Merry Christmas TASVideos! This is our improvement of 1367 frames, or ~22 seconds over the previous run.

Game objectives

  • Emulator used: Snes9x 1.43 v17
  • Aims for fastest time
  • Abuses programming errors
  • Uses death to save time

Comments

This run has taken us a few months and we found many more improvements than we initially thought was possible. If you wish to enjoy the movie to the fullest we suggest you watch before reading ahead.

New Tricks

Chuck-Eat Glitch
This glitch was discovered by nathanisbored. The glitch abused the fact that chucks erroneously have the "Give Power-Up when eaten by Yoshi" bit set. Watching sprite table $7E:167A shows this. Thus, when we eat a chuck, we get something strange in our reserve box or a crash. When we eat a Clappin' Chuck as Fire Mario, it produces a goal sphere in reserve. Other chucks with other power-up statuses give other strange results but these aren't relevant.

Level comments

Level comments in (brackets) are over our previous smv, which is for the same submission but replaced 07/01/12 (dd/mm/yy)

Yoshi's Island 2 (4)

ISM suggested an idea of losing subpixels in order to perform additional corner boosts in this level. These are lost early in the level when 6/5ing. We perform 2 additional corner boosts. Then it became clear that performing one more boost would cause the goal to spawn a frame earlier; we managed a corner clip on a pipe near the end, which is set up via the corner boosts.
Lag is reduced at the goal too, by landing further to the left it seems. Lag is temperamental in this area.

Yoshi's Island 3 (27)

Many frames are saved on the duplications due to PangaeaPanga. Improvement of the sky level is taken from ISM's 96-exit wip.

Yoshi's Island 1 (-2559)

For information on the Item Swap glitch, please see the previous movie's comments.
Many optimization improvements occurred in this level. The aim of yi1 is to eat the Clappin' Chuck as fire mario by using the item-swap glitch; this yields a Goal Point Question Sphere, sprite 4A. We have two strategies to perform this. The first one (smv) is to collect a sprite coin from Jumping Piranha Plant while it is on Yoshi's tongue. We burn the Piranha Plant with Yoshi flames from the red shell near the midpoint. We turn the Piranha Plant into a coin, then collect it while it is on Yoshi's tongue. This triggers an item-swap. This strategy might be thought to be very fast, but because of bad placement of the Piranha Plant and the Clappin' Chuck, it costs a lot of time to die. The second one, which we've chosen in our run, is to fill sprite slots by dropping four resereved powerups and then use the fifth reserved powerup to item-swap. This strategy has an advantage, that we can die with no cost of time, but it costs a little time to collect powerups and we also had to R-Scroll.
(Until recently, these strategies matched after heavy optimization. But thanks to insight from regnum0nline (Dawn), 8 frames were gained for the latter strategy. This idea involved 49 hopping most of the way.)

Yoshi's Island 4 (-2559)

No change.

Iggy's Castle (1287)

The large portion of the improvement in this run happens here. The Goal Point Question Sphere we acquired from Yoshi's Island 1 is used.

Donut Plains 1 (1287)

No change.

Donut Secret 1 (1291)

Just before the first pipe entry, speed oscillation is changed from 17¨15 to 17¨16, which saves a subpixel and hence a frame. In addition, this enables us to beat the frame rule of sprite interaction so we can hit the red koopa a few pixels further to the right as it is unnecessary to slow down and it saves another frame.
To reduce key entry lag, we hit the sleeping fish to the left, with no loss of time thanks to the timing of cape spin.
(A frame is lost due to level transition lag.)

Donut Secret House (1290)

(A frame is lost due to level transition lag.)

Star Warp (1292)

We beat the frame rule of star warp and gain 4 frames on previous movie.
Related addresses:
  • $7E0013: In-Game Frame Counter
  • $7E1F19: Mario's Y Position in OW (2 bytes)
  • $7E1DF7: Star Warp speed
  • $7E1DF8: Star Warp Rotating Timer
OW transition starts as soon as $7E1F19 hits a certain value or less, and warp speed increments depending on the in-game frame counter, or more precisely, it increments when the value of the counter modulo 8 equals 1. So the star warp has a frame rule whose cycle is 8 frames long. Due to its strange behaviour, beating just a half of its cycle may save a frame , and hence the other half 7 frames.

Star World 1 (1296)

New optimization idea of smashola saves 4 frames. This gives us glitchy colours in this level, also.
(A frame is lost from level transition lag.)
Delaying key entry by 2 frames may remove 2 frames of lag.

Star World 2 (1296)

No change.
Delaying key entry by 4 frames may remove 2 frames of lag.

Star World 3 (1296)

No change.

Star World 4 (1302)

By platform boosting at the beginning, we gain some subpixels allowing us to corner clip once more. This saves us a frame. At the keyhole we reduce lag.
(A frame is gained from level transition lag.)
Delaying key entry by 3 frames may remove 2 frames of lag.

Star Warp (1302)

We managed to beat just the whole cycle of the frame rule, which means no time is lost due to the frame rule. If this was not the case, we could sacrifice a few in-game frames, in such a way we would end up losing no time after star warp, to remove lag frames. We were previously on this plan, but the new sw4 improvement finally enabled us to beat the frame rule and this old plan became useless.

Bowser (1367)

A new innovative method involving manipulating Bowser phases allows us to end input much earlier. We can reset a long counter during a phase by hitting Bowser at an appropriate time. For example in the second phase, by hitting Bowser as he begins his routine to drop a Big Steelie, we can avoid this and instead have him throw Mechakoopas on next round. Addresses 7E14B0-7E14B9 are important in this.
The credits would never be triggered by goal spheres being used at any places of Bowsers Castle. It only puts Mario to the overworld.

Thanks

  • Mister thanks #TASers and pirohiko
  • #tasvideos
  • amaurea and gocha for their helpful Lua scripts
  • nathanisbored for finding Chuck-Eat glitch
  • ISM for some ideas used in this run
Thanks for your patience while we worked on this! Enjoy.

Translation into Japanese - 対訳

メリークリスマス!前記録から1367フレーム、およそ23秒の更新です。
この対訳は Mister (Mr.) によるものです。ところどころ私の主観が入っています。

仕様

  • 使用したエミュレータ:Snes9x 1.43 v17
  • 最速を目指す
  • プログラムエラーを悪用する
  • 時間短縮のための死亡を許す

コメント

この動画を製作するにあたって数ヶ月を費やすこととなりましたが、当初想定していたよりもはるかに多くの更新点が見つかりました。この動画を最大限に楽しみたい場合には、先に動画を見てから続きを読むことをお勧めします。

新テクニック

ブル食べバグ
このバグは nathanisbored によって発見されました。このバグは「ヨッシーがブルを食べるとパワーアップとして認識される」という(誤って設定されたであろう)仕様を利用したものです。本来食べることのできないブルを口寄せバグ等によって食べると、マリオのパワーアップ状態が変化したり、本来想定されていないストックアイテムが得られたりします。この動画では、ファイアマリオで手拍子ブルを食べることで、ゴール玉(ラムネ海溝の沈没船のゴール)をストックしています。

各コースのコメント

括弧の中の数字は前記録との差を示しています。ルート変更が著しいため、ヨースター島コース1からイギーの城まではあまり参考になりません。

ヨースター島コース2 (4)

ISM のアイディアをもとにした二つの変更点があります。ひとつは、ゴール口寄せ直前での挙動の最適化、もうひとつは、コーナーブースト等を可能にするために数サブピクセルを犠牲にするというものです。ISM によると「コーナーブーストを2回追加することはできても、ゴールを1フレーム早く出現させるには1サブピクセルだけ足りない」とのことだったので、更に数サブピクセルの犠牲を払い、土管での角抜けを1回追加し、1フレームの短縮につなげました。ヨッシーが居ればこのタイプの角抜けが出来るというのは、この時点ではまだ誰も知らなかったようです(もちろん私も)。

ヨースター島コース3 (27)

PangaeaPanga の妥協のない最適化によって、ブロック増殖部分が大幅に短縮されました。今回はヨースター島コース1でヨッシーを使うため、ここで捨てずに持ち越します。翼を取ったあとの空の面では、ISM の全クリ TAS を参考にしています。

ヨースター島コース1 (-2559)

アイテム入れ替えバグ(または口寄せ)については前記録のコメントを参照してください。
この面での目的は、ファイアマリオになり、ゴール手前に配置されている手拍子ブルをヨッシーで食べることです。ブルを食べることはできないので口寄せバグを使うのですが、これを実行するための作戦として、以下の二つが考えられます。まずひとつ目は、スプライトコインを用いた口寄せを使う方法です。まずコースなかほどに設置されている赤甲羅を使い、ブル手前のPパックンをスプライトコインに変えます。そのコインを舌で捕らえ、舌の上のあるうちにコインを取ります。すると舌の上にコインと同じスプライトインデックスを持った無が一時的に生成され、そのインデックスにブルが出現するようにすると、ブルが舌の上にワープして口寄せ成功です。この方法はRスクロールを必要としない分、後述するふたつ目の方法よりもずいぶんと速いように思えますが、手拍子ブルとPパックンの配置が悪く、死ぬために大きく時間をロスしてしまいます。
次にふたつ目の方法です。口寄せそのものに使われるテクニックは、ストックアイテムの二個食いです。これはヨッシーが舌でストックアイテムを捕らえるのと同時に、体当たりでストックアイテムを取ってしまうというもので、この方法でも上と同様にして舌の上に一時的に無が生成されます。ここで生成される無のスプライトインデックスは、ストックアイテムと同じものです。スプライトのインデックスは0番から11番までの12個あり、ストックアイテムやブロックから出るアイテムは、空いているインデックスのうちで一番大きいインデックスに出現します。これに対してヨースター島コース1では、ブルや通常の敵は0番から7番のインデックスにしか出現することができないので、口寄せを行うためにはストックを7番以下に出現させなければなりません。ストックから出現するキノコやファイアフラワーは画面外に出ても消えないという性質を持っていて、したがってストックを4つ落とすことで、インデックス8番から11番を埋める事ができます。このためにはストックが合計で5つ必要になる計算ですが、この面に配置されているパワーアップアイテムは羽つきブロックの中のファイアフラワー、草陰のキノコ、コース終盤にあるはてなブロックの中のファイアフラワー、の3つです。草陰のキノコのみ、ブロックから出るアイテムではないため、二個食いで増殖することができますが、それでもまだストックは4つ分しかありません。ここで赤甲羅の上のブロックにある1UPキノコを利用します。もちろん1UPをそのまま取ってもストックを増やすことはできませんが、ストックを二個食い増殖するのに使うことができます。この動画では、この地点までに既に2つストックを落としているので、10番と11番のインデックスが埋まっている状態です。この状態では、ストックは9番に出現することになります。そのストックを二個食いすると、ストックにキノコが得られ、舌の上にはインデックス9番の無が生成されます。この無をそのまま食べるとキノコとして扱われるのですが、ストックに入った1個目のキノコを落としてからでなければ増殖に成功したことになりません。しかし、このストックのキノコはインデックス9番に落ちてくるので、口寄せによって舌の上にワープしてしまいます。そこで、1UPキノコをあらかじめ9番に出現させ、この1UPキノコが消える直前にストックキノコを出すと、このキノコは8番に現れます。そして1UPキノコが消えて9番が空いたあとに、キノコを二個食いしてストックを落とすと、無が8番、ストックが9番となって口寄せは起こらず、無を食べることができ、結果としてストックを増殖したことになります。この手法は最後のファイアフラワーでも使われています。
検証の結果、このふたつの方法はまったく同じ速さになると思われていましたが、regnum0nline (Dawn) の最適化案により、ふたつ目の方法が8フレーム速くなると結論付けられました。彼のアイディアは草陰のキノコ以降慣性移動するというものです。
ヨースター島コース1でゴールすると、ゴール後の長いデモが流れ、マップ画面ではマリオがかっぱ山黄色スイッチまで歩いていってしまうため、大きく時間をロスしてしまいます。これを回避するために、ゴール玉をストックしたら即座に角抜けで死亡しています。死亡によってヨッシーを失ってしまうので、ヨッシーが必要になるヨースター島コース3を先にクリアし、そのあとヨースター島コース1に戻るというルートを取っています。

ヨースター島コース4 (-2559)

変更点はありません。

ヨースター島の城 (1287)

ヨースター島コース1で取得したゴール玉を使いゴールします。強制スクロール部分と、卵救出&城破壊デモをスキップすることができ、大きな更新につながりました。

ドーナツ平野コース1 (1287)

変更点はありません。

ドーナツ平野秘密のコース1 (1290)

最初の土管に入る直前に、1/1を一瞬だけ解除して1フレームの短縮。スプライトの当たり判定は2フレームに一度しか行われないため、先の1フレームの更新により赤パタパタを1フレーム分近くで叩くことができ、更に1フレーム短縮。鍵穴エフェクトのラグを軽減するため、グースカを左側にはじきました。これによる数ピクセルのロスがありますが、マントアタックの周期の関係で、そのロスがなくとも鍵ブロックを叩くことのできるタイミングは同じため、結果としてロスはありません。前記録では、ラグを軽減するために鍵の挿入を1フレーム遅らせていましたが、今回はそれを回避して1フレームの短縮。さらに鍵穴エフェクトのラグを1フレーム軽減、ゴール後のマップ読み込みラグで1フレームロス。
マリオの呼吸による気泡は拡張スプライトになっていて、これももちろんラグの増加の一因となりえます。この気泡は128フレーム周期で発生し、また真ん中の64フレーム目で特定のキー入力が行われていると、追加で気泡が発生します。つまり、周期の真ん中の気泡は入力を避けることで、発生しないようにできるということです。今回の動画では気泡調整でラグを軽減することはできませんでしたが、スコアが100050の場合には、気泡を一つ減らすことでラグを完全になくして2フレーム短縮することができます。

ドーナツ平野の隠れ屋敷 (1289)

スコア調整のためにクルクルブロックを壊して50点を取得しました。ロスなく50点を取れるのはここだけです。また、アトミックテレサを倒したあとのマップ読み込みで1フレームロスしています。

スターワープ (1292)

スターワープのフレームルールを破って3フレーム更新しました。スターワープには以下のメモリアドレスが関係しています。
  • $7E0013:ゲーム内フレームカウンタ
  • $7E1F19:マップ画面でのマリオのY座標
  • $7E1DF7:スターワープ時のマリオの速度
  • $7E1Df8:スターワープ時のマリオの回転タイマー
ワープによるマップ移動は $7E1F19 が特定の値以下になったときに始まります。また、ワープ速度はゲーム内フレームカウンタに依存して1から4まで増加します。より正確には、カウンタの値を8で割った余りが1のとき、ワープ速度が増加します。つまりスターワープは8フレーム周期のフレームルールを持っていることになります。一方でスターワープは少し変わった挙動をするので、周期のちょうど半分、4フレームだけ短縮した場合、ワープ後の更新量は1フレームあるいは7フレームとなります。1フレームの場合は更に4フレーム短縮すると残りの7フレームが更新されます。逆もまたしかりです。

ネイティブスターコース1 (1296)

前記録ではマリオの押し出し判定のフレームルールが合わず、このコースに入るのを1フレーム遅らせていましたが、今回はそれを回避することができました。コース内では smashola の新最適化案により、4フレーム短縮しています。また、クリア後のマップ読み込みで1フレームロスしています。
注意として、今回は適用していませんが、鍵の挿入を2フレーム遅らせるとラグが2フレーム減ります。

ネイティブスターコース2 (1296)

変更点はありません。ここも鍵の挿入を4フレーム遅らせることでラグを2フレーム減らすことができます。

ネイティブスターコース3 (1296)

変更点はありません。

ネイティブスターコース4 (1301)

コース内に入るときのマップ読み込みで1フレーム更新。bahamete によるリフト加速のアイディアに、改良型角抜けを追加することで1フレーム短縮。リフト加速のみや、通常の角抜けを追加するだけでは、次の角抜けにつながらず、短縮には至りませんでした。改良型角抜けとは、角抜け中に数サブピクセル犠牲にすることで、ブースト回数を増やし、結果的に通常の角抜けよりも数サブピクセル分のアドバンテージを得るための方法です。また、鍵の挿し込みを最適化して2フレーム短縮し、鍵穴エフェクトのラグも1フレーム軽減しています。
ここでも鍵の挿入を3フレーム遅らせることで、ラグを2フレーム軽減することができます。

スターワープ (1302)

スターワープのフレームルールの周期をちょうど破ったため、フレームルールによるロスは一切ありません(つまり1フレームでも遅れるとフレームルールで7フレーム遅くなります)。
ネイティブスターコース4のリフト加速による更新が見つかるまでは、フレームルールを破るには1フレームだけ足りませんでした。そこで、鍵の挿入をネイティブスターコース2で4フレーム、コース4で2フレーム遅らせることで、ラグを4フレーム軽減していました。一見遅くなっているように見えますが、6フレーム遅らせた分は、フレームルールのおかげで1フレームのロスに丸め込まれているので、トータルでは3フレームの短縮となります。ラグフレームではゲーム内フレームカウンタは増加しないので、ラグフレームの増減はフレームルールに影響しないというのが重要なポイントです。このように、フレームルールを利用して、ゲーム内フレームとラグフレームをトレードするというアイディアもあります。この記録が更新されるときにはきっと参考になると思います。

クッパ城入り口 (1367)

クッパの行動パターンを調整するという革新的なアイディアにより、キー入力の終了を大幅に早めることができました。クッパの行動パターンはあるタイマーで制御されていて、適切なタイミングでクッパにダメージを与えると、そのタイマーを強制的に特定の値にリセットすることができます。つまり、長い行動パターンの先頭でダメージを与えることで、クッパの時間を削り取ることができるということです。そのために1巡目でメカクッパを余分に出させて1体を持ち越し(444フレームのロス)、2巡目では2つ目の鉄球を投げるモーションに入った瞬間にメカクッパを当て(240フレームリカバリ)、3巡目の開始と同時にメカクッパを当てています(196フレームリカバリ)。この時点ではまだ8フレームロスしている計算ですが、ダメージ1発分のアドバンテージがあるため、実際には短縮に成功しています。また、そのアドバンテージにより、クッパ上空で入力を終えられるようになりました。完全にテンプレ化していたクッパ戦の見た目を大きく変えることができたので、今回の更新案はとても気に入っています。
ゴール玉をクッパ城のどの部分(クッパ戦を含む)で使っても、マップ画面に戻されるだけで、エンディングへと行くことは出来ません。

DarkKobold: Judging.

DarkKobold: This run will obsolete the current any%. We will not be changing the branch name.
DarkKobold:Delaying submission while the authors improve the movie. And you only thought it happened to Mario in 3 dimensions.

DarkKobold: Verified and replaced, setting back to accepted. Nice new bowser fight. This run will obsolete the current any%. We will not be changing the branch name.
ledauphinbenoit:Processing...


Editor
Joined: 3/31/2010
Posts: 1466
Location: Not playing Puyo Tetris
Voting no since this run: http://tasvideos.org/forum/viewtopic.php?t=12215 is faster and MUCH better at demonstrating how TAS differs from a "normal" speed run. I did enjoy this run. I saw the WIP on the forum. It was quite interesting to see the first castle skipped in such a fancy way. But, the new run, glitches the game so much that the game gives up and says, "The End."
When TAS does Quake 1, SDA will declare war. The Prince doth arrive he doth please.
Active player (451)
Joined: 3/21/2011
Posts: 127
Location: Virginia (United States)
hegyak wrote:
Voting no since this run: http://tasvideos.org/forum/viewtopic.php?t=12215 is faster and MUCH better at demonstrating how TAS differs from a "normal" speed run. I did enjoy this run. I saw the WIP on the forum. It was quite interesting to see the first castle skipped in such a fancy way. But, the new run, glitches the game so much that the game gives up and says, "The End."
I'm not sure if you realize that that run is aiming towards a completely different category.
YouTube Channel - Twitter Current projects: Sutte Hakkun, Hyper VI, RTDL, own hacking projects
Former player
Joined: 1/17/2006
Posts: 775
Location: Deign
Oh! Oh! This looks like fun! >>That's even more arbitrary. If you want to stay within the Mario universe, take SM64 for example. There used to be a 16-star run, but a 1-star run soon made short work of that. Now call me crazy, but that run seems to skip just as big a percentage of the previous run as the SMW credits glitch TAS. If I am wrong, why did the 1-star run obsolete the 16-star run instead of being put into a separate category? sm64 16 star run, what's the goal? Complete the game as much faster as possible than the fastest intended path (hint: it wasn't a 17 star run, or a 38 star run, or a 69 star run, etc, because 16 was fastest). So when it is discovered that you can do it faster than faster than the fastest intended path, it's still the same goal: to complete the game faster than the fastest intended path. Since fastest is now 0 star (hint: it still isn't 17 star run or 38 star run or 69 star run), it has the same goal as the 16 star. So it is the same category/branch/voodoo magic. sm64 70 star run, what's the goal? Complete the game in the fastest intended path (I think? not a sm64 guy myself). Why no accepted? Beats me. Perhaps it was too difficult to stomach using a glitch (BLJ seto007!!!!!!!!!!!!!!!!!!!!!!!) that would let you take a non-intended path through the game but collecting the same amount of stars as the intended path, thus making your goal choice self contradictory. Since the goal is the intended path. bwaaah?! smw11: the goal is to complete the game in the fastest intended path. If it were faster to go through vanilla dome like the smw small 15 exit run does, then that is the path it would take. But it's faster to take 11 exits (in an order that was intended by game devs). Which this run does do. This is why I pretty much always call it smw11 rather than smw any%. smw glitched: the goal is to complete the game faster than the fastest intended path. PS: I lied when I said Joey says Nyeh! 750 times. I forgot to multiply by 8. He says Nyeh 6000 times.
Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign aqfaq Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign Deign
Active player (264)
Joined: 4/15/2010
Posts: 198
Location: England
The run has already been accepted, just delayed for an improvement pointed out by Dawn, and the decision for separate categories has been confirmed by DarkKobold. I just hope he sticks to his guns :p You're free to do as you wish but I feel seperate categories discussion is best in here.
Retired smw-96, smw any%
Joined: 12/28/2011
Posts: 14
DarkMoon wrote:
DJWebb32 wrote:
voting no is quite frankly disrepectful to kaizoman666, bahamete, mister, and pangaeapanga for all the hard work they put into redoing a tas that most of them just submitted a few months ago. they started and completed this TAS with the goal of obsoleting the current any%, which they will have done once the new improvement mister found is put in, while the same cannot be said for masterjun, who made his TAS with the hopes of showing off a completely new and gamebreaking glitch and making a new category. great job to all 4 of you as this was a top notch TAS
Both TASes were exceptionally made and both should be accepted to the site, in my honest opinion.
I agree completely
Joined: 2/16/2005
Posts: 462
This run caused me to log in for the first time in over 3 years. Very amazing, voting yes.
This signature is much better than its previous version.
Player (13)
Joined: 6/17/2006
Posts: 511
DarkMoon wrote:
So using any glitch at all constitutes a "glitched" category?
In my opinion, there should be no such thing as a "glitched" category in the first place, but rather a "no player-defined restriction" category which should be the default. The existence of any other category by introducing player-defined restrictions should be based on whether or not introducing these restrictions is an interesting goal in itself and if the final result offers something significantly new and/or different from the others, such as "no save data corruption by resetting" or "no wall clipping" for example. In the case of this particular submission, the restriction added is "no credits glitch". In my opinion, this is not an interesting goal in itself, and the only thing new and different in such a category when compared to no restrictions or a glitchless restriction is the item box glitch itself, which would most likely be used in a (glitched) 96-exit restriction in more interesting ways. Sure, the item box glitch is interesting in itself, but I fail to see the pertinence to add a separate category just for swapping a level-skipping glitch with another one.
DarkMoon wrote:
Because you need Yoshi, of which there are none in YI1.
Ah I forgot about that. However, there's a Yoshi in YI2, so why wait until after beating YI3 before backtracking to YI1 to perform the glitch?
Joined: 12/22/2009
Posts: 291
Location: Michigan
SmashManiac wrote:
Ah I forgot about that. However, there's a Yoshi in YI2, so why wait until after beating YI3 before backtracking to YI1 to perform the glitch?
You need the Yoshi in YI3 to do the Yoshi-Wing exit.
Current projects: Yoshi's Island Disassembly Yoshi's Island any% TAS with Carl Sagan
Player (13)
Joined: 6/17/2006
Posts: 511
DarkMoon wrote:
SmashManiac wrote:
Ah I forgot about that. However, there's a Yoshi in YI2, so why wait until after beating YI3 before backtracking to YI1 to perform the glitch?
You need the Yoshi in YI3 to do the Yoshi-Wing exit.
Oh I see, and then there's no quick way to exit YI1 with Yoshi after triggering the glitch! Thanks DarkMoon for the explanation!
Joined: 2/16/2005
Posts: 462
I agree with the "glitched" run being a different category. While it's true both runs use "glitches", using memory corruption to jump directly to the credits is particularly destructive and no longer resembles normal gameplay. It is hard to call that "beating the game" when the actual game is bypassed. You're not talking about taking short cuts on the way to the finish line, you're talking about reconfiguring the course so the finish line is next to the start line. That is approaching the game with a different objective and warrants a different category.
This signature is much better than its previous version.
Active player (417)
Joined: 8/22/2008
Posts: 301
Location: Brazil
I think the better choice is: A glitched run, a low glitch using warps, a without Warps and a small only :P (I'm missing some? lol)
Editor
Joined: 3/31/2010
Posts: 1466
Location: Not playing Puyo Tetris
Marcokarty, then you run into arbitration. I still hate that Bowser boss battle. It's always so dull and drawn out.
When TAS does Quake 1, SDA will declare war. The Prince doth arrive he doth please.
Joined: 12/22/2009
Posts: 291
Location: Michigan
hegyak wrote:
I still hate that Bowser boss battle. It's always so dull and drawn out.
Hey, at least they brought up a new strategy this time!
Current projects: Yoshi's Island Disassembly Yoshi's Island any% TAS with Carl Sagan
Player (71)
Joined: 8/24/2004
Posts: 2562
Location: Sweden
Clear yes vote. Nice xmas gift. :)
Joined: 7/30/2010
Posts: 16
I'm way late to the party here, I'm sure everything's been discussed to death already, but I'll just throw this out there anyways just in case. What's wrong with SDA's method of differentiating these kinds of runs? The term "(With large skip glitches)" is appended to their speedrun titles with, well, with large skip glitches. It seems specific enough to draw a reasonably hard line between runs, but vague enough to include most runs of this kind. And as for categories, just speaking as a lay-person who generally just watches these TASs, I think the current system of new categories being decided by the community is best. New categories are great, especially for popular games, but need to be different/interesting enough to please the community at large.
Personman
Other
Joined: 4/20/2008
Posts: 465
PotatoHandle wrote:
What's wrong with SDA's method of differentiating these kinds of runs? The term "(With large skip glitches)" is appended to their speedrun titles with, well, with large skip glitches.
What's wrong with it is that (some of us think) this site should be primarily about pushing games to their absolute limit, no matter what. There's plenty of room for people to see entertaining runs with objectives that show more varied gameplay, of course, but the run that a new user first sees when searching for a game should be the absolute fastest, most broken thing we have to offer, and if one branch is going to receive most of the frame-warring attention, it should usually be that one. Other people obviously disagree, which is why there's all this brouhaha. In the end, it doesn't matter a huge amount -- everyone's runs can on the site, it's just a small matter of labeling. But it does feel unfair for the fastest run of a game on a site supposedly dedicated to, to quote the submission rules, "break[ing] all existing records" not to be displayed most prominently.
A warb degombs the brangy. Your gitch zanks and leils the warb.
Joined: 7/30/2010
Posts: 16
Maybe a slight redesign (Easier said than done, I know.) would be appropriate then? Give each game a single entry on the console's page, with a priority on speedrunning but would have a "View other categories" link or something similar for alternate runs, walkathons, PAL roms, warpless, playarounds, etc, etc. You could decide which run gets main page exposure by either drawing a hard line at fastest completion, or you could use the community ratings to choose the community's favourite of the categories, which would usually give you the most impressive and/or glitchiest run available. If you draw a hard line, then stuff like SMW where the desire for a more full run is likely, the 2-minute run would take precedence, but you could easily link to the 10-minute run in the description for added exposure, like I've seen done before to great effect; I use those links all the time. While all this wouldn't solve the problem per se, it would certainly lessen the blow a bit for those who make brilliant runs like this that many people will want to see; only to have it potentially rejected, or runs like chrono trigger, and still allow viewers such as myself access to all our beloved categories. It would also open up the doors a bit for people to add more creative categories, as they wouldn't be cluttering up the game page anymore. And would still be moderated by the community to filter out the garbage. Just some thoughts, hopefully there's at least something in there that hasn't been said yet, haha.
Joined: 6/26/2011
Posts: 167
Personman wrote:
In the end, it doesn't matter a huge amount -- everyone's runs can on the site, it's just a small matter of labeling. But it does feel unfair for the fastest run of a game on a site supposedly dedicated to, to quote the submission rules, "break[ing] all existing records" not to be displayed most prominently.
Technically speaking, the easiest way to fix this situation of "what should or shouldn't be labeled" is to force a label for every game with more than one category. So, hypothetically speaking, Super Mario World would end up with a "fastest glitched" run, a "fastest non-glitched" run, a "100%" run, and a "max% small-only" run. (Names tentative, of course.) No runs that are simply "Super Mario World". Of course, like any option, this one would also have its detractors. However, it would strike a compromise in allowing the absolute fastest runs to be showcased, while also permitting appearance of runs with fewer game-skips to get attention (to combat the "TASes are only glitchfests" misconception).
First a movie gets submitted, and ends up accepted despite breaking rules other runs have been rejected for. And when I vote less than spectacularly on this movie, I become the victim of harassment and threats. Yay, favoritism.
Personman
Other
Joined: 4/20/2008
Posts: 465
Saethori wrote:
Technically speaking, the easiest way to fix this situation of "what should or shouldn't be labeled" is to force a label for every game with more than one category.
This is pretty darn reasonable.
Potato Handle wrote:
Maybe a slight redesign (Easier said than done, I know.) would be appropriate then?
I think you're exaggerating the size of the problem. No good runs on the table right now are in danger of rejection, nor are runs for movies with multiple categories hard to find.
A warb degombs the brangy. Your gitch zanks and leils the warb.
Skilled player (1653)
Joined: 11/15/2004
Posts: 2202
Location: Killjoy
PotatoHandle wrote:
What's wrong with SDA's method of differentiating these kinds of runs? The term "(With large skip glitches)" is appended to their speedrun titles with, well, with large skip glitches.
That is sort of the same as "glitched", just slightly more verbose. I fail to see how adding 3 words makes it magically better. I forgot who mentioned it, but we do need to add "glitched" to our glossary, to explain it.
Saethori wrote:
Super Mario World would end up with a "fastest glitched" run, a "fastest non-glitched" run, a "100%" run, and a "max% small-only" run. (Names tentative, of course.)
Glitched covers the first. The second is any%, etc, etc. Not to mention, any% isn't non-glitched, it is just missing a glitch which fundamentally alters the flow of the game.
Sage advice from a friend of Jim: So put your tinfoil hat back in the closet, open your eyes to the truth, and realize that the government is in fact causing austismal cancer with it's 9/11 fluoride vaccinations of your water supply.
Joined: 6/26/2011
Posts: 167
DarkKobold wrote:
I forgot who mentioned it, but we do need to add "glitched" to our glossary, to explain it.
Guilty as charged. >_>;
DarkKobold wrote:
Glitched covers the first. The second is any%, etc, etc. Not to mention, any% isn't non-glitched, it is just missing a glitch which fundamentally alters the flow of the game.
I wasn't actually suggesting category renaming. I was attempting to meet a compromise between mentioning all the categories, and my mind outright blanking. And by "non-glitched", I was referring to the stipulation you yourself brought up to me, about how "glitched" meant "significantly warped/skipped through the use of glitches", as opposed to "a run with one or more glitches of any degree". I was just simply suggesting a compromise that would give the pair of runs that each complete the game as quickly as possible, but with one using a game-breaking glitch and the other avoiding it, a sort of equal footing. Otherwise, we'll always kind of be stuck with people arguing which one "deserves" to have the base, non-category-marked, version, or questions on what "deserves" to obsolete what.
First a movie gets submitted, and ends up accepted despite breaking rules other runs have been rejected for. And when I vote less than spectacularly on this movie, I become the victim of harassment and threats. Yay, favoritism.
Joined: 7/30/2010
Posts: 16
DarkKobold wrote:
PotatoHandle wrote:
What's wrong with SDA's method of differentiating these kinds of runs? The term "(With large skip glitches)" is appended to their speedrun titles with, well, with large skip glitches.
That is sort of the same as "glitched", just slightly more verbose. I fail to see how adding 3 words makes it magically better. I forgot who mentioned it, but we do need to add "glitched" to our glossary, to explain it.
Wasn't expecting a reply from my favourite TASer, big fan, haha. There was a lot of discussion over the use of "glitched" being too general, as nearly every TAS uses a glitch somewhere, which is why I thought a slightly more specific version might've made things a bit more clear; however, a glossary entry defining the term in a TAS context could work just as well.
Demon_Lord
He/Him
Joined: 2/20/2011
Posts: 80
Location: Chicoutimi, Qc, Canada
I see a difference between data and code corruption. Manipulating the program stack or instruction pointer is a different beast than abusing an algorithm to change standard variables' values.
Active player (264)
Joined: 4/15/2010
Posts: 198
Location: England
Here is our new smv, I don't think I can replace it myself. We're sorry it took so long to update. This contains an 8 frame improvement in yi1. Thank you to Dawn for his insight. However, due to level transition lag, we lose 2 of these frames so this is in fact a 6 frame improvement. Also, we had to redo the Bowser Battle due to a desync, which I think is now more entertaining than before. ^^
Retired smw-96, smw any%
Player (173)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
Nico encode of the updated movie: http://www.nicovideo.jp/watch/sm16636523 As to the update, the precise changes of time are: 8 frames gained in yi1 1 frame lost at ds1 exit (level-transition lag) 1 frame lost at dsh exit (same) 1 frame lost at sw1 exit (same) 1 frame gained at sw4 entry (same)
Retired because of that deletion event. Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"