Editor, Player (175)
Joined: 4/7/2015
Posts: 331
Location: Porto Alegre, RS, Brazil
Since the old Glitchfest is dead, IgorSantos777 started a fresh new one, and it's really nice so far: Link to video
Games are basically math with a visual representation of this math, that's why I make the scripts, to re-see games as math. My things: YouTube, GitHub, Pastebin, Twitter
Masterjun
He/Him
Site Developer, Skilled player (1987)
Joined: 10/12/2010
Posts: 1185
Location: Germany
I already said this in a chat, but I'll repeat it here just for reference: Of course I agree this glitchfest is entertaining so far (seriously, watch it, it has some neat ideas). However, especially this run will get repetitive really fast. Almost every level starts with the usual combination of Yoshi, Feather, and Flower. This is obviously the setup that allows for the most glitches (or rather, the most glitchy ones), which probably allows for the most effective glitchfest. But here's the point, the author intends to continue to do all 96 exits. Considering the pace is pretty slow (spending 3 minutes only in the first level), we're looking at a 2.5-3 hour movie. It's very hard (maybe impossible) to keep up the entertainment like this while avoiding repetition. (This wouldn't really be a problem if he decides to quickly proceed to Bowser and end it at around the 1 hour mark) However, there is another way to approach this problem. Instead of focusing on a glitchfest, focus on a playaround. By changing up the challenge in each level, you could create a long movie which stays entertaining, and also shows off interesting things no other run can show off. My personal ideas are:
  • Try an entertaining playaround in the level while staying Small Mario.
  • Get all the items and coins.
  • Try beating the level without jumping.
There are definitely other things you can do instead of using the stun glitch in every level and create glitchy sprites, and it would be a shame if those wouldn't be shown off.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
TheKDX7
He/Him
Player (118)
Joined: 7/9/2011
Posts: 392
Location: Switzerland
Masterjun wrote:
I already said this in a chat, but I'll repeat it here just for reference: Of course I agree this glitchfest is entertaining so far (seriously, watch it, it has some neat ideas). However, especially this run will get repetitive really fast. Almost every level starts with the usual combination of Yoshi, Feather, and Flower. This is obviously the setup that allows for the most glitches (or rather, the most glitchy ones), which probably allows for the most effective glitchfest. But here's the point, the author intends to continue to do all 96 exits. Considering the pace is pretty slow (spending 3 minutes only in the first level), we're looking at a 2.5-3 hour movie. It's very hard (maybe impossible) to keep up the entertainment like this while avoiding repetition. (This wouldn't really be a problem if he decides to quickly proceed to Bowser and end it at around the 1 hour mark) However, there is another way to approach this problem. Instead of focusing on a glitchfest, focus on a playaround. By changing up the challenge in each level, you could create a long movie which stays entertaining, and also shows off interesting things no other run can show off. My personal ideas are:
  • Try an entertaining playaround in the level while staying Small Mario.
  • Get all the items and coins.
  • Try beating the level without jumping.
There are definitely other things you can do instead of using the stun glitch in every level and create glitchy sprites, and it would be a shame if those wouldn't be shown off.
Many glitchfests on the site are repetitive... I think it's very interesting to show the limit that can present this game. The performance levels are still done pretty fast, 20 minutes for 2 worlds. I do not feel that even when doing all the outings, it will last 3 hours ... It's hard to argue, already because I do not speak English, but also because, like many people (I guess), I finally want to see a glitchfest of this game. Just like SMB3 had his TAS. Even using the same process of glitches at the beginning of the levels, IgorSantos is very good at showing something amazing and varied every time. Even having already watched many glitches on this game, many people might still be surprised at what awaits them. And after all, if it is possible to accept any% TAS of more than 5 hours or even TAS containing more than 5 or 6 categories as is the case with SMB. Why not accept a SMW "glitchfest" ? It is a category so little highlighted. Even if it had to last more than 2 hours why not ! I encourage you IgorSantos777 ! Continue this beautiful TAS, it's very technical, difficult and I hope you will bring this project to an end.
Masterjun
He/Him
Site Developer, Skilled player (1987)
Joined: 10/12/2010
Posts: 1185
Location: Germany
You missed my point. I did not say we can't accept such a glitchfest (because we absolutely will). I said a run with more variety would be superior.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Editor, Skilled player (1344)
Joined: 12/28/2013
Posts: 396
Location: Rio de Janeiro, Brasil
I agree with most of Masterjun's points. Igor is doing a great work and if it's submitted as it is, it surely should be accepted, but there are some ways to improve even more the freerun. The stun glitch is obviously overused. It is an awesome glitch and it must be performed more than once in the glitchfest, to show off many of its possible outcomes, but Igor is doing it in almost every stage in a way it gets really repetitive. Also, I believe he should spend less time in each level. You don't need to show everything a level has to offer if most of it can be shown later on. Lastly, I consider it's a very bad idea to do 96 exit. I would like to see every world done, but through a route that only shows the levels with the most entertainment potential.
My YouTube channel: https://www.youtube.com/channel/UCVoUfT49xN9TU-gDMHv57sw Projects: SMW 96 exit. SDW any%, with Amaraticando. SMA2 SMW small only Kaizo Mario World 3
Lil_Gecko
He/Him
Player (98)
Joined: 4/7/2011
Posts: 520
Oh, that's a surprise. It does indeed look very good so far ! Have you already plan ahead for the next world ? I don't think the length will be an issue, my main concern is that you already use a lot of glitches and it might get a little repetitive in the end. The amount of Item Swap and Stun Glitch so far didn't bother me, but it's just the first two worlds yet, so let's see where this is going. I also agree with Masterjun's ideas. For example you can showcase the new small only levels that just became possible ?
Amaraticando
It/Its
Editor, Player (159)
Joined: 1/10/2012
Posts: 673
Location: Brazil
Lil_Gecko wrote:
For example you can showcase the new small only levels that just became possible ?
There's a WIP of this category.
Editor, Skilled player (1439)
Joined: 3/31/2010
Posts: 2108
brunovalads wrote:
Since the old Glitchfest is dead, IgorSantos777 started a fresh new one, and it's really nice so far: Link to video
Man, this was seriously awesome. It's been a while since I last went WTF at a TAS.
Experienced player (634)
Joined: 5/2/2014
Posts: 21
Well, I quite liked Masterjun's suggestions. I was thinking and it really has many levels that there is not much to do and with that it would be unnecessary to do them. So, I'll take a good look at what each level has to offer and decide which ones are worth doing. With that I intend no more to do the 96 exits, but yes, only the exits in which there is a great potential of entertainment. I'm also going to try to pass some levels without jumping, I've never tried to do this, but I think I can. Receiving all the items and coins is a cool suggestion and I can do it from now on. About the repeatability of the stun glitch, I agree with Bruno and after having seen my video a few times I realized that it really is too much and so I decided to do it only when the result is satisfactory not to get so repetitive. In short ...I will try to make the suggestions of Masterjun to make this run the best possible, I will try not to spend much time in the levels, I will try not to make it so repetitive and try to show as many glitches as possible. And finally, thanks to everyone for the suggestions, I will do my best! and sorry for the bad english :/
Experienced player (634)
Joined: 5/2/2014
Posts: 21
Lil_Gecko wrote:
Have you already plan ahead for the next world ?
Certainly. I already have a base to do in practically every world.
Joined: 6/18/2014
Posts: 11
NEW CREDITS WARP WITHOUT ACE FOUND BY AMARATICANDO No ACE used! Yes, it is as simple as it looks. For once there's not a lot of technical behind-the-scenes code-writing here; just writing directly to the gamemode! In fact, this glitch is an extension of what we used to call the Null Sprite glitch, so if you're familiar with that, it's the same thing except we can now get more slot numbers than just 0xFF in Yoshi's mouth. Specifically the one used to get to the gamemode is slot 0x1C. This glitch is just one step away from the normal Suicide Swap (where you drop a Yoshi in a pit as a new sprite spawns, and it teleports to Mario). bahamete found this glitch ages ago, and it's actually used in the current 96 exit TAS. Now what's new is, if Yoshi's mouth is full when you drop him in the pit, there is actually one frame where you can spit it out as the platform spawns, effectively making you spit out the item from the platform's mouth instead. There is a miscellaneous sprite table at $160E (different sprites use it for different things), which Yoshi uses to determine the slot number of the item in his mouth. When his mouth is empty, it's set to 0xFF, but there are various ways to get this value while his mouth is still full (Null Sprite glitch). Now we can get more values than just 0xFF, based on what the platform uses $160E for. According to this chart, the platform uses $160E for "Tile number of the rope the sprite is currently touching, minus #$76." So this gives us a range of values that the game will interpret as the slot number in Yoshi's mouth. If you spit something out on this magic frame, the game will try to update 11 different addresses at 11 different property tables for the thing coming out of his mouth. It uses the slot number to index these tables, but the tables are only 12 bytes long, so if our slot number is higher than that, it will start writing to other memory. Specifically in the gif, the value at $160E (and thus the slot number), is 0x1C so it uses that value to index the 11 tables. One of these tables is $E4, which is supposed to be the X position (low byte) of the item coming from Yoshi's mouth. But it writes this value to 0xE4 + 0x1C which is $100, which is the gamemode! It normally writes the value based on Yoshi's position, but since in this case Yoshi is the platform, it writes based on the platform's new position after teleporting. In the gif, this position is 0x20 so the gamemode is set to 0x20, which is the "Fade to enemies" gamemode. (Full list of gamemode values can be found here). By altering the position, we can get more values for the gamemode as well, but this is just an example. Is this useful? Well, it's rather slow (compared to ACE), since you have to go to Cheese Bridge Area to get line-guided sprites to swap with. You also need to bring something into the level to make Yoshi's mouth full, which requires either PI, or eat-chuck. If nothing else, this is a new way to do stun glitch, if you swap with a sprite who sets $160E to a value between 0-11. You can even make "Yoshi" spit out a Yoshi, which was previously only possible with PI to get a reserve Yoshi. You can also possibly get more values than just 0x1C when swapping with the platform, to get some new possibilities for addresses to write to. These have not been exhaustively looked into or tested, but it looks like vanilla SMW will be pretty limiting in this regard. The bytes that get written with slot 0x1C are:
Address----Purpose-----------------------------What's supposed to update--------Possible values
0x00C6-----misc table $C2 for slot 4-----------sprite Y speed-------------------0x00 
0x00D2-----player Xpos high byte---------------sprite X speed-------------------0x10 or 0x30
0x00DE-----sprite Ypos for slot 5--------------misc sprite table----------------0x00 
0x00F4-----empty-------------------------------Ypos low byte--------------------near platform's Ypos low byte
0x0100-----gamemode----------------------------Xpos low byte--------------------near platform's Xpos low byte
0x14E4-----sprite Xpos high byte for slot 4----sprite state---------------------0x00 or 0x09 or 0x0A
0x14F0-----sprite Y subspeed for slot 4--------Ypos high byte-------------------near platform's Ypos high byte
0x14FC-----sprite X subspeed for slot 4--------Xpos high byte-------------------near platform's Xpos high byte
0x1598-----misc table $1594 for slot 4---------horizontal direction-------------0x00 or 0x01
0x15EC-----sprite OAM index for slot 2---------licked flag----------------------0x00 
0x1642-----misc table $163E for slot 4---------kill counter---------------------0x00
Amaraticando
It/Its
Editor, Player (159)
Joined: 1/10/2012
Posts: 673
Location: Brazil
Thanks for this text, Nathan. Quoting bahamete:
Yoshi uses sprite table $7E00C2 to determine his current state. #$00 = not riding, #$01 = riding, #$02 = running away (hurt). The bug occurs when a sprite spawns in Yoshi's (now unused) slot using an initial value of #$01. The game determines that Mario must be riding this sprite and, as such, teleports the new sprite to underneath him. It quickly corrects itself (as there are no Yoshi's present) but the one frame is sufficient to trigger the glitch. This isn't such a technical explanation but I hope it is sufficient.
The problem is, many sprites use $C2, but only a few set it to 1 between the 1st and the 2nd frame of existence.
Editor, Skilled player (1344)
Joined: 12/28/2013
Posts: 396
Location: Rio de Janeiro, Brasil
One of the most famous SMW pit hacks, Item Abuse 3, has just been TASed. And what a brilliant TAS. https://www.youtube.com/watch?v=VXZ54MGWgN0 Panga's video has been taken out by Nintendo, so you might want to watch the previous TAS before in case you don't know the intended solutions: https://www.youtube.com/watch?v=JuZMvMtZCWU
My YouTube channel: https://www.youtube.com/channel/UCVoUfT49xN9TU-gDMHv57sw Projects: SMW 96 exit. SDW any%, with Amaraticando. SMA2 SMW small only Kaizo Mario World 3
Lil_Gecko
He/Him
Player (98)
Joined: 4/7/2011
Posts: 520
This is awesome. Thanks for sharing.
Amaraticando
It/Its
Editor, Player (159)
Joined: 1/10/2012
Posts: 673
Location: Brazil
Yet another rope mechanism climbing technique.
Challenger
He/Him
Skilled player (1689)
Joined: 2/23/2016
Posts: 1061
Amaraticando wrote:
Yet another rope mechanism climbing technique.
Cool find! Looks definitely useful for a new 96-exit TAS.
My homepage --Currently not much motived for TASing as before...-- But I'm still working.
Editor, Skilled player (1344)
Joined: 12/28/2013
Posts: 396
Location: Rio de Janeiro, Brasil
Challenger wrote:
Cool find! Looks definitely useful for a new 96-exit TAS.
I should announce here as well, there is a 96-exit TAS in progress already. I'm currently at Butter Bridge 2. Thankfully for me, this new trick won't improve Gnarly. It would suck to line up every frame rule again! So far, 58/96 exits have been done, and improvement is of 1 minute and 37.8 seconds. Rerecord count is around 627k.
My YouTube channel: https://www.youtube.com/channel/UCVoUfT49xN9TU-gDMHv57sw Projects: SMW 96 exit. SDW any%, with Amaraticando. SMA2 SMW small only Kaizo Mario World 3
Joined: 5/12/2009
Posts: 748
Location: Brazil
BrunoVisnadi wrote:
So far, 58/96 exits have been done, and improvement is of 1 minute and 37.8 seconds.
That would be already crazy enough if the improvement was for the whole run, and to think it's possible that it can be over 2 minutes when the run is done. Congratulations on the great work, man!
Editor, Skilled player (1344)
Joined: 12/28/2013
Posts: 396
Location: Rio de Janeiro, Brasil
Reeve wrote:
BrunoVisnadi wrote:
So far, 58/96 exits have been done, and improvement is of 1 minute and 37.8 seconds.
That would be already crazy enough if the improvement was for the whole run, and to think it's possible that it can be over 2 minutes when the run is done. Congratulations on the great work, man!
Thanks for the compliment! I will make a post here occasionally about the progress of the run. Also I want to say I initially intended to publish at least one WIP of this run, but I realized doing so would very likely result in someone finding some improvement early on, and I know I would want to go back and fix it, which would make this run to never be finished. So I rather not knowing about possible improvements until I'm done :) Thus, I'm just sharing some nice strats or levels on Twitter, for those interested. This was the final version of Butter Bridge 2, for instance. A rare use of double stun glitch: https://twitter.com/Bruno_Visnadi/status/1013565795790663681
My YouTube channel: https://www.youtube.com/channel/UCVoUfT49xN9TU-gDMHv57sw Projects: SMW 96 exit. SDW any%, with Amaraticando. SMA2 SMW small only Kaizo Mario World 3
Expert player (2567)
Joined: 12/23/2007
Posts: 830
BrunoVisnadi wrote:
Thankfully for me, this new trick won't improve Gnarly.
Amen to that, so you don't have to go through the redo process again.
BrunoVisnadi wrote:
Also I want to say I initially intended to publish at least one WIP of this run, but I realized doing so would very likely result in someone finding some improvement early on, and I know I would want to go back and fix it, which would make this run to never be finished. So I rather not knowing about possible improvements until I'm done :)
So that's why. I finally get it. I guess it's what happened to ISM back when he was making the 96-exit TAS. Maybe it's better to keep in discussion with people who are capable of finding improvements once a stage is done, so you'll have more confidence with no improvement being left? Looking forwards to your masterpiece!
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Editor, Skilled player (1344)
Joined: 12/28/2013
Posts: 396
Location: Rio de Janeiro, Brasil
HappyLee wrote:
So that's why. I finally get it. I guess it's what happened to ISM back when he was making the 96-exit TAS. Maybe it's better to keep in discussion with people who are capable of finding improvements once a stage is done, so you'll have more confidence with no improvement being left?
That could help, but it's hard to catch an improvement without the movie file. And I just can't TAS if I know there are improvements behind (if I happen to find one myself, I will definitely go back and fix it). So if I make the movie file public, chances are (considering this is a very big TAS) somebody will find an improvement somewhere, possibly more than one. I will end up going back to try to fix them, and god knows how many months it will take to sync everything, re-do score, coin and frame rule plan, etc. So I rather not knowing about any possible improvements, and thus, not sharing the file (neither a video of the whole run so far). Makes it much more likely that this run will be completed and submitted one day.
My YouTube channel: https://www.youtube.com/channel/UCVoUfT49xN9TU-gDMHv57sw Projects: SMW 96 exit. SDW any%, with Amaraticando. SMA2 SMW small only Kaizo Mario World 3
Expert player (2567)
Joined: 12/23/2007
Posts: 830
BrunoVisnadi wrote:
So I rather not knowing about any possible improvements, and thus, not sharing the file (neither a video of the whole run so far). Makes it much more likely that this run will be completed and submitted one day.
That's perfectly understandable. SMW TASing (especially the 96-exit TAS) is so complicated that very few people in the world can truly understand what you're dealing with. Best wishes to you and your project.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Editor, Skilled player (1439)
Joined: 3/31/2010
Posts: 2108
BrunoVisnadi wrote:
Challenger wrote:
Cool find! Looks definitely useful for a new 96-exit TAS.
I should announce here as well, there is a 96-exit TAS in progress already. I'm currently at Butter Bridge 2. Thankfully for me, this new trick won't improve Gnarly. It would suck to line up every frame rule again! So far, 58/96 exits have been done, and improvement is of 1 minute and 37.8 seconds. Rerecord count is around 627k.
Yo that's hella dope my man I'm so looking forward to the final TAS
Challenger
He/Him
Skilled player (1689)
Joined: 2/23/2016
Posts: 1061
I'm really surprised how you managed to save a great amount of time even when the current run looks very solid (considering how TASing this game needs a lot of skill/knownledge mechanics overall). 627k = lots of brute force management for every level.
My homepage --Currently not much motived for TASing as before...-- But I'm still working.
Editor, Skilled player (1344)
Joined: 12/28/2013
Posts: 396
Location: Rio de Janeiro, Brasil
As it turned out, there were some improvements early on to the 96 exit TAS (some of them apply to any% too). My plan of not sharing WIPs so that people don't find improvements didn't work. So I decided to share the same WIP ISM did in the past, Star World + Special World: https://www.youtube.com/watch?v=TdaqW4TeS40
My YouTube channel: https://www.youtube.com/channel/UCVoUfT49xN9TU-gDMHv57sw Projects: SMW 96 exit. SDW any%, with Amaraticando. SMA2 SMW small only Kaizo Mario World 3