This text is not supposed to be read, because there is this video which should be watched instead.
Comments
I decided to make a 1.43 instead of a 1.41 or 1.172/2.5 exit version because I actually started this TAS one year ago(?)!
Why this should be accepted
I'm totally sure this will be accepted because it has a valid goal and very optimized strategies. (As you can('t) see) I'm using tricks that aren't even in the Game Resource page of SMW (eg. not 49 hopping or 6/5ing but just pressing the direction is about 1.43 times faster). I should also mention that my input is very entertaining in the loading scenes, which is why I included it in the video. I unfortunately didn't end input as fast as possible, because then the rest of the run wouldn't sync (this game is really hating hex edits), but this shouldn't be a reason to reject this run (that wouldn't make sense at all then).
Possible improvements
Since this is a glitchfest, there are always improvements:
Pressing random buttons at loading screens can make them load faster (that is because the controller else has to be latched multiple times in a frame, which slows down the loading process).
Ending input faster without desyncing the run.
Avoiding too many glitches. That is just decreasing the entertainment value and might affect the speed of this run.
Thanks to
the people that helped me creating this run.
<- these lists which are super neat and handy to use.
Nach: This was a nice run, and is the kind of thing we're looking for at TASVideos. However, TASVideos only accepts runs which complete the game, and this sadly does not. Also, Snes9x 1.4x is no longer accepted. Please however don't get discouraged, and try making a nice full run on a newer emulator, and I'm sure you can get published!
Oh man, what an impressive playaround! I enjoy watching it a lot, and really hope to see a serious playaround submitted here again someday. :)
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Gentlemen and ladies, I would like to submit this thing which I haven't seen exhibited on the site, and which most definitely is a game glitch.
http://www.youtube.com/watch?v=bpoV5imBbZw
The stage starts at 3:38, and the payoff is at 5:17. I don't know the precise process here, but there's something with the bubbles and the Yoshi falling in that pit. There's at least one other example of this trick being done by these guys on this user's channel, so I suppose it's not very difficult, but the effect is distinctive and amusing. It's not really my place to make a request, but if this glitch were explored and understood and exhibited in later runs, I would just be super jazzed.
On for this run, I dig it a lot. I've seen a lot of tricks and glitches in SMW since I started watching videos here, and there were still a good number of jawdrops to be had. I applaud!
I am not? :(
/Amaraticando: My main focus is on the part of reading the controller's buttons multiple times per frame. That just blew my mind. Well, somebody's trollin'...
There's usually no particular reason why a game can't read the controller buttons multiple times in a frame; they just typically read the buttons once per frame so that their game logic and rendering logic stay in sync. For instance, in the NetHack TAS I'm working on, during luck manipulation we press buttons 12 and a half times per frame on average, because that's the maximum speed the game can read, in order to manipulate luck as fast as possible.
EDIT: A great example of this is Super Mario 64. It only reads the controller when it actually cares about the input, and not when it's lagging, so it ends up reading less than once per frame. The N64 console verifications of Super Mario 64 rely on this to sync properly; although Mupen isn't very accurate, they know how much lag Mupen thinks is in the game, and can measure the lag the game is actually experiencing because you can actually electronically detect that the controller is being read, and so they can adjust for lag on the fly. Many other N64 games, like Ocarina of Time, can't currently be console-verified because they read the controller every frame, and so there's no way to adjust for lag.
Please please please make a full game run out of this. This game has so many entertaining glitches and to see a video like this an hour long would be amazing. :)
Please please please make a full game run out of this. This game has so many entertaining glitches and to see a video like this an hour long would be amazing. :)
I'd even be satisfied by a glitchfest that beats the game via the any% route.
The shame about SMW freeruns is that it'd have to spread out the glitches over a large area which would make individual levels less entertaining. Thus it'd have to keep the route short but long enough to justify a "glitchfest"-type run. I think the ideal route would be something of involving taking the Vanilla Dome star warp through to special world, then onto Bowser (or completing the run via code execution or crash). A lot of planning would have to be done to ensure the run gets the maximum amount of glitches in the route without repitition. We'd probably want to include things such as arbitrary code execution too, which can be done as far as I know in Yoshi's Island 2 and Yoshi's Island 3. If that glitch is shown off there, the rest of the run may seem dull in comparison.
I'd love to work on such a project with someone (or a group) though. It's a shame my previous glitchfest project failed due to lack of motivation but I think this game is adequate enough for such a run.
Also, neat work here Masterjun, as I said on Skype. It's an interesting little freerun. :)
Okay, please explain what inside joke I'm too ignorant to get with the "1.43" being everywhere here.
Emulator: Snes9x rerecording 1.43 v17
At the rate things were going, my first instinct was to believe that was just hacked in to perpetuate the running gag. But I'll trust you.
The movie was entertaining, but based on the precedent of the SMB3 playaround, I do believe that Bowser should be beaten at the end, or else the ending screen be glitched in a way that's significantly different from the current memory-corrupting speedrun.
this mini glitchfest is entertaining but I think it would have been better a showing glitchfest most worlds as shown in this videos
Link to videoLink to video
this mini glitchfest is entertaining but I think it would have been better a showing glitchfest most worlds as shown in this videos
Link to videoLink to video
Of course these glitchfest videos are much better than this one. The problem however is that, first of all, they are not very easy to plan. Then there is the problem that you can't put too many glitches into one level, because then you would have no for later levels (or they would fit better in previous levels).
Here I had no problems with that and I could just do whatever I wanted to do :P.
But I'm sure that with enough planning, such a run could be done and you can really do awesome stuff with the physics of this game.
Warning: Might glitch to creditsI will finish this ACE soon as possible
(or will I?)
Of course these glitchfest videos are much better than this one. The problem however is that, first of all, they are not very easy to plan. Then there is the problem that you can't put too many glitches into one level, because then you would have no for later levels (or they would fit better in previous levels).
Here I had no problems with that and I could just do whatever I wanted to do :P.
But I'm sure that with enough planning, such a run could be done and you can really do awesome stuff with the physics of this game.
While you're right about that but what should I do first is to create a route that is the glitches and things that are going to perform at each level but for a game like this is difficult but it is possible but realizing it between 3 or more tas glitchfest one would be easier to perform at least I think so because you and made 2 TASes playaround glitchfest.
