Not sure if this is the right place to post this request, but would it be possible to publish this game under the original title "Kou Dai Yao Guai: Bai Jin Ban", or as something like "Pokemon: Platinum Edition (Bootleg)"?
Just saying because it's frustrating to do a search for "Pokemon" looking for main series games, only to encounter bootlegs in the search results whose name is nearly identical to an actual main series game.
Cross-quoted from the glitchless thread to push that this run's branch name should be "warpless", "no glitch warps", or "no void glitch". It's not that "no save/reset" is actively wrong for the movie, it's that it's not the best branch name:
1) In this movie, the no save/reset isn't fundamentally part of the route; avoiding using the void is. Suppose a future improvement to this movie uses the tweaking glitch, avoids the void glitch, and also performs a save/reset for luck manipulation elsewhere like the glitchless submission does. Isn't it still the same category and could obsolete this movie despite the save/reset?
2) In the fastest movie, the save/reset isn't fundamentally part of the route; the void glitch is. Suppose someone found a void path that could reach the Hall of Fame without saving/resetting. It shouldn't obsolete this movie on the grounds of being no save/reset, should it?
3) For any other game, if I saw three categories "any%", "no X", and "glitchless" from fastest to slowest, I'm going to assume that the fastest movie uses some game-breaking maneuver X, the middle movie avoids X, and the glitchless movie avoids more than just X. It would be very weird to hear that (a) X is not a game-breaking maneuver, and (b) the glitchless movie also performs X.
Thoughts?
I think all 3 of glitchless/no tweaking, tweaking but no void glitch, and void glitch deserve to be published. Branch names would be up for debate though. (I don't like "no save/reset" for the middle one because save/reset can be used for non-glitch stuff totally unrelated to the void. "Warpless" or "no glitch warps" would be preferred.)
I'm positive that crits completely ignore Attack drops from at least Gen III onward. There shouldn't be any damage difference between a critical hit at -1 Attack and a critical hit at +0 Attack.
Should there be a note in the publication description saying that unlike previous Total Control movies which more or less abandoned the game's original code and ran entirely their own code, this movie's Total Control setup was designed to piggyback off the game's original code and essentially allow switching at will between it and the injected code, which is much more technically challenging? (NES assembly primitives, etc.)
Of note it looks like only one block lost its collision (the second block of the ceiling). You can see that Mario bumps into the side of the third block as he's falling and loses forward momentum.
I remember the Super Mario 64 any% run that was shown off at the very first AGDQ TAS block. Even though everyone on this site already knew what was coming, the audience still found it very entertaining due to 1) the popularity of the game among real-time speedrunners, and 2) the obviously infeasible physics glitches that took place, like the BLJ on the BitDW elevator to basically zip through the rest of the level.
I've always heard it as long jumping accelerating Mario until his speed hits an upper limit and can't go any higher. But there's no lower limit programmed in, so if Mario decelerates (and then accelerates backward) repeatedly, his speed will keep dropping to arbitrarily low negative numbers (really fast backward). The BLJ setups are simply places where Mario can perform a couple dozen of these within one or two seconds.
The game's RNG generates eight-'digit' hex values between 00000000 and FFFFFFFF, which is what all the letters mean. The description is saying that when Metronome is used, the game first generates two values and throws them away (that's what the 'x' means), followed by generating a third value and only pulling out the 2nd, 3rd, and 4th digits from them (throwing the other five digits away).
The three hex digits (labeled 'a') are then converted into twelve binary digits (bits), and the first three of those twelve are thrown away. This finally yields nine bits (a value from 0 to 511), which is what 'bbbbbbbbb' corresponds to.
If 1 + bbbbbbbbb is an invalid move, the third RNG value is generated again (followed by only keeping the 2nd, 3rd, and 4th digits from that, etc.), until the result is a valid move.
Ah, that makes sense. Thanks!
So it's basically process of elimination, except that you search for values that are changing rather than values that are equal to some known quantity. I actually wasn't aware that such a RAM searching filter existed. Thanks again!
A little off-topic, but how does one go about finding RNG addresses in the first place? I know about breakpoints as well as RAM searching, but don't know how to use these to e.g. find an address that when set to a specific value makes all moves critical hit.
Any in-depth insight would be hugely appreciated!
Holy crap, thanks so much for pulling the subs! I know I'm not the only one on here who enjoys reading long submission texts but doesn't have 3+ hours of free time to watch the whole movie. Reading everything immediately. :)
We need a "worst abuse of the rules" award just like the one for the Obfuscated C Code contest.
Technically, the only thing barring this from publication is that it doesn't break existing records.
I will vote yes if you prove this will finish running before monkeys finish typing Shakespeare's plays.
Almost certainly yes. There are enough of Shakespeare's plays and each play is long enough that the number of possible combinations of letters (whose length equals that of all of Shakespeare's plays put together) is much, much greater than the number of possible combinations of inputs whose length equals the number of frames needed to beat the game.
Time for that yes vote.
Oh crap, that slipped my mind since I usually don't think of Regis as event-only 'mons. In that case one would need to wait until B2/W2 for NDS-constrained Regis to be available.
This has nothing to do with the run, and mostly with the concept, but...
This is completely possible to do for later generations (G3+), just with 3 different consoles (GBA, GCN, NDS). Not actually expecting it to ever come to light, but how would that even be TASed?
By intentional design, you've never needed more than one type of console to complete any Pokedex:
151 and 251: Possible with only GBx
386: Possible with only GBA (need all five of R/S/FR/LG/E)
493: Possible with only NDS (need all five of D/P/Pt/HG/SS)
649: Possible with only NDS (need either all seven of DPPtHGSSBW, or all seven of Pt/HG/SS/B/W/B2/W2)
721: Possible with only 3DS (need all four of X/Y/OR/AS)
Event-only 'mons at the end of the Pokedex excluded, naturally.
The release of Emerald made it possible to get the Gen III diploma without Colosseum or a GCN. The release of HGSS made it possible to get the Gen IV diploma without Pal Park or a GBA (or a DS Lite since the DSi came out around that time). The release of ORAS made it possible to get the Gen VI diploma without Pokemon Bank or a NDS. These are all games released at the end of a generation. None of this is a coincidence.
(Less impactfully, the release of XD made it possible to get the Gen III diploma without R/S, and the release of B2W2 made it possible to get the Gen V diploma without D/P.)