Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!
You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!
I should really start trying to map out more of the game. I'd probably know so much more if my hacks weren't so surface level. For example, clearly this was a structure that I found months ago (that you just identified) that I didn't even think to look into.
p1 when s0 = 8013BA88
p2 when s0 = 8013BB44
p3 when s0 = 8013BC00
p4 when s0 = 8013BCBCNot at all. Extensive testing through every emulator known including CEN64 (the most accurate emulator) says it should be but it doesn't. DMA calls simply do not work on console which effectively make 0.10 useless.
One day I'll collect everything from the hundred of text files and my nemu bookmarks and try to start documenting things haha.
Yeah, it's weird. I should share my rom/asm file/full write-up on my code (i guess there's no hurry until you get your everdrive anyways), but i made a simple color changer for the CSS screen (Z+R to loop through colors like melee, and L to loop through shades. #makethecbuttonsgreatagain). It works fine in emulators etc, and the game run fine until you hover over a character on the CSS. Once you did that, the game would softlock (no input, but the music still plays.) [I really wanted to make a D-PAD to choose Polygon/MM/GDK with changing backgrounds and the announcer saying "FIGHTING POLYGON TEAM", but I made this code first for practice, and now I'm stuck with all this DMA trouble.]
That's really interesting. Like really interstesting
Can't say I knew exacty what I was doing but I also wrote a code that attempted to use the games DMA function (osEPiStart DMA(?)) which, like you said is super complicated. I broke down a few parameters and tried to run it and even got it working on emulator through the in game function. Still soft locked.
Wow my brain got broken trying to understand all this but I'm glad you actually tried to at least make Custom CSS screens to make selectable the non-playable characters. Fun fact: In 20XX Pichu and Pikachu are the characters used to toggle "Crazy Hand and Master Hand", (press Z button to toggle alternate character iirc) respectively.
tehz
@othermipswizards So let's make the assumption that interrupts are the problem (I have no way to test this right now). If I were to do this, would I enable PI interrupts? 
Enable_PI_Interrupts:
lui a1, $a430 ; load mi regs
ori a0, r0, $200 ; enable "set pi mask"
sw a0, 0xc(a1) ; writeYeah, the structure of the actual DMA thing (OSIoMesg) is easy, but the structure of the "OSPiHandle" is crazy, and probably unmakeable
lui a1, 0xA430 ;mi_regs
lw at, 0xC(a1) ;get mi_reg state
ori at, at, 0x0200 ;set enable PI bits
sw at, 0xC(a1) ;store maskNo to the .cod file. I pull my osFunctions from the Zelda debug rom since it's all documented.
That's five more minutes than I've looked into it haha. Looking at it more closely, since you write and read to different actual address, just writing to +0xC would probably work. I guess it doesn't hurt to try. I'm sure there's something about setting interrupts in that giant MIPS 4.0 pdf
Wait, does that means the toggles you put in 0.10 testing will work too? Will you include these toggles in 0.11?
Yes. Everything will be working. 0.11 = what 0.10 should have been + a little.
This is on the to do list.
Probably not in the near future.
The RANDOM function will likely become a stage in the future.
Probably not. Definitely not in the near future.
The biggest bug were the toggle behaving strangely on Dolphin (no longer supported) and the options bugging out/causing the screen to go red. The source of the issue has been resolved with the new DMA method. I'll scroll through the thread to see if I missed any as well.
I think it should be possible to make a "page" button on the SSS, but I really haven't looked into because of the DMA issue. So, we'll see... I start really figuring out if it possible so I might have something when jorgasms releases his console-safe DMA code.
Can't believe I never thought to do this. I could easily do this with and incorporate a function to retexture it when you "change pages."
I'm confused by your question, because you mention that the GameCube sticks have issues with range, but then begin talking about a toggle for deadzone, which is something completely different. Those GameCube sticks you refer to have a large range and if you were to implement a larger deadzone, that would make them even harder to do tilts and such.
I clearly remember that someone suggested this earlier in the thread, but way even before you mentioned the DMA issues and probably even before 0.9G itself. I think maybe, just like many other suggestions, no one did notice it at that time.
My friends do play with these sticks, sure it has their pros and cons. It would be useful for those who don't rely on much tilt usage but terrible bad for attacks that need exact accuracy in the tilting like angled tilts, and specially. pivot tilts. HOWEVER, in the real gamecube controllers it is easier to do such stuff since they don't have that much of the sensibility/deadzone analog range. It is easier to perform any tilt related technique in an original gamecube controller, rather than their modded N64 controller counterpart.
T̶h̶e̶ ̶i̶s̶s̶u̶e̶ ̶w̶i̶t̶h̶ ̶t̶h̶e̶s̶e̶ ̶s̶t̶i̶c̶k̶s̶ ̶i̶s̶ ̶t̶h̶a̶t̶ ̶t̶h̶e̶y̶ ̶h̶a̶v̶e̶ ̶a̶ ̶m̶o̶r̶e̶ ̶l̶i̶m̶i̶t̶e̶d̶ ̶a̶n̶a̶l̶o̶g̶i̶c̶ ̶r̶a̶n̶g̶e̶ ̶i̶n̶ ̶i̶t̶'̶s̶ ̶s̶e̶n̶s̶o̶r̶s̶ ̶/̶ ̶d̶e̶a̶d̶z̶o̶n̶e̶ ̶/̶ ̶w̶h̶a̶t̶e̶v̶e̶r̶ ̶i̶s̶ ̶t̶h̶a̶t̶.̶ ̶F̶o̶r̶ ̶e̶x̶a̶m̶p̶l̶e̶,̶ ̶l̶e̶t̶'̶s̶ ̶a̶s̶s̶u̶m̶e̶ ̶6̶4̶'̶s̶ ̶a̶n̶a̶l̶o̶g̶i̶c̶ ̶r̶a̶n̶g̶e̶ ̶i̶s̶ ̶f̶r̶o̶m̶ ̶0̶ ̶t̶o̶ ̶1̶0̶0̶.̶ ̶T̶h̶e̶ ̶d̶e̶f̶a̶u̶l̶t̶ ̶r̶a̶n̶g̶e̶ ̶f̶o̶r̶ ̶a̶ ̶v̶a̶n̶i̶l̶l̶a̶,̶ ̶u̶n̶m̶o̶d̶i̶f̶i̶e̶d̶ ̶c̶o̶n̶t̶r̶o̶l̶l̶e̶r̶.̶ ̶N̶e̶w̶ ̶g̶e̶n̶e̶r̶a̶t̶i̶o̶n̶ ̶s̶t̶i̶c̶k̶ ̶r̶e̶p̶l̶a̶c̶e̶m̶e̶n̶t̶ ̶(̶t̶h̶e̶ ̶g̶a̶m̶e̶c̶u̶b̶e̶ ̶o̶n̶e̶s̶)̶ ̶h̶a̶v̶e̶ ̶a̶n̶ ̶a̶n̶a̶l̶o̶g̶i̶c̶ ̶r̶a̶n̶g̶e̶ ̶f̶r̶o̶m̶ ̶0̶ ̶t̶o̶ ̶5̶0̶.̶ ̶T̶h̶e̶r̶e̶f̶o̶r̶e̶ ̶m̶a̶k̶i̶n̶g̶ ̶m̶u̶c̶h̶ ̶d̶i̶f̶f̶i̶c̶u̶l̶t̶ ̶d̶o̶i̶n̶g̶ ̶t̶i̶l̶t̶-̶r̶e̶l̶a̶t̶e̶d̶ ̶m̶o̶v̶e̶s̶.̶ ̶S̶o̶ ̶i̶f̶ ̶y̶o̶u̶ ̶c̶a̶n̶ ̶m̶a̶k̶e̶ ̶a̶ ̶t̶o̶g̶g̶l̶e̶ ̶t̶o̶ ̶f̶i̶x̶ ̶t̶h̶i̶s̶,̶ ̶I̶ ̶d̶o̶n̶'̶t̶ ̶s̶e̶e̶ ̶w̶h̶y̶ ̶i̶t̶ ̶w̶o̶u̶l̶d̶ ̶b̶e̶ ̶s̶u̶c̶h̶ ̶a̶ ̶b̶i̶g̶ ̶d̶e̶a̶l̶.̶ ̶B̶u̶t̶,̶ ̶I̶ ̶d̶o̶ ̶h̶a̶v̶e̶ ̶t̶o̶ ̶r̶e̶c̶k̶o̶n̶ ̶t̶h̶a̶t̶ ̶i̶t̶ ̶w̶o̶u̶l̶d̶ ̶g̶i̶v̶e̶ ̶m̶o̶r̶e̶ ̶a̶d̶v̶a̶n̶t̶a̶g̶e̶ ̶f̶o̶r̶ ̶t̶h̶e̶ ̶g̶a̶m̶e̶c̶u̶b̶e̶-̶i̶s̶h̶ ̶s̶t̶i̶c̶k̶ ̶u̶s̶e̶r̶s̶.̶ ̶B̶u̶t̶ ̶I̶ ̶d̶o̶n̶'̶t̶ ̶t̶h̶i̶n̶k̶ ̶t̶h̶a̶t̶ ̶w̶o̶u̶l̶d̶ ̶b̶e̶ ̶u̶n̶f̶a̶i̶r̶ ̶b̶e̶c̶a̶u̶s̶e̶ ̶w̶h̶e̶n̶ ̶S̶m̶a̶s̶h̶ ̶6̶4̶ ̶t̶o̶u̶r̶n̶a̶m̶e̶n̶t̶s̶ ̶a̶r̶e̶ ̶h̶e̶l̶d̶ ̶i̶n̶ ̶P̶C̶ ̶t̶h̶e̶ ̶K̶e̶y̶b̶o̶a̶r̶d̶ ̶a̶n̶d̶ ̶a̶n̶y̶ ̶n̶o̶n̶-̶t̶u̶r̶b̶o̶ ̶c̶o̶n̶t̶r̶o̶l̶l̶e̶r̶s̶ ̶a̶r̶e̶ ̶a̶l̶l̶o̶w̶e̶d̶
0.9G doesn't work on Dolphin/VC. However, it does work on Homebrew emulators like Not64, but I don't recommend it until someone figures out a way to completely disable frameskip to make it playable at a stable framerate. I think it would be a thing of asking the devs of the Not64/Wi64 team and ask them if it is possible to do so. Now that I have mentioned this so many times. Anyone could try it. If you play the game and figure out for yourself you will know what I'm referring to.
This isn't right.
also unfixed gc replacement sticks (possibly this years batch fixed the issue still untested) skipped steps. so instead of 1.2.3......50 they went 2-.4.6.8.50 etc. the gc replacement sticks can be fixed and modified in the hardware to have reduced sensitivity.
I knew of the inaccurate unit issues in the unfixed gc sticks. But I didn't know that last thing you said. Can you please guide me how to fix these replacement sticks? My friends and sometimes me uses those sticks to competitive play Smash 64 and we would like to know how to fix the sensitivity issues ourselves.
yeah, it should be really easy in theory. Just have to get the additional images in memory (and deal with the pointers,etc. unless you just want to load the images over the old ones). There's already a function that draws the images (i just don't know what will happen when you call the routine out of order.) Everything else should be able to be "hot swapped" whenever since it only matters when you hover over it. The only thing that would be missing is "stage previews" for FD/Race to the Finish (lol). One day I'll look into what draws the "random" preview to see if that can be co-opted or used a model for a custom routine. Or, someone could figure out how the other models are stored/loaded and fun stuff like that...
either the AVRhttp://nfggames.com/forum2/index.php?topic=5803.0 or pichttps://code.google.com/archive/p/n64-joystick-translator/.
afaik, the problem is even worse than this. Basically, the stock gc-replacement sticks hit their full-activation (+/-125-128) before the stick hits the physical edge of the restrictor gate. So, you have an outside deadzone which makes the functional area of stick even smaller than it is (and use already lose 2-4mm going from the stock stick to the gc stick). I think that makes it uncorrectable in software. You could remap 0-125 to 0-85, and maybe make some tilts easier since you're reducing the values (you'll have multiple stick positions that map to the same 0-85 value), but you'll have to make the map really carefully. And, you still have that giant deadzone on the outside of the stick that's just wasted space...
http://nfggames.com/forum2/index.php?topic=5803.0
I wouldn't map it. I'd just use a multiplier. 120 * 2/3 = 80. Like you said, some stick positions would have the same value but it's still better than nothing. The deadzone on the outside is news to me and you're right, that's unfixable.
Adding a deadzone would make it easier to tilt. If the middle ~40 was literally blank space, you could tilt at 41 (because 41 = 1) and up. However that's a bad solution since you couldn't tilt at ~30. Like I said, I was at work and just thinking because making pizzas is boring as ****. The multiplier, however, should work since there will literally be more value for you to hit tilts
Old New
000 = 000
001 = 000
002 = 001
003 = 002
004 = 002
005 = 003
006 = 004
007 = 004
008 = 005
009 = 006
010 = 006
011 = 007
012 = 008
013 = 008
014 = 009
015 = 010
016 = 010
017 = 011
018 = 012
019 = 012
020 = 013
021 = 014
022 = 014
023 = 015
024 = 016
025 = 016
026 = 017
027 = 018
028 = 018
029 = 019
030 = 020
031 = 020
032 = 021
033 = 022
034 = 022
035 = 023
036 = 024
037 = 024
038 = 025
039 = 026
040 = 026
041 = 027
042 = 028
043 = 028
044 = 029
045 = 030
046 = 030
047 = 031
048 = 032
049 = 032
050 = 033
051 = 034
052 = 034
053 = 035
054 = 036
055 = 036
056 = 037
057 = 038
058 = 038
059 = 039
060 = 040
061 = 040
062 = 041
063 = 042
064 = 042
065 = 043
066 = 044
067 = 044
068 = 045
069 = 046
070 = 046
071 = 047
072 = 048
073 = 048
074 = 049
075 = 050
076 = 050
077 = 051
078 = 052
079 = 052
080 = 053
081 = 054
082 = 054
083 = 055
084 = 056
085 = 056
086 = 057
087 = 058
088 = 058
089 = 059
090 = 060
091 = 060
092 = 061
093 = 062
094 = 062
095 = 063
096 = 064
097 = 064
098 = 065
099 = 066
100 = 066
101 = 067
102 = 068
103 = 068
104 = 069
105 = 070
106 = 070
107 = 071
108 = 072
109 = 072
110 = 073
111 = 074
112 = 074
113 = 075
114 = 076
115 = 076
116 = 077
117 = 078
118 = 078
119 = 079
120 = 080
121 = 080
122 = 081
123 = 082
124 = 082
125 = 083
126 = 084
127 = 084
128 = 085
Didn't work on console
This is what my limited knowledge and 5 minutes of googling has discovered. AFAIK, I can't just change this variable via toggle since you'd never see it. I could maybe make it so you have universal button modifier/button modifier doing boot or produce multiple ROMs. Expect a PM in the future with a test patch for these things but for now, this should solve your problem.
Curious, why is it that you decided to remove the anti-aliasing, even just temporarily?
10/10 problem. The link has been updated to include all of the other files in the folder (although they're hidden).You might need to include more stuff in the distribution for the patch to work properly for everyone.
![]()
The best answer is it didn't work on console. It cause all sorts of graphical issue
Interesting. I've used the code on original SSB64 and didn't experience any graphical issues, so I assume it has something to do with 19XX specifically? Or maybe I just didn't notice the graphical issues on the original.
No it was just my way of implementing the GS code didn't work. If 0.11 is a success, I'll be able to disable anti-aliasing with ease at the hardware level instead.
