, jlun2, nrg_zam?
Sorry for the late reply, but at the latest interim, the buttons are back for me.
How about the scroll bars?
@feos: In the latest dev build copying/cutting cells isn't working for me and when I do that it just leaves the clipboard empty and pasting doesn't do anything.
Scroll bars seems to appear again, but for GBC core, whenever I misclick and click on this region next to "A" in the interim I get :
http://imgur.com/a/guHiw
System.NullReferenceException: Object reference not set to an instance of an object.
at BizHawk.Client.EmuHawk.TAStudio.TasView_MouseDoubleClick(Object sender, MouseEventArgs e)
at System.Windows.Forms.Control.OnMouseDoubleClick(MouseEventArgs e)
at System.Windows.Forms.Control.WmMouseUp(Message& m, MouseButtons button, Int32 clicks)
at System.Windows.Forms.Control.WndProc(Message& m)
at System.Windows.Forms.Control.ControlNativeWindow.OnMessage(Message& m)
at System.Windows.Forms.Control.ControlNativeWindow.WndProc(Message& m)
at System.Windows.Forms.NativeWindow.Callback(IntPtr hWnd, Int32 msg, IntPtr wparam, IntPtr lparam)
Everything works and looks correct to me.
What OS are you working with jlun2? You and nrg_zam seem to be having the same issue.
@feos: do you think my commit that fixed the original issue i was having should be reverted? It seems like it might be cause of problems here. Maybe then i can find a more surgical fix for windows 10, since I'll at least be able to see the problem, without messing things up for other users.
http://imgur.com/p9n4kIR
Edition: Windows 10 Home
Version 1607
OS Build: 14393.576
How about you?
I'll see if optimizing the input, but pressing "A" on Harry at the same frame as the movie would allow the "28" to occur.
Also, can you please check if there's some 2nd RNG affecting this?
Edit: It doesn't work; even if the last "A" press is the same frame, changing the previous A makes it 13.
Changing 0x00F8 in WRAM ~5 frames before 0BCA (damage dealt) is set changes to 13. Although the addresses around it seems to be for sprite animation. I suspect its the frame that changes 0x0BCA that's responsible, since if the hit changes it at frame 26166 it hits 28, but it hits 13 at other frames.
Edit 2: Nevermind, landed a different hit at 26166 and it was 13.
Edit 3: Brute force check every frame naively by delaying 1 frame then pressing "A"
Language: lua
memory.usememorydomain("System Bus")
local frame_start = emu.framecount()
local state = {"Checklag","Wait","Reload","Fix","Done"}
function waitframe(frame)
while (frame >= 0) do
emu.frameadvance()
frame = frame-1
end
end
--[[
algorthim for checking rng
start before land hits
1. save state at enemy select
2. press "A" not on lag frame
3. wait 200 frames and see if crit
4. reload save
5. advance frame
6. repeat
]]--
--0BCA
local damage_dealt = 0 --needs to be outside loop so it won't get reset
while true do
local frame_now = emu.framecount()
local frame_dif = 0 --difference between this frame and start frame
local statenow = state[1]
savestate.saveslot(1)
while statenow == state[1] do
frame_now = emu.framecount()
--C3A1 at 37 == cursor gone. 36 == cursor here
savestate.saveslot(1)
console.log("Reached 1 "..emu.framecount())
if memory.readbyte(0xC3A1) == 37 then --just in case
statenow = state[2]
break
end
if memory.readbyte(0xC3A1) == 36 then
joypad.set({A = 1})
emu.frameadvance()
statenow = state[2]
break
end
emu.frameadvance()
end
while statenow == state[2] do
console.log("Reached 2")
waitframe(5)
if memory.readbyte(0xC3A1) == 36 then --the enemy select cursor still exists, something went wrong with state1
statenow = state[4]
break
else
waitframe(200)--200 if you want to crit, 400 for npc
statenow = state[3]
break
end
end
while statenow == state[3] do
console.log("Damage: "..memory.readbyte(0x0BCA,"WRAM").." prev "..damage_dealt)
if damage_dealt == 0 then
damage_dealt = memory.readbyte(0x0BCA,"WRAM")
elseif damage_dealt == memory.readbyte(0x0BCA,"WRAM") then
frame_dif = frame_dif + 1
savestate.loadslot(1)
emu.frameadvance()
statenow = state[1]
break
elseif damage_dealt ~= memory.readbyte(0x0BCA,"WRAM") then
console.log("Done. Frame pressed at: "..frame_now.." Started at frame: "..frame_start)
console.log("You waited "..frame_now-frame_start.." frames")
client.pause()
statenow = state[5]
break
else
console.log("IDFK?")
client.pause()
statenow = state[5]
break
end
end
while statenow == state[4] do --something went wrong with state2, go try reload
console.log("Reached 4")
savestate.loadslot(1)
emu.frameadvance()
joypad.set({A = 1})
emu.frameadvance()
statenow = state[2]
break
end
while statenow == state[5] do
emu.frameadvance()
end
emu.frameadvance()
end
Edit: Made the script take account into damage and damage from previous delay, in case some other goal is in mind.
In the latest interim build:
1. Open a game
2. Open a tasproj file using the "File>Movie>Recent" dialogue
3. Don't frame advance; it should be at frame 0/whatever
4. Open up TAStudios
See the end of this message for details on invoking
just-in-time (JIT) debugging instead of this dialog box.
************** Exception Text **************
System.ArgumentOutOfRangeException: Index was out of range. Must be non-negative and less than the size of the collection.
Parameter name: index
at System.ThrowHelper.ThrowArgumentOutOfRangeException(ExceptionArgument argument, ExceptionResource resource)
at BizHawk.Client.EmuHawk.InputRoll.DoBackGroundCallback(PaintEventArgs e, List`1 visibleColumns)
at BizHawk.Client.EmuHawk.InputRoll.DrawBg(PaintEventArgs e, List`1 visibleColumns)
at BizHawk.Client.EmuHawk.InputRoll.OnPaint(PaintEventArgs e)
at System.Windows.Forms.Control.PaintWithErrorHandling(PaintEventArgs e, Int16 layer)
at System.Windows.Forms.Control.WmPaint(Message& m)
at System.Windows.Forms.Control.WndProc(Message& m)
at System.Windows.Forms.Control.ControlNativeWindow.OnMessage(Message& m)
at System.Windows.Forms.Control.ControlNativeWindow.WndProc(Message& m)
at System.Windows.Forms.NativeWindow.Callback(IntPtr hWnd, Int32 msg, IntPtr wparam, IntPtr lparam)
http://i.imgur.com/7Blrfpu.png
It then gives the errors shown.
Also completely unrelated, but whenever I open TAStudios, my RAM watch window clears out. Can this please be changed so that it doesn't do that?
Edit: Why were the buttons in TAStudio regarding the branches and markers removed? Similarly with the scrollbar.
http://tasvideos.org/userfiles/info/36050262009454538
I believe I did it. This beats the boss in 5 turns, with both alive. However, attempting to optimize the input causes the last hit to land 13 rather than 28. I'm worried it may be some desync, given the RNG appears to be the same, so can you please check if this syncs for you?
Edit: Delaying the last "A" press also changes the hit from 28 to 13 for some reason.
Hey, while messing around with Harry and Ron, I seemed to have accidentally skipped an enemy's turn. This happened at the first boss fight, where Ron went first, then I tried to flee with Harry. The game was paused a bit, so I pressed "A", and suddenly it was Ron's turn again. I think the first boss is much more approachable now, just let me try and see if I could make something work first.
Edit: Nevermind; unfortunately, this does not advance the RNG, so it still won't work since I would only be able to use Ron, and he'll end up out of MP. I mean, of course unless using only Ron to cast Flipendo 17+ times while Harry skips the turn over and over ends up being faster. lol
In other words, right now, if I get the knight to 2 more hits left, it appears I'm left with only Harry. Allowing Ron to live would work, but ends up using Flipendo since Harry has to keep skipping turns. Maybe advance the RNG with Flipendo? IDK.
Edit 2:
http://tasvideos.org/userfiles/info/35962672901060452
Quick input file where I used this trick + Flipendo to advance the RNG so Ron could crit the knight. It works, but I'm not sure if it would be faster than to heal.
I just discovered how to prevent the graphical glitch to "THE END" screen, tough the game will still freeze: instead of swapping Pokémon 7 with 8, swap 6 with 8.
But as I said, this only corrects the graphical part of the problem; the game will still freeze and not save the progress.
I was wrong about this. What made it change was not the different Pokémon swap, but the different amount of time taken in order to close the menu.
I'm doing some experiments, and looks like for every frame you close the menu later, a different script is triggered after the credits. I already got some very interesting results, like getting the game softresetted or getting teleported to a glitched location. This may even bring a way to mimic the correct credits end perfectly!
Edit: this is the most interesting one I've encountered so far. It happens if you close the menu 3 frames later in FractalFusion's movie file:
http://i.imgur.com/ggPIEMD.png
Edit 2: it looks like the game executes the next commands found in the WRAM after finishing the credits.
Given that you cancelled it, I suppose this means you failed to find a way to make the post-game contents work properly?
