I still think that 100% is not definable for this game. Go for a best ending run, but get all the notes, cheatos, and honeycomb pieces. Nothing else in the game will trigger a unique Bottles/Grunty/Mumbo dialog. (except Mumbo randomly changing you into a washing machine).
Collecting too much kill routes, takes too long, and is NOT integral to the game. If I had to define 100%, that would be it.
Weren't there other things you could take pictures of to get that heart piece, though?
Maybe. My memory is fuzzy. A big-octo picture gets you something good, I'm not sure if it's a heart-piece, or just rupees... Anyways, tingle is everywhere, so its probably faster to snap a photo of him than do some other stuff.
At least he's not needed to beat the game in this one... *grumbles at him taking forever to decipher charts in Windwaker*
Ohhh, you are SO wrong.
If you want to beat the game 100%, you need to take a photo of him for a heart-piece.
And in a speedrun/TAS the photo would probably be an upskirt photo... ;)
I see no reason to attempt a 100% run that isn't well defined. Getting CLOSE to 100% (leaving out notes) gets you the best ending.
IMO, that is 100%, as the game has nothing more to offer.
So I'd say, aim for the best ending, as 100% is not really definable w/out a huge debate.
Oh yeah, just for good measure, get all 1000 notes, and cheatos ;).
So RAM is the "status" or output of the whole game, while button presses is the input. So I take it, it's impossible to directly manipulate RAM?
But you can't affect it with your controller input, unless the game is programmed to do so.
So RAM can affect your button presses? That doesn't seem to make sense. It might be able to restrict button presses... Ok, add crazy feedback loops to the whole I/O subsystem...
Technically a movie could be tidied up by emulating the game two frames at time and checking if a particular button of the frame before previous carried any effect to the RAM.
Usually games cache the buttons for 1 or 2 frames, but they don't distribute them everywhere in the RAM. Some games use them in randomness checks, but if they do, it would be found out by observing that the RAM is changed across more than two frames.
So RAM is the "status" or output of the whole game, while button presses is the input. So I take it, it's impossible to directly manipulate RAM?
Or given a certain RAM state, generate all the button presses that creates that state? I'm guessing that the whole I/O aspect is not memoryless (might be causal), probably not linear, and probably not time invariant. Thus all my communications knowledge goes out the window.
Some button presses that look out of place are actually manipulating luck. So if you said "hey why is this guy pressing B and holding left while this area loads" and you hex-edit the file to not have those presses, maybe stuff in the level is different and the file desyncs. I think most runners don't have anything pressed when pressing does nothing like during a long unskippable cutscene.
Yeah, desyncs would probably occur a lot. But I'm sure a few button presses can be removed in each movie file (maybe not donkey kong).
I might be slightly out of it due to jet lag, but I'm quite curious...
How hard is it to take any existing movie file and remove all button possible that will still allow the user to movie the game.
I'm quite curious to see if there are a lot of extraneous button presses. Granted a lot of "extraneous" input is due to "acting, extra glitches in scrolling levels..."... Or take this one step further: The movie playback needs to be visually the same. Could be an interesting project, and help decrease server load (reduce movie file size ;))
Another related idea: it would be interesting to compare the least number of button presses to complete a game (regardless of time).
FYI:
Actually an old topic that randomly popped into my head prompted this question. It was related to the SMB walkathon run, I think, where Bisqwit trims the final movie file to exclude every button press of "b".
Sorry for the convoluted post. I'm tired as hell. And I can't think of a proper title (I've gone through 2 edits).
Other possible improvements:
More L/R/U/D glitches, I'm quite sure there are more to be found! You shall be able to use those kind of glitches and slash enemies from distance.
Run more! There are some places I've noticed where you could have started running earlier. Like when your'e opening the 4 keylocks in dungeon 3. You can start running before you unluck it and get the boost right after.
More S&Q's can be used to save time.
More text skips, like when taking Bow-wow.
Damage boosting.
Talk closer to the dog who gives you the riboon, can be manipulated I think:).
Try to have PoP more often. It might save time to get a PoP sometimes instead of S&Q warping.
Another Bomb/Arrow route. You have to go away a few rooms to get a refill. I think I can improve my route.
Jump when a door closes/opens to avoid standing still.
Make bow-wow eat faster.
Take the 50-rupee chest before entering dungeon 2 instead of after.
Light the left torch instead of the right first in beginning of dungeon 2.
Push left block before the right one in dungeon 2.
Go Up-Right instead of Right-Up after killing miniboss in dungeon 2.
I will write more later...
Just curious, how much time do you think all this would save? Obviously not enough for the run not to be published, but I'm just curious.
I've been watching the WIPS as of recent (only recently was I able to get the right version of VBA), so I couldn't really offer any constructive criticism.
Great run. The quality is consistent all throughout. However, I was having some graphical glitches with Link in some areas (esp with bow-wow). I take it that it's something you wanted to put in the game?
And I'm not too big of a fan of save-warping, cause I didn't know where the hell you were going ;). As I haven't played LA in a few years, it took me a while to gain my bearings, but after that, this run seemed like butter on bread!
All in all: I liked your first WIP (that I viewed) a lot. As soon as I re-familiarized myself with the game, I LOVED your run.
A yes vote. No questions asked.
Try "The Alchemist". It's a quick but VERY satisfying read. It leaves you with a great feeling.
I also love the Silmarllion (by J.R.R Tolkien), but as some people say, "pick your poison." In fact, a Silmarllion movie would be 100 times more epic than LOTR. There is a freaking BALROG army!
Sounds perfectly fine for me. It would certainly be useful for optimising paths.
In theory, it sounds like it should work, but we know that theory goes out the window in real life.
Any predictions on what a "short time" would be. With Bisqbot, I think it was more like a few 10s of frames took overnight to compute, and even Bisqbot was no means an exhaustive search.
Don't stress yourself out :). If it causes too much trouble, forget it, and just make a note of it in the comments when you submit the movie.
But then again, looking at your runs, Sami, you tend to strive for perfection :).
but I'm pretty sure it can be beaten under.. what, 20 minutes? 30 tops, for sure. So take 20 minutes and multiply that by 3 since you need to unlock all the characters, and that's one full hour of SSB TAS running.
Can I have some of the shit you are smoking? ;)
It would take quite a bit less than 10 minutes to beat SSB. As a benchmark, consider the adventure speedrun of SSBM takes less than 8 minutes:
http://speeddemosarchive.com/ssbm.html#adv
So beating the game 3 times could be feasible.
Nice, Nice! Spiral mountain's route will be complete soon!
So in terms of route planning, it seems that the overworld may have some strange movement because of the locations of puzzles, entrances, and of course witch switch jiggies. Then we can factor in the warp cauldrons... It seems quite daunting, much more than the Mario 64 overworld route. Maybe even more than an optimal Majora's Mask route.
Anyways, good luck!
I have decided my goal is to the Special cup on 2-player GP, 150cc and beat it as fast as possible. This could, potentially be a published movie as it defeats the game as fast as possible (I would see the ending).
Looking forward to seeing WIPs of this run!
Is 150cc open immediately? Or do you need to beat 100cc first? If so, you'll probably need to changes your plans, and see if you can start a movie from SRAM...
Which characters are you using? Bowser/DK? If you do take a lot of shortcuts that make Lakitu tow you, it may be advantagous to use a character that accelerates faster, but that's up to your discretion.
Good luck! :)
The form factor means nothing It's just cosmetic. Hit detection stays where it is as if you were normal (I think. 99% sure.) You can get the washing machine once you've gotten the jiggy in the sandcastle in TTC.
WISHWASHYBANJO...
I really remember reading about someone who got stuck in a small area as a washing machine, an area you couldn't normally get into, so hit detection may be different.
I mean, hit detection is different for each regular transformation, right? As a pumpkin, you are very small, and as a walrus, you are relatively large, and the collision detection is different.