Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I appreciate the efforts here to keep the author list comprehensive, but I'd actually prefer to not be included in this (or any future) author list, even if some entire levels from my previous run are used in the finished product.
Though I suspect as Evil_3D said, that they are improvable even without zips since a lot of Sonic engine knowledge has been gained by the community since this project. Still I look forward to checking this out later today.
(And also exciting to hear kaan is improving this with zips!)
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
My personal S-Rank encode would have fixed graphics and cutscenes removed, but with some additional tweaks to help with some of the things that happen when you yank out cutscenes:
Sound fixed - export the result without music, and manually mix it back in so it doesn't restart constantly and skip portions during door transitions. (I manually did this for my very old 100% TAS. I also cleaned up some grating sounds manually, like reducing the volume of some of the annoying or persistent sound effects. Unfortunately the video quality is poor due to my lack of understanding at the time, but it does display the potential nicely.)
Some small amount of atlas-like encoding: Not zoomed-out at all, but mostly I'm thinking that when Samus is near a door she's about to enter, the camera keeps her in the middle of the screen, showing the room she's about to enter in a pre-loaded state, to reduce the jarring effect of Samus quickly going from one edge of the screen to the other when entering a door without the door transition cutscene.
I assumed this would all be unrealistic, but maybe not with the rate these enhanced encodes are being cranked out? (It also would be quite removed from the actual game output at that point, but it turns out that doesn't bother me.)
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I loved this run and had no trouble following what was going on. I didn't find the graphical oddities distracting enough to be bothersome. Now, maybe this is because I watched the fixed graphics run first, or because I know this game's map and TASing techniques better than I know most any other game, but whatever the cause, that's my stance.
From a standpoint of "technical appreciation", I kind of understand concerns about the introduction of a powerful technique obsoleting the need for a lot of smaller and maybe trickier optimizations, but I also don't think this technique just makes the game simple to TAS. There seems to be a wide breadth to the new tricks here, which require their own optimization that is different, but to me, technically impressive all the same (and more often than not, more impressive to me from a non-technical perspective).
And I also like the idea artistically of the game's graphics getting a little more screwed up as the run goes on. It's like a visual indication that the game is buckling under the pressure of the TAS.
---
Notes as I watched the non-fixed version:
The bowling room was slightly confusing to me at the beginning, probably because it's using a technique I don't understand to fall through the floor initially, but I figured out what was going on as items were collected.
In the pre-spring-ball room, the fixed-graphics version does make much cleared what happens at the end (which was just silly clever).
"Swinging into a door via grapple nearby" before and after Draygon was so fun to see.
The final escape is too good, and I think even better on the non-fixed graphics version. It adds to the chaos of the planet at that point. Perfect ending to the run.
(Edit: Turns out the "list" markup doesn't work well with the "spoiler" markup. You gotta spoiler each individual list entry.)
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
There are so many cool things that have been implemented here. You can see how every magnetized ring eventually gets to Sonic or Tails. This will look real cool on some of the places where this is used heavily, like Hydro City 1 or Ice Cap 2. The view window I think is the perfect compromise between being zoomed out far enough, but not too far. The water effects and sprite layering is handled well. The important objects like rings are always shown. I think the camera's focal point between Sonic and Tails does a fantastic job of not making either of them look "more important" and really complements the cooperative dynamic in the gameplay.
I realize I'm giving a lot of opinions on this tool and I don't want to dominate the discussion, nor do I want to discourage any progress in this endeavor, regardless of the direction. I think this is such an exciting development and the idea it could also be used for other games is crazy to me. With all that being said, here are a couple more thoughts I have:
I worry about adding too many bells to this - things like dynamic zooming and dynamic split screen. Maybe they would look better than I realize, but part of what I think makes the atlas encodes so cool, especially for this game, is to be able to understand the path to the overall level much better, even not having memorized the map. You can see where paths double back on themselves and why. My worry here might also be bogged down in "this sounds hard and would delay the encode if efforts were put toward it".
This example looks significantly better than Hydro City 2 and Angel Island, partly because Carnival Night has a lot of decorative elements in the foreground, and the background of the stage is quite dark already, so the absence of the background layer is not as jarring. r57shell's atlas work includes some of this, though I don't know how it looks for every level, and also isn't zoomed out as far as Selicre's atlas encodes are. This is one of those things that seems like it would look amazing, but the practicalities of it may make it impossible to do in a visually satisfying way (especially in the general case).
In any case, I can't wait to see where this goes next!
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
Evil_3D wrote:
I wonder if you can increase the camera zoom between both players are far away from each other, this can fix some of the camera bugs?
Unless this can be done smoothly, it will be really jarring. It might look weird even if done smoothly. Zooming on pixelated stuff can look odd sometimes if it's not part of the original game, though I can't really articulate why. Maybe it's kinda disorienting? I'm thinking about the way that the original arcade Samurai Showdown dynamically zooms in and out during the battle as an example.
In LRZ2 there is a level warp, how the camera behaves in this particular spot?
Right now it would probably snap to the Sonic-Tails midpoint like any other place in the run, which I guess would be the middle of the level until Tails warped to Sonic's level-wrapped location, where it would snap again. If the camera had its own momentum that would accelerate/decelerate toward here the camera "should" be on that frame, maybe it could swing to the right with considerable speed until it "catches up" to Sonic?
In any case, it looks like some spots will need manual handling anyway - for example at 1:49, I dunno what the game does to handle that spinning orange pipe section, but it messes with the camera in an unexpected way. There's probably more spots like that throughout the run.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
Selicre, from the Youtube description, wrote:
I know how to smooth the camera panning
Is this what you're referring to around 0:35-0:38, where the movement seems slightly jerky as Sonic moves away from, then back towards, a stationary Tails? Cuz I was noticing that today on my one millionth rewatch of this video and don't really know how to describe it.
If so, I wondered if it would be smoothed out by having the camera position adjusted each frame based on having its own momentum which would accelerate the camera position towards that midpoint, rather than just following the midpoint exactly each frame. (Or maybe some combination of the two, to avoid the camera overshooting its intended target when there's a lot of momentum.)
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
The Angel Island 2 backtracking is to set up many trips for Tails to zoom through the floor with some version of the slope glitch (I think) to collect rings one-by-one in Knuckles' section near the stage's death barrier (i.e. the horizontal line that the game interprets as "you fell into a pit and died"). The death barrier is higher than it would be when actually playing through as Knuckles, so Sonic can't go down there himself without dying, and most rings are close enough to the death barrier that Tails has to simultaneously collect one while dying, so many trips are necessary. (Some rings here are so far below the death barrier that even Tails can't reach them.)
This early-stages atlas encode of the act may help shed some light on it, too:
Link to video
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
This level must have been a pain to route-plan. It doesn't seem to flow very well because of the amount of different places you need to go for rings that don't connect in a graceful way.
Some questions!
At 0:31 you send Tails up to get the rings above so you don't later have to go on that route, at the delay of waiting for him to respawn so you can get the 20 rings in the impossible wall. But later on, at 1:22, you take the lower route to get the 5x4 grid of rings at 1:26. What if you took the upper route instead with Tails, getting the 6 rings up top, and then sending Tails down below to get those 20 rings while Sonic got the 3, 5, and 3 ring boxes? Would that delay be shorter than the current delay at 0:31 waiting for Tails to respawn?
At 1:20, the gravity thing pushes you up pretty high. If you jumped straight up first and fell more into its effect, would that counteract its upward push enough to let you hit the ground faster?
At 1:46, the pinball section, I'm surprised this wasn't faster. It seems like Tails could hit the 4th switch sooner. What's he doing between 1:50 and 1:52, does he not interact with the bumpers off screen? What if you sent Tails below to collect that set of rings while Sonic activated the switches on the right side, could this speed things up? Or is collecting those rings without the electric shield much slower?
Why do you lose the shield at 2:12?
Edit: I found the answer to the last question just a few posts ago:
kaan55 wrote:
You can get rid of the lightning shield like I did in the testrun because the insta-shield is better for the start of Act 2.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
My opinion on a camhack is that I think one for Tails would work very well for this TAS. There is so much happening off screen that is impressive, if you can see it. All sorts of places where Tails goes off and grabs some rings, sometimes in an interesting way that only he can do as an off-screen object.
One random example, in FB1, 0:33, Sonic misses 4 rings with the electric shield. You can hardly tell in real-time. And then at 0:39, the ring counter goes up by 4 (also very easy to miss in real-time), because Tails stayed below ground and doubled back to get these rings before despawning while Sonic can continue forward above.
Taking it farther, a series of cam-hacked encodes combined with a NESAtlas approach would be an enormous undertaking (which I've thought a lot about how to practically accomplish.. first step would be to create full-res maps like exist for Sonic 1 and 2), but it would make for an amazing result. These maps are sprawling, beautiful, and packed with interesting things, and getting a sense of the entire route and how it's split up among the two characters simultaneously would have a huge amount of light bulbs going off as you immediately understood how much more was happening outside of the traditional screen boundaries.
---
Here is the list of questions I came up with over the past few days that could be used as a guide some of the likely questions during the submission process. I'm not looking for all of these to be answered myself, just to use as a guide for what someone of my knowledge level on TASing this game (slightly above the basics; I did some of the old Sonic 1 TASes many years ago).
AI2
A general explanation of the "death border" with maybe a map would be helpful. I remember doing this manually and counting the rings to figure out where that line was, though unfortunately I don't have the picture anymore. People may also ask why you can't zip through the wall in AI1 to fight the Knuckles boss and if that would allow for getting more of the rings in his area.
HC1
2:24 - Why does Sonic bounce up high at the end? (I assume for a favorable global timer in HC2)
HC2
1:20 - Does Sonic bouncing up in the hand spinning thing cost any time? Does it help anything or just look different?
2:03 - How does Sonic survive on these spikes?
2:05 - Would it be faster/possible to get Tails to spawn/land earlier and fly Sonic across these spikes? (Maybe that would lose time on the boss?)
MG1
Why generate the fire shield at the end of act 1? (Spend time for a global timer again, just with something cool looking?)
CN1
Same question on generating unnecessary shields at the end of act 1 - global timer manipulation?
CN2
1:40 - Could you repeatedly bubble shield this red spinner downward instead of going around the fans?
2:26 - The balloon bouncing *seems* a little snow but it's unclear whether that would matter anyway due to the timed object (spinning set of 3 bumpers) at 2:30
3:31 - How does Sonic go through the wall here? Is this a consequence of being on the Knuckles route somewhere else?
IC1
1:47 - Why not use the springs and platforms here? (I actually know it's to preserve the slope glitch but that's not obvious at first)
2:10 - An explanation what happens here would be helpful. I think I understand it - both bosses are spawned by Tails falling below, the one above is killed, the one below is almost killed and saved for later in Act 2 to kill midway through, in order to get 2 signposts to get both sets of underground ringboxes.
LB1
Is it possible to fight the Knuckles Act 1 boss instead of the Sonic boss to keep Act 2's water level down (while still getting all of Act 1's rings, or maybe having to come back for them from Area 2 via bonus stage level border resetting, if that's even possible...)
LB2
2:13 - Why detour to the left for a second, up the side of the wall, before turning around?
4:05 - Could you detour back to the Knuckles boss around 3:38 instead and save time overall here?
MH1
At the end of the act, could you regain control of Sonic to begin Act 2 like in some previous TASes? Or does this make total ring collection impossible?
FB1
0:55 - Why the extra little hop before spin-dashing to the left?
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
So I'm familiar with Sonic TASes but not many specifics of the mechanics behind this particular run. Would it be helpful if I sent a list of questions I would have as a somewhat-casual viewer as suggestions on what points would be useful to comment on in the submission text?
An example of the type of questions I would have: "In LB2 at 2:13, why did Sonic dash up the hill for a moment before turning around and going back the way he came?". (I assume the answer is related to Tails spawning mechanics, but I'm not sure.)
I can probably find some time to look through the entire thing so far - I've watched these WIPs so many times, it's silly :) - if you think that would be helpful.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
It all looks so smooth! LB2 looks fantastic and I'm happy you were able to avoid going into the bonus area and Tails' drowning death, funny as it was :)
Is the extra delay near the end of SO1 (just before you get the last big ring) in order to change the global timer for SO2 for the spiked pillar moving up and down at 0:57? I could not tell what you did differently in SO2 (except that you didn't spend as long in the walls around the 0:55 mark).
Oh, and for anyone else wanting to watch this who maybe hasn't been following the entire time, I got the savestate needed to play this movie from this post. (Or this direct link.)
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I'm loving this ring attack. I am consistently surprised at the amount of time saved with each new iteration. Seeing you find improvements of 30 seconds in HC2 and IC2 is really surprising. I know it's a very tough balance between optimizing as much as possible, and knowing when to move on to the next level in order to ever make long-term progress, but it seems like just about every level is being tightened up quite a bit by this point. It's impressive!
Looking at the ring totals from Sonic Zone 0 across what's been done so far (up to Lava Reef 2) and comparing with the Knuckles ring attack where applicable, you've gotten every available ring except for the ones in AI2 below the "death line" and 10 in LB1 due to mutually exclusive boss ring monitors. But! You've also picked up an extra ring in FB2. I'm wondering how you found out about this one ring. Is there some kind of map-searching utility or part of the LUA script which scrapes level data for the total number of rings? Is it possible there are other goofy hidden wall rings that aren't known about yet like this one?
Anyway, thanks for keeping on with this run. It's been a joy to watch.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
Bisqwit wrote:
Some of the detours to kill enemies looked weird. Such as at 6:31 in 3–3 to kill a red turtle. I guess they were because of slack time waiting for the global timer run platforms/turtles in the end… Would have been nice if this was addressed in the submission comments. But in the end (of 3–3), there were no global timer run events. What’s going on? Was it because of the coins? But weren’t coins categorically excluded? Ehh. I’m a third into the movie already and I still don’t know what the goals are.
This explains some of it, from the submission text:
1UP Mushrooms in X-1 (except 1-1) only appear when we get all the coins in (X-1)-3 ( For example 1UP Mushroom in 4-1 will appear when we get all the coins in 3-3). So in this run you can not only see all the ways of getting items , but also see coin collection in X-3 stages. In 1-3 we can miss two coins.
So I think in all* the #-3 stages, all the coins are collected - not for their own sake, but so that the 1UP appears in the next world's #-1. So the 1UPs matter directly for the item collection goal, but the coins do not, except indirectly via this mechanism. (Whether or not individuals agree with this definition is a different discussion.)
*Except for 1-3 which has a (hard-coded?) exception, and 8-3 where it doesn't matter because it's the last #-3.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I liked the less-glitchy Super Metroid TASes from years ago, but I like these newer ones, too. I think this round of new techniques is just that. I don't see them as breaking some definition of "major", in my mind. They certainly allow routes and skips that weren't possible before, but that's been the case with all the previous large-scope tricks, too. The type of things that the moonfall opens up seem similar (though silghtly smaller in scope) to me as what the short charge made possible. I see these X-Ray/jump/speed tricks as very analogous to when arm-pumping was discovered: it's a method that allows for new techniques and routes all over the place, yet it doesn't look that great.
Even so, the messed up graphics here seem very minor to me and even provide some entertainment (for example, in Maridia, Samus X-Ray sparking while scanning up and down was fun). Most of the graphical glitches I saw were a door in the middle of the screen instead of at the edge. At no point was I confused as to which graphics were "real" (note that this doesn't mean I understand why the visual glitch happens). Mother Brain's room was the only place where I thought, "eh, that's a bummer". But it's not like I watch v12 of this TAS to watch the scripted Mother Brain fight anyway. I couldn't tell you the last time I watched a SMW Bowser fight, either.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I've been following this TAS's progress on YouTube for a while now and my opinion on AI2 is that it's better to get the rings than not. I have found a lot of the enjoyment of this TAS to be the careful coordination between Sonic and Tails. The AI2 from Tails's perspective really opened up how complex this got, and has me noticing and enjoying a lot more of the coordination in the other levels.
Maybe a side publication with the entire run from Tails's perspective would help with others understanding the delays in AI2 and some other levels for the sake of maximum rings. My pie-in-the-sky encode of this run would be an Atlas style video that had Sonic, Tails, and the nearby areas to show some of the crazy paths that electrified rings take.
In either case, I would still vote Yes for the submission.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
BrunoVisnadi wrote:
I think this is a valid pattern of covering an nxn grid with n-1 lines.
Yes, this is the same pattern I've found for nxn. Nicely done!
BrunoVisnadi wrote:
for larger values of n, this pattern doesn't work.
As far as I can tell, it does work. By drawing a series of n-2 parallel lines across the grid that cause the pierced squares to form a pattern of 2 in one direction (either x dimension or y dimension), and 1 in the other dimensional direction, etc., you can fill everything but 4 squares in n-2 lines, and one more line can cover those last four squares. Here's an example for 6x6 using 5 piercing lines:
[img size=400x400]https://i.imgur.com/pXOkyJJ.png[/img]
The colors of the squares show which lines have pierced which squares. After working on these for a while, I've found it helpful to think about these examples in terms of the squares they hit rather than the line itself. (I just have to be careful I'm actually selecting a 'path' which is possible for that line.)
I haven't formally worded the algorithm above yet, but I believe that n-1 piercing lines is an upper bound for all nxn grids, where n>2. I have proven this value is also a lower bound for n=3, 4, and 5. The proof is related to how many squares it's possible for an individual piercing line to pierce in the best case, but the grid size increases too rapidly for this approach to work for n>5. It feels like it should be right to me, but I've struggled to formalize anything about it so far.
thatguy wrote:
it would have lines which crossed each other - but of course, this introduces inefficiencies to the problem, as some squares would necessarily be pierced more than once.
As a general strategy, it's better to go for not-yet-pierced squares with subsequent lines, but sometimes you end up getting already-pierced squares "for free" with a later line, as seen above with the blue line.
thatguy wrote:
Just start with the configuration you mentioned, each line piercing a single column of squares. Now tilt the first line slightly so it pierces the top square in the second column, then tilt the second line a little more so it misses the top square in its own column, but pierces the top two squares of the third column
This is a good observation. To generalize it a bit: there are some piercing lines which are not "optimal", meaning, there exists a separate piercing line which pierces all of the same squares as the original line, plus one or more additional squares.
Something else to notice is that the number of squares pierced by a particular piercing line is equal to the number of gridlines that line intersects, plus 1. This is because in order for the path of a piercing line to "move" from one square into another square, it must "cross" a gridline boundary. (Add one more for the square that it started in.)
In my analysis, these two observations lead me to these properties of "optimal" piercing lines:
1) Piercing lines should not be parallel with respect to any dimension.
2) Piercing lines should start and end at the opposite ends of the grid of at least one of the dimensions of the grid.
Example of 1, not optimal:
[img size=400x400]https://imgur.com/6HwaFUf.png[/img]
Optimal:
[img size=400x400]https://imgur.com/Lax2TNQ.png[/img]
Example of 2, not optimal:
[img size=400x400]https://imgur.com/V166XZR.png[/img]
Optimal:
[img size=400x400]https://imgur.com/M2Tt2sN.png[/img]
These strategies seem to hold true in higher dimensions as well. If a piercing line doesn't cross at least n-1 gridlines across one dimension, and cross at least 2 gridlines (or 1, if near an edge of the grid) in all other dimensions, there is a piercing line which can capture all the same squares, and more.
BrunoVisnadi wrote:
I suspect for larger n it might be possible to do it with less than n-1 lines.
I have kind-of similar suspicions, but they are about adding dimensions rather than increasing the grid's edge size.
For example, a 3x3x3 grid of cubes could be solved with 6 piercing lines, by using the solution above for each 2D slice of the 3D grid. But none of these 6 lines are optimal because they are all parallel with respect to the new dimension. Could more optimal lines allow for a 5-line solution?
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
This is a problem I came up with a little while back and have enjoyed kicking around, especially if I have time to kill and not much around to help kill it. I initially thought of this in an airport when I noticed some patterns in the floor tile.
The initial problem:
You are given a 3x3 square grid in the plane, through which you may draw as many straight "piercing lines" as you would like. A particular square on the grid is considered "pierced" if it contains one or more piercing lines within the area defined by its edges. The goal is to "skewer" the grid: pierce all the squares in this grid.
(Note: piercing lines drawn exactly along a border between two rows/columns of squares do not count as piercing any squares - otherwise the problem becomes trivial).
You can clearly do this in 3 lines, by drawing a line through each row or each column. Can you do it with just 2 lines? How about 1 line? Can you prove the lowest value (prove that a solution with less than a certain number of lines does not exist)?
Larger grids:
How about a 4x4 or a 5x5 grid? Do they follow similar patterns from the 3x3 case?
How about an NxN grid? Can you come up with an algorithm for the general case to get an upper bound less than N? Can you prove that this is indeed the lowest number of lines possible?
Grids in higher dimensions:
Given a 3x3x3 grid of cubes, define piercing lines in a similar way: a cube is pierced if it contains one or more piercing lines within the volume defined by its edges.
We can clearly skewer this grid in at least 3 times the optimal solution for the 3x3 grid, as we could draw piercing lines for each "slice" of the third dimension. Can it be done with fewer lines?
From here the problem can continue to expand into higher dimensions and larger sizes. I haven't solved all of the aspects of the problem which I've put forth below, but they've been fun to think about. Formalizing geometric stuff into constructs that are more mathematically.. uh, generalizable(?) is not my strong suit, so I haven't gotten very far with any of the proofs.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I struggle to understand the other perspective of the underflow glitch because it seems to amount to, "I don't like how it feels". I'm not saying that isn't valid, but I can't identify with it personally.
We could have said the same thing about arm-pumping a decade ago that people are saying about the ammo underflow now. Places all across the game become simpler to solve because you can cover more ground in the same amount of time (like heat-room runs using less health). Instead, it was taken on as an essential part of the future of Super Metroid TASing. There is no "arm-pump-less" branch, and there shouldn't be one.
The comparison to the GT code doesn't hold water for me, either. There are probably TASes of games with stage select cheats that aren't used which find other glitched ways to skip levels. I'll edit in an example if I think of one.
Another point of comparison: This run also shinesparks before getting the speed booster. Does that feel clean? If there were a GT-2 type code discovered that gave you the speed booster when you entered the Maridia tube room and held L, R, down, and select on a second controller, would that affect how the pre-boost sparks feel?
Underflow doesn't break the game, and doesn't split things into a significantly different (I understand this is a subjective judgment) type of run, the way that something like Space-Time Beam or the X-Ray wall clipping does.
Bring on the future!
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I don't see the underflow glitch as anything to specifically have a restriction around. I don't think it breaks the feel of this run, because the route selection and execution are my favorite parts of this game's TASing. Many other TASes also use underflow for game-breaking results, and don't generate this type of discussion.
Symphony of the Night underflows the number of gems, and then uses them all to get a lot of money quickly.
Many games that allow things on one edge of the screen to leak onto the other edge (horizontal or vertical) are likely due to over/underflow. Bionic Commando, Little Samson, Mega Man 2 (Crash Man stage at least), and probably many more.
I assume that that zipping past the left edge of the level to get to the right edge of the level in Genesis Sonic games is due to underflowing of Sonic's position.
I think Deadly Towers underflows the number of bells, which is used to beat the game without playing most of the towers, but I'm not sure on that one.
Anyway, I liked this TAS a lot. The number of shinesparks ending at the perfect moment was especially fun to see!
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I just watched this on YouTube and as soon as I saw that part in Star Light 2 I laughed and said "please let this be the TASVideos screenshot".
And it is, and I rejoiced. Apparently also a bunch of Sonic runs happened a while back that I either missed or forgot about.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Experienced Forum User, Published Author, Experienced player
(761)
Joined: 5/6/2005
Posts: 3132
I think this is one of the most entertaining TASes I've ever seen. Thanks to the authors for making it! Seeing that the Oil Ocean speed glitch was stackable absolutely cracked me up!
(I previously thought it was an "activate Super Sonic physics" glitch, but I'm now guessing it's an "increase current speed for tube journey without actually going through the tube sequence" glitch.)
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
