Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
Fethy wrote:
Masterjun wrote:
is nothing better than 0
eg 400 and 4000
what have less lag?
I just made some tests and result with the same amount of lag for
400, 4 000, 40 000 and 400 000.
The number of digits doesn't seem to matter. Only their value does.
IIRC, the number of digits does matter in much more laggy situations, although there rarely appear any difference in usual situations. But I think you don't have to care about it because it's almost impossible to manage the number of digits. Lag description can also be found at #2494: ISM & Mister's SNES Super Mario World in 10:22.17.
For the pipe thing, that trick actually doesn't shave off lag frames, but just shortens mario's entrance motion. It's impossible to reduce lag frames during pipe/door transitions as far as I know.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
Though it's nicely done and entertaining much, I'd rather say no. I compared it with the small 100% wip and it beats the wip in some levels but loses to the wip in other levels. Besides the points andymac and Tompa mentioned above, it wins in total between dp1b and sw4b but at least 120 frames are lost there, which can never be ignored, and I believe it's faster to 6/5 and dodge the statue at the end of room 7 of the bowser castle, instead of using R scroll which brings a loss of dozens of frames.
I'm also doubtful if this type of run can have a concept good enough to get published. I think one of the most important points of the published 100% small run is that it aims to break/complete the levels which looks impossible to be broken/completed with small mario (while this run has dp1b and sw4b as levels of this type). It's just my opiniton however.
Voting no because of its large improvability we cannot overlook.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
TheFlameBoy wrote:
Before anything else: Hi everyone, this is my first post!
Well, I tried to use Mister's One-sided Wall-jump and it worked good. But why I couldn't catch the wall again? I did exactly the same as the previous wall-jumps. Here is a Video:
(video)
Also, thanks Mister!.
Well, an input sequence after the first jump depends on the amount of subpixels (7E13DA) at a frame you catch the wall: if it's 0x00 (when the wall is on your right), that sequence works fine, and it depends also on your initial x speed: it must be (-)2, so you have to modify your speed and (sub-)position to apply it with no change. If you can't modify your subposition well enough for some reason (you are in a narrow space for example), try to control the subpixel as 0x10 or 0x30 and change the sequence into -21 -23 -24 -24 -26 -26 -21, or -21 -23 -24 -26 -26 -26 -21 respectively. That may work.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
Masterjun wrote:
independently of the speed, is it easier to do a corner glitch when you're flying or when you're jumping?
There seems to be no difference other than the speed between flying and just jumping. Imo, it's just because you may easily recover the top speed and control mario's position that it's easier when you're flying.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
A variant: duplication away from blocks.
[...]
The following table shows when you should throw up an item to get block duplication; distance means “how far an item is from the block”, and subpixel “subpixel of the item”.
What means "how far an item is from the block"? how far what? pixel? subpixel?
and "subpixel of the item" his X or Y?^^
Ah, sorry for unkind description. :p The "how far" part only counts vertical pixel parts, and the "subpixel" means the vertical one.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
DJSecret wrote:
Mister, in your SMW Any% TAS with ISM, who did the Yoshi-Wing exit in YI3? A friend and I are needing some help with it. We cannot, for the life of us, clone that block into the message box.
Any tips?
Both of us did. Tips are just 1) to try changing an item's vertical position (in subpixel) by releasing and regrabbing it, and 2) to kick it up at each of mario's possible vertical positions (in pixel, because mario's subpixel does not affect the subpixel of the item). These are just enough to clone a block away from it. If you would like more exact conditions, see the Block duplication section.
TheFinalBoss726 wrote:
@ dawn- i tried your method about 1k times, and i still havent gotten it... then i tried on a different corner and still havent gotten it :(
overall now i have spent about 7k rerecords on corner clips and havent gotten one... it makes me very insanely sad... :(
not to mention pissed off...
Which corner are you trying to pass through, and to which direction are you going? You might have tried the glitch in impossible situations (wrong direction, sprite-type corners, etc.).
Well, I show you an example with concrete values. For instance, suppose that you are going to the left with 49 hopping, and that mario's horizontal position is 50 (+ any subpixels) when you're touching the block from the right (not sticking yourself into the block). Then mario's position should vary like 57.3, 54.2, 51.1, 48.0, 44.F, 41.E (no other patterns get the glitch to be pulled off). It's much easier to control the vertical position, so first you have to pay attention to the horizontal position. If you want to keep p-meter after the glitch, try to 6/5 into the corner in such a way that mario's position varies like 56.B, 53.D, 50.E, 48.0, 44.F, 41.F with speed varying like 47, 48, 47, 49, 49. As you may notice, "48.0, 44.F" is the most important part; in order to get this pattern, you must have a speed of 49 when you're at x position 48.0. Moreover, any other patterns having "48.0, 44.F" as their intermediate would have you pass through the corner. The glitch probably looks very complicated, but its logic is rather simple once you understand. Sorry for my poor explanation. Good luck!
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
dnnzao wrote:
I think Im probably dumm to not understand what is X speed, hit with the certaint ammount of pixels
Im saiyng this, because you said "would sometimes make you succeed", so how does you and a lot of players do it so easily :/
PS: Sorry about the poor english, but im from Brazil :P
That means "oscillation of items in a solid object affects whether we succeed in block duplication to the side or not" (taken from here). So if the position and the speed is appropriate, cloning upward is always successful but cloning to the side is not. You have only 50% chance of success.
TheFinalBoss726: Which glitch do you mean by corner clipping? Anyway, any of corner glitches have very little chance of success in sprinting speed (say about 0.4%), so it's hard to get those things depending only on your luck. Watch the speed and subpixel carefully, and do some calculation to know roughly what amount of (sub)pixels you should shave off. I think that's the quickest way to get corner glitches. Note that the x speed should be 49 just before touching a corner because 3-pixel sticking is needed.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
dnnzao wrote:
well, there's always jerks on every forum :/
I need help on 2 glitches:
1) block duplicate
2) wall jump
does anyone can explain, with patience, how do I do then?
1) You can get blocks to be cloned when the x speed of an item is 0 and you stick the item into the block by a certain amount of pixels. To make the item's x speed zero, mario's x speed should be 0 or 1 (but not -1), and kicking it up nearby the block would sometimes make you succeed in cloning it. More detailed description can be found at our comments of any% run.
2) Wall jumps can be pulled off by jumping at the same time mario catches the wall. To catch the wall you have to stick mario into the wall by 2 pixels (so mario needs at least a speed of 33 going toward the wall).
See also this page.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
Fethy wrote:
Are you guys talking about SDW ? I haven't played this game for ages but I think Lakitu's cloud is only available after the Big Boo's tower no ?
Yeah... We are talking about sdw and I was kinda confused. You're right. So then it's still worth beating our brains out to break the very tall tower.
andymac wrote:
http://dehacked.2y.net/microstorage.php/info/1431966319/SDW.smv
Here's a WIP of my SDW 120 TAS that I was doing before, I picked it up again, and now I'm up to where ISM was when he made his WIP. This run beats ISM by ~ 1120 frames, and beats ChefStef by over 50 seconds, through two new big shortcuts and a few minor improvements elsewhere.
I'm not too sure about a lot of the new yoshi glitches. Instead of trying to explain them to me, I can figure that out myself, could someone tell me what setups are needed to use these glitches? So far I have seen no real opportunities to use any of these glitches, but if I'm wrong, I don't want to have to redo large portions of the run, so I'll ask now, when it's still relatively early. Where I'm at, I can easily go back and get a yoshi at the mushroom house thing without too much backtracking.
Nice wip, andymac. The new shortcut of Desert Pyramid is great. ;) It seems like there remain a few improvable places (I might probably miss some others but who knows).
Desert Castle (secret): stick mario into the wall by 3 pixels to make the screen scroll upwards, so we can ignore a jump on the yellow block.
Water World 2: I CAN FLY!! \o/
Attracting a goal tape to yoshi is rarely faster unless you're already taking yoshi, and I'm not sure where the other yoshi glitches work in sdw. I'm still a stranger in sdw (as you saw above :p).
You're doing quite well, keep on going!
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
jimsfriend wrote:
Could that L+R trick in Run Shell Run be used in boos tower to skip boos tower?
Yoshi + spring gets through the ceiling in the first room. Fly to the right to get near the last room.
1) cram the spring into the top left edge of the super thick wall. L+R it over until it falls. Step on the spring as it falls. This will only work if the jiggling movement it does is enough to make it go high enough that mario can step on it.
2) cram the spring into the top left edge of the super thick wall. mario precedes it and also crammed himself into the wall, so the spring is on his left. L+R the spring over so it pushes mario until he falls.
3) you guys are crazy, i'm sure you will figure out something (and probably already have made the .smv).
What crazy ideas :p I tried those two and another one (using yoshi + spring at the second room from below) but I could pull off none of them... 1) I can't step on the springboard, 2) mario can't precede the crammed springboard, 3) the springboard vanishes before I step on it. If yoshi's id # were #3, "yoshi eats nothing" glitch could be applied to pass through the ceiling.
After I finished these tests, I noticed that this level can be skipped by lakitu's could. It's much faster and easier but maybe kinda disappointing.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
TheFinalBoss726 wrote:
though how did you figure it out? were you just testing random combinations of buttons and hoping to get good results, or was there an actual logic behind which buttons you pressed and when while you were trying to improve the old one sided wall jump (or did it just happen to you randomly?)
I just did a simple calculation on a paper, because I knew that x speed should be 33 or more to stick mario into the wall and I also knew how many pixels I should move by to get one-sided walljumps. First consider the positive speed part; the most lossless pattern should be like (-1), 4, 6, 7, 9,..., 30, 31, 33 ("lossless" means "the total sum is the smallest and the last speed is 33"). On the other hand, the negative speed part needs to have as a large sum as it can, and its pattern is expected to start like -2, -3, -5,..., and to end like -11, -6, -1, (4, 6,...). So the sequence I first thought of was:
with the sum = 81. But the older one-sided walljump said that the sum should be 32, so I put additional -23 and -26 into the sequence. Thus I got a one-sided walljump of length = 44 frames. Any other possible patterns easily turned out to be slower (by putting addditional values to modify a sequence in a similar way). That's all I did.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
We have used one-sided walljumps which take 45 frames in each jump, and I found this method experimentally not in the theoretic view point. Today I calculated the theoretically most optimal way of one-sided walljumps and I found it should take only 44 frames in each jump. It's 1 frame faster than the older one. ;)
demo: http://dehacked.2y.net/microstorage.php/info/121644255/smw-newonesidedwalljump.smv
When you use an input macro of snes9x, it's like
[(<BY)]15(BY)[(<BY)]2Y[(>BY)]24(>Y)
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
TheFinalBoss726 wrote:
Mister wrote:
bahamete wrote:
Mister wrote:
WillDaBeast wrote:
okay, wow.
I bet yoshi's id # is #1 and the thrown block's is #3. ;)
Correct. How do you know?
God told me so.
lol, maybe you can explain this:
(youtube video)
nathan found it a while ago, but i have no idea how the hell it works :P
Okay, I can explain it :p It's the same trick as the one that an empty shell respawns a naked shell, or a stumped p-switch respawns a fish, cheep cheep, using a yoshi.
A spinjumped shell has a death timer which is also used in other usages. If yoshi eats it and spits it out before the timer reaches 0, the timer will work in another way depending on the status of the shell; it respawns a sprite (naked shell in this case) when yoshi spits it out with ducking. The p-switch case is similar.
The glitch in the video is similar too, but it is combined with an item overwrite glitch. The key in yoshi's mouth is overwritten by a feather having a timer, and it respawns a sprite unexpectedly by the timer. Note that an overwritten sprite usually has a glitched graphic and glitched status (which seem to be related to the original sprite's ones).
If you would like to go further into the game, these addresses are useful (see this page for details as usual):
0x7E14C8 + (sprite id #) sprite status table
0x7E1540 + (sprite id #) timer used in various situations
I think watching them clarifies your understandings better. You may also watch sprite properties (0x7E1656-) to see what happens to the glitched property of an overwritten sprite. I haven't done (and won't do) a research much on that matter, though.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
I'll write down what we have gotten in our research on the "yoshi eats nothing" glitch, most of which are due to ISM and 33953YoShI. You may skip the technical parts of the following if you need. I would appreciate it if someone summarize my post in order to handle the glitch easier. It's too long. :(
1. Why and how does yoshi eat a null object?
Yoshi uses one common address 0x7E18AC to indicate whether he has an item in his mouth and when he swallows it. It starts at 255 and decrements every four frames, and he swallows the item when it reaches 0. The timer will be reset to 0 if he spits out or swallows the item. So if we could ignore an item in his mouth, yoshi would eat "nothing". It's indeed possible in two different ways:
I) Use two yoshis: By killing a visible yoshi who has an item in his mouth, the other yoshi will be visualized without the swallow timer getting reset, so he'll be eating "something" we'll never know.
II) Use powerup/down, room transition, respawning an egg: When yoshi puts out his tongue at the same time mario takes a damage or gets a powerup etc., yoshi will puts his tongue out twice. If he eats an item (which can stay in his mouth) at the first time, it will vanish at the second time without the swallow timer getting reset, so he'll be eating "nothing".
2. What will happen when yoshi spits out a null object?
From now on, I assume you using the lua script. First of all, make the situation clearer:
a) Yoshi can spit out "something" because the swallow timer is non-zero.
b) Because the sprite info of the item in yoshi's mouth got reset after yoshi eats a null object, the game regards the id number of the sprite yoshi is eating as #255, which is indicated by the address 0x7E160E + (yoshi's id number). This address is originally set to 255 if yoshi eats nothing in a usual sense.
As a result, yoshi can spit out an "item" of id nubmer #255. On the other hand, when yoshi spits out a normal item, some of its info get overwritten by certain values: its position is overwritten by yoshi's position plus a little somthing extra, its y speed by 0, and so on. This also happens with a null object of id number #255. The following list shows which addresses will change by yoshi's spitting out a null object:
address role role regarded as ID#255 sprite's possible values
0x7E01A9 stack y speed 0x00
0x7E01B5 stack x speed 0x10 or 0x30
0x7E01C1 stack general-purpose table 0x00
0x7E01D7 stack ypos low byte near Yoshi Ypos Low byte
0x7E01E3 stack xpos low byte near Yoshi Xpos Low byte
0x7E15C7 off-screen flag (ID#3) sprite status 0x09 or 0x0C
0x7E15D3 being eaten flag (ID#3) ypos high byte near Yoshi Ypos High byte
0x7E15DF interaction with objects (ID#3) xpos high byte near Yoshi Xpos high byte
0x7E167B sprite properties (ID#1) horizontal direction 0x00 or 0x01
0x7E16CF block bounce sprite exists flag being eaten flag 0x00
0x7E1725 extended sprite xpos low byte consecutive enemies it killed 0x00
I shall pick up some of them, which I think are important.
i) When the address 0x7E15D3 is a non-zero value, the ID#3 sprite can't touch the other sprites of smaller id numbers, and it behaves somewhat strangely. We can pull off this stuff by cancelling to eat a sprite which the tongue catches by taking damage.
ii) When the address 0x7E15DF takes a non-zero value, the ID#3 sprite loses its interaction with objects so falls through objects.
iii) The address 0x7E167B indicates the ID#3ID#1 sprite's property (taken from SMWC's RAM map):
bit 0: Don't disable clipping when killed with star
bit 1: Invincible to star/cape/fire/boucing blocks
bit 2: Process while off screen
bit 3: Don't change into a shell when stunned
bit 4: Can't be kicked like a shell
bit 5: Process interaction with mario every frame
bit 6: Gives power-up when eaten by yoshi
bit 7: Don't use default interaction with mario
Because the moving direction takes 0 or 1 as its value, bit 0 is set to 0 or 1 depending on the direction yoshi faces, and all the other bits are set to 0. The infinitely usable p-switch has something to do with bit 3.
iv) Stacks are used while mario is standing on a brown platform. If spitting out a null object on it, an unexpected program would run, which causes a freeze, or change of mario's powerup status, or nothing (these are all what we can see). The status change happens when yoshi's id number is #4 and when (yoshi's x position) + Z mod 256 = 214, where Z is 16 if yoshi is facing right, and -16 if facing left. (More precisely, when (yoshi's x position) + Z mod 256 = 214, the address 0x7E0015 + (yoshi's id number) will be added by 1. So yoshi's id number should be #4 if we want to glitch the powerup status 0x7E0019.)
3. Possible applications.
1) Use directly the results of i), ii) and iii):
i) To let yoshi take no damage when touching enemies, or to prevent an enemy from turning or getting killed by touching another enemy or a shell or something when we are guiding the enemy.
ii) To get a key or a shell etc equipped over a ceiling, or to kill an enemy by kicking a shell across a wall.
iii) To use a p-switch infinitely, to carry a p-switch or a key by kicking it like a shell, or to kill an obstacle lift by a fireball.
2) Pass through the ground: the effect of ii) will make yoshi fall through the ground, so by riding yoshi after yoshi falls a little through the ground, we can pass through the ground. The same is possible with a springboard.
3) Mario's powerup status change: mario's powerup status changes as small → big → cape → fire → red (glitched), etc., so we can get caped by any chance in a level where mario can't be caped for some reason. The powerup status of red mario (status 4) behaves strangely when he gets a powerup:
mushroom → fire
fireflower → fire
feather → cape
star → cape
1-up → freeze?
and when cutting a goal tape with having an item, it turns into a star. Another powerup status would bring another result. We've not known details yet.
4) Delayed star: When an item turns into a star at the goal, the address 0x7E13CB (whose value is usually 0) is added by 1 per star. This value is doubled each time we switch a room in a level, and when it reaches 0x80 (=128) or greater, mario will start the room with the star power.
Edit: fixed ID#1 in iii).
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
WillDaBeast wrote:
New (I think) Glitch!
Now, although useful, (Notice that the shell WAS yellow) it is a very restricted glitch.
Here's why:
It MUST be with a purple or yellow shell. If you try with red or green, Yoshi just eats it.
You have to be on Yoshi and have a fire flower.
I don't know if this is new or not, but I haven't seen it before. Feedback would be nice, and stuff.
Yeah cool find! It must be a new way to get an item swap glitch. Though I've not tested yet, I think this is somthing like: getting a coin on the tongue makes the tongue catch a null object, so by scrolling the screen a newly appearing shell gets swapped onto the tongue. This method is somewhat easier to utilize than the method using power-ups because you only need to be firemario and need an enemy that will turn into a coin by a fireball, and we will often encouter such situations. :D
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
When you hit a goal tape, and spawn a starman (which is never), this is set to $01. Now each time you switch an area in a level, this gets multiplied by 2. When this reaches $80 (changed area 7 times), you will start the area with the star power. The instructions which sets this address to $01 is located at $00:FB5C.
Look at this:
"When this reaches $80 (changed area 7 times), you will start the area with the star power."
Video^^:
Helpful vid, thanks. But just two things I noticed. You need only one item to pull off the delayed star glitch, and the 2 stars are both real stars. When his powerup status is glitched (value 4), a star makes him caped, and after he's caped a star behaves as a usual star.
There is another way to get a delayed star in the iggy demo: carry one p-switch and hit the goal tape (without entering the extra room) after his status is glitched, complete yi4 without entering the extra room, and defeat iggy. In this way mario is being caped at yi4 so the run would not lose entertainment imo.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
Very impressive run. There're many things I've never seen. :)
That star glitch would be indicated by the addresse 7E13CB as this page says. I found this after I watched your wip :p So well, according to its description, you only need 1 star (=p-switch) to get a feather, and a star for the castle? If so, YI3 can be shortened? or you can get a star in demo after iggy fight? That also would look nice, I guess.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
bahamete wrote:
Mister wrote:
What an awesome glitch! Is there going any numerical research heavily watching memory addresses? Anyway, I should be testing it out soon...
Perhaps I was too ambiguous with the term "research". I only mean we are testing this in different circumstances with different sprites. I posted it here so someone more... "experienced", for lack of a better word, could do research into this.
Okay, thanks anyway. You guys' samples are very useful. I hope we can find any laws which make the glitch be more easily handled.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Experienced Forum User, Published Author, Player
(172)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
bahamete wrote:
A new glitch, discovered by nathanisbored, is currently being researched.
The glitch basically involves eating a sprite at the same time a powerup is collected. This will permanently despawn the sprite and make Yoshi eat a "null" sprite. This can make sprites act very, very strangely, and can even freeze the game.
A few samples:
Freezing the gameChanging properties of a sprite(Youtube) Changing properties of a various sprites(Youtube) The original video
Like I said, we are still testing this glitch and are hoping to find a practical use for it.
What an awesome glitch! Is there going any numerical research heavily watching memory addresses? Anyway, I should be testing it out soon...
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
