Hey, I'd like to keep this short because I wanna go to sleep...
Heavy glitch abuse
Uses death to save time
Genre: Platform
Emulator: VBA 23, should sync on VBA21. Does not sync on VBA24m.
This run improves the published run by 1145 frames (~19.052 seconds), mostly by finishing the first level quicker, by taking a shell and performing pipe glitch death (entering a pipe and dying in the process, then re-entering the level and going through the ground - this only works on the v1.0 ROM). This trick was already used in the previous run in Macro Zone 1 in order to acquire pipe glitched state so we can enter through the ground in Pumpkin Zone 1 in order to finish that level early. When I originally tested all the levels for their ability to be completed quick using the pipe glitch, I messed up and deemed the first level "unpipe-glitchable", but I recently found out you can move left after entering through the ground and thus reach the end of the level.
I didn't want to keep this insane 2 minute run outdated like this, so I took a look at it since yesterday and created this run. I was always afraid of using the new pixel trick because I couldn't quite understand it (and I didn't really bother understanding it since I didn't think I'd TAS the game again since andymac's around and doing a good job with the game himself). Maybe I messed up in a few spots during the run, I can't tell for sure. I tried my best, and it seems I even managed to save 4 frames outside of the first level, probably by better maneuvering in the OoB area. I didn't try to find a quicker way through the second half of the OoB area though, as there are quite many ways to reach 0xA2D5 (the byte that needs to be altered; which we enter the OoB area for) and I didn't feel like testing those.
I'd like to list andymac as a co-author due to his invaluable effort he put into SML2 TASing, without him this run wouldn't be possible the way it is.
For detailed info about the pixel trick and more, visit the previous run's submission text.
Possible improvements
Let the screen scroll up in Macro Zone 1 in order to unspawn sooner, and therefore leaving the level sooner (saves ~3 frames)
First level optimization, I'm expecting a few frames can be saved here...
Nach: Well, I still have no idea what's going on without dissassembly and tons of data in front of me, but it's certainly better than before. That suicide thing is also impressive, making the run much more exciting. Accepting as improvement to existing run.
I didn't get it to sync. I'm sure I can get it to sync with more testing etc. but I'm not in the mood for it, and I'd like to gather more improvements before doing another TAS.
No, it doesn't make sense because it DOESN'T USE THE SAME GLITCH. The pipe glitch is not the glitch. The glitch is the fact that you are able go outside the map area of memory to locations in memory that are not intended, allowing you to edit the game's RAM from within the game. You shouldn't be able to do that. The game should restart the level when it detects that you are far enough outside the intended game area. (Like, say, what Mario 64 does if you go too far.)
And, yes, I know that's a tradeoff they made to get the system to work on the console. So is every other glitch.
You all even admitted in the previous publication that this was what was interesting about the run. If it weren't for those glitched screens, would you have even accepted this? No, because it wouldn't have completed the game.
And I just went back and looked at the first submission in this branch--it flat out says it doesn't use the pipe glitch. So what happened here is that one run that didn't use a glitch got superseded by one that does got superseded by one that uses another glitch.
Joined: 4/17/2010
Posts: 11475
Location: Lake Chargoggagoggmanchauggagoggchaubunagungamaugg
trlky: You might be interested in discussing this topic.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.