Posts for XIF


XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
i thought this might have been entertaining with the battle animations considering how many crits you get, but I suppose with 94 turns there's no need to expand this run another couple hours.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
anybody wanna drop some helpful suggestions on where to get a good dump of the game for my view pleasure? Seems interesting, especially with the image in my head I get with "handles like a tank with a jet engine" Aside from that... I felt like posting this somewhere... So the past week saw the publication of 24 different runs, where as before we were getting 1 run about every 3 to 5 days. All the runs on queue seemed to be up there for weeks or months, and even saw a mario run staying in queue for over a month I think, where as before over a day was like an eternity for such runs. wtf is going on?
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
PostPosted: 2008-07-12 18:51:50 Post subject: #2029: SilentSlayers & z0MG's N64 Super Mario 64 in 05:33.85 today: 2008-7-25 this game: Standard Mario game. submission to publication time= end of the world.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
hmmmm... Well what pointer variables are there in Mario 64 in the first place? Entrance to the castle + key doors perhaps? I was under the assumption Mario 64 didnt have any real pointer variables to use. My only other awesome idea is black room of death to test out... I dunno...
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
So it seems tricking the doors isnt likely... The only other entrance is the moat then, no? The door is completely inactive while there's water though IIRC... so what that guy was saying about there being a byte to store the moats state, if there's a way to overflow it and cause one piece of data to cause the moat to sink prematurely, then there's a way to simply skip the first bowser. Of course I have no practical experience TASing this game, so this could be a load of hogwash.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
Swordless Link's avatar: an appropriate reaction to the topic at hand. Anyone wanna follow up on my suggestion before going crazy with loopy BLJs?
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
man, people be having forum drama no matter what... LOL at TAS drama. Lucky this isnt the MVC2 scene. MVC2 there's drama and niggas get jumped or start losing thousands of dollars ROFL anyway, as far as the "tricking the game to thinking the first key opens the door to third floor"... surely it wouldnt be difficult to shark it so that you activate a code to automatically set your X, Y, and Z coordinates. (I dunno how to use a memory watcher myself, but I imagine some seasoned TASer (marinated, perhaps) could create the code and just activate it during the open the door scene and see what happens. If it works then you can concentrate on doing it with a BLJ. Maybe you can avoid that all together and just observe when the bit determining lock or unlock changes, cause if it changes the frame you approach the door, it seems unlikely that the trick will work (Or maybe it means you need to get from their to the upstairs in a single frame <_<), if its after the entire animation, maybe its worth investigating.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
So i'm doing a little test run of the rocket belt. 814 frames with 60 rerecords, mainly just going through menu's, but it took me about an hour to find what I thought was the best way to pass the first ring. There's probably still better ways though. The main concern is positioning yourself for the next ring, since momentum is carried over so much, anyway, i pass the first ring on the 814th frame already a fairly interesting screen: EDIT: 1154 FRAMES, second ring passed. about rerecords: the way i've been quitting and saving my movie for later works has been to do movie stop and then do another movie record using that file name and just load the savestate and savestate once again and then continue. I dunno if this is the norm or how your technically supposed to do it, but regardless it resets my rerecord count. I thought it seemed a little low, and now I know why. I should have something like 200-300 or so re records. I know it might seem silly to record such small steps, but this is my first time TASing, so I feel every little step is awesome =P
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
I was thinking it'd be better like that as well, which is why I wanna go for top score, but also some other things that may be slower but cooler. Like doing rocket belt 5 feet above ground at all times and things like that.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
the thing is that basically to get anything out of the bonus stage you need to be bouncing around for a bit for the extra 20 points needed to pass the first course, and that takes quit a bit as well, because you stay in the air for so long, and then you need to hit 4 target, not to mention the first 2 jumps before hand. Whereas the plane pretty much going through the stage as fast as possible means getting a perfect score, and then after that you just dive bomb three rings and break your neck on landing. This is going to be far more complicated than expected @_@
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
anyway, I tested out that glitch you were talking about and this is what I got: There was absolutely no graphical errors or game crashes on level 3 hang glider stage, however the light plane lesson first gave me a weird scramble of graphics (it was more like saving something as a gif in paint and having it appear fuzzy) and then the game crashed before the last ring was passed. I dunno what caused it but the error it gave me was something along the lines of that snes couldnt read the memory correctly. I have a couple of questions. Right now, I want to do a straight run through. Speed and score. But there are bonus stages for when you land in the moving targets that allow you to get more points. Right now I think I'd like to just get 100 on every stage and land on the normal targets, but I could also go for every bonus, or I could also simply go for straight speed. For instance: In the very first level, what would probably be the fastest way to simply pass would be to get a perfect on the light plane (All beams, really fast, perfect landing) and the just hop over to sky diving and start off immediately head first and pass the first 2 rings, head up for a couple frames or so to pass the third, and then just head straight down again and crash. The rings alone give you 20 points and pushes you barely equal to the passing score. While undoubtedly funny (crashing is just funny in pilotwings) its hardly impressive. so what? Normal playthrough with perfect scores or complete playthrough with bonus stages or barebones fastest way possible?
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
sorry about that my laptop has a super high resolution (1920X1200) so to me the picture wasnt incredibly large but I forget that not alot of people have a resolution nearing 2K pixel width. I am unsure about the hanglider glitch, i'll test it out later.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
Where do I know you from again neophos? I recall the name... Its funny though, I've been looking at the boards for so long I feel like I know all the users here, and all the drama too. Like Saturn with Super Metroid, and that stuff in the OoT thread about being arrogant. Whoever says this is correct but most people who watch the movies here respect all movie makers simply for making great TAS's. Most dont give a damn about board drama. But I digress. edit: sorry.
<3 time attacks
Post subject: Pilotwings
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
So i'm a lurker... I mean, a really really old lurker. I've been following I dunno how many TASes from start to finish and reading up on various strategies since like 2004 or earlier. I have a little experience with speed running (used to do break the targets in Smash Melee before I got competitive in vs.) but I've always wanted to try my hands at a TAS. Specifically Pilotwings just cause there isnt a TAS already. (Or am I mistaken?) I know some of the basics and the dos and donts. I'm gonna use SNES9x v 1.43+ v 9, and I know how to use Frame advance and save states. But I wanted to ask some questions. First off, I know for alot of games, people use memory watchers for position, speed, etc. Where do I get it, how do I use it properly, and whats the right way to go about finding the correct memory address for such values? Pilotwings is (for me anyway) notorious for loopy animations trying to make it 3-D, so the ground is always flipping out in disorienting ways, and just knowing the value for the acceleration, and especially the turning would be immensely helpful. What is the right way to go about the actual recording? Do I just create a savestate from the start and test one by one the optimal frames? (For instance, I couldnt tell you the earliest frame I could press start at to bring up the menu automatically at the start screen). Would I just simply set a savestate at the beginning, create another one before each frame I test to see if start then will bring up the menu, and just repeat it until I get the result? Are there any other helpful tools I might want to use outside of that? Is there any menu option in snes9x to perhaps set and stop a frame count (like a timer for frames) and what is there to display the overall frame number of recording? I did a couple of preliminary testing on the skydiving stage in the A Class (first) stage. Like: 42 frames to open the parachute roughly 105 frames if you open it while falling at max speed (pointing straight down) for your rate of descent to drop to a non crashing speed You dont start turning for about 2 to 3 frames on button press and you stop turning for around 3 to 15 afterwards (How long you were turning affects this). for each change in body position for flipping, starting at full horizontal face down, it takes 8 frames, except the first takes 7, and to position yourself full vertical, it takes 63 frames of continuous input of up or down to point straight down (or up if you will). Altitude gets updated once every 4 frames The values generally oscillate between 16 feet and 17 feet at each update of altitude. My tests for face or feet down showed that while the values were equal half the time, feet down showed it as 1 foot closer to the ground than face down. I dont know whether either one will be faster than the other, but I think face down is better, pointing face down at the very beginning allows you pass the first 2 rings without at problem, and simply facing back up for a little to get to the third, whereas feet down threw me back too much, which brings me to: A break down of flipping (minimum frames is the amount of frames for body position to change minus 2) 7 frames to change body position the first time you ever turn 8 frames to change body position to any position not previously at 6 frames the first time you go to the previous position you were just in (i.e forward 8 frames, then back 6 the first time to return to previous position) 5 frames every time after that to go to the previous position. 10 frames if you input for the minimum frames, and then resume flipping. This applies only to the first body position change and it doesnt alter subsequent changes if the button is held, and only if the position is not one you were just in, you may also resmue flipping before the actual change has happened, you can go two positions forward by doing 16 frames up; or 6 frames up, wait 2, and then 10 frames up for 18 total; or 6 frames up, wait 1, and 10 frames up for a total of 17 frames 4 frames if you do minimum frames to change body position, and use the 2 extra frames to go the other direction. It is then 3 frames if on that stacking of flipping, you use minimum frames to then use the extra time to go back once again. After than it is 4,3,4,3,4,3, etc... going forward or backward doesnt affect this, only what the previous position was (in essence, which direction you were going in the first place. now it gets a little tricky, this was continuous input (holding down.) the frames are cumulative, you could push forward for 1 frame, wait a second, and then press forward for the other 7 and do the turn. minimum frames apply here too, like this: up wait wait wait wait up wait wait wait wait up wait wait wait wait up wait wait wait wait up wait wait wait wait up so over 30 frames you press up interspersed for 6, and then the next frames nothing will happen and then the one after that body position will change (8th frame total) If you hold up the extra 2 frames until body position changes, it changes your falling angle from if you were to press the minimum number of frames this is from like 2 or 3 hours of testing, it was confusing at first because at first it seemed kind of random when I was turning when my body position would change but I think this is the entire breakdown. I need to do further testing for turning as well, and the actually parachute is gonna be a nightmare, and even though alot of this info is unnecessary, I think its nice to have an idea of how the game works. EDIT--------------------- I have a couple smv files showcasing the different things that happen under the above rules. I'll try and put them up tommorrow.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
POP IT INTO OVERDRIVE GENERAL! SHE CAN HANDLE MORE YOU JUST GOTTA BELIIIIIEEEEVEEEE!!!! EDIT: uploading faster than your downloading should be punishable by law.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
SEED MORE.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
perfect answer, thanks. I had always wondered but knew there had to be a good explanation so i never bothered but i was reminded just now. It just seems the elec beam is the problem when its really the solution. <_<
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
this may be bringing up old news, but I was proud of myself for thinking it up. if: john doesnt like the use of excessive glitches and: this movie uses excessive glitches then: john doesnt like this movie. (and therefore votes no) even arguing that its a protest not an honest vote, my perfect syllogism owns all arguments BWAHAHAHAHA!!! ... ... hi, i'm a lurker, i dont post much, but i download almost 3/4ths of the videos here. I do have a question about wily1, when you pull out elecman's weapon, it seems to lag caused by the shots you use... whats up with that?
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
when was this finished?!?! i'm sooooo behind it seems. I checked on this like a month or 2 ago and i assumed at 80+ stars you had WAY more time to go but i guess you know no bounds. downloading the avi now, i'm super pumped.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
i creamed a little during webwoods. good stuff, no johns.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
Amazing video with a perfectly fine pacing, and I liked your choice of goals. This video made me go and get the GSF's for the game. =) I actually wanted to put my own input into one of your comments.
nitsuja wrote:
(I must admit, though, that I'm skeptical of the records at TSC that involve following a route that requires nailing 1/60 of a second timing windows well into the level. I wouldn't consider such routes possible outside of a TAS, but maybe my reflexes just suck.)
Its very possible to do actions which require 1/60th of a second timing, and perhaps even harder on consoles. Even possible to do so consistently. My only example as usual is SSBM, but there's a glitch with Samus that makes her fly across the floor really far and fast from her bomb. You have to press one direction on the 42nd frame of the move and the other direction on the 43rd. Many people can do this almost constantly... the fact that its individual level runs compounds the idea of its possibility as well. BTW, Sunset Hill 1 song and stage rocks.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
TSA simply came off as argumentative and a little arrogant, but I dont think anyone reacted the way they should've =/ this is basically what I saw TSA yells TSSAAAAA!!!! at everyone, but in a nice way. Anger Fist thrusts an angry fist through TSA TSA goes TSSSAAAAA!!! but in a mean way >=( GuanoBowl prepares to dump guano on TSA >_> *Other unoriginal play on words* About the cam hack: I thought that the view point was OK overall but it would have been nice to see a larger portion of the world he enters. It would simply affect the visuals of the game though, yes? I because when I think different camera angels I think that forward becomes backwards and vice versa.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
nice work, i dabbled in BtP and BtT a little in 64, my only notable score was Yoshi's BtT which was something like within .4 of the WR.
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
remember kids, rules for anything always end up turning into a "loose guideline" dont get all yer panties all ker-fluffle now. I've enjoyed the idea of cam hacks ever since one was made for Sonic Advance. Doubtful, but any reasonable way to get the cam in this run or future runs to show the front of Link instead, since he's facing backwards most of the time anyway... or would that perhaps screw up the input of the game...
<3 time attacks
XIF
Experienced Forum User
Joined: 2/7/2006
Posts: 58
Yeah, i'm not THAT knowledgeable about 64 smash, just some things I picked up from board whoring and from moogle at smash aid 7.
<3 time attacks