Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
3 years later I have a new WIP3 for this game. I am now happy with it up to 03:01.50 which is right around an extra 1.25 minutes or about 5% further. Most of the delay from my last WIP to now is not having time to do TAS at all in the past few years, but also because every time I have started up this game I have rage quit because optimizing around the lag in this game is absolutely awful. NES was not designed for 4-player.
At the end of the last completed level exit I am 00:07.73 seconds faster than chupapi's published run. That comes out to be an improvement of 4.42% which has held pretty steady so I'm still shooting for an eventual 47 second improvement if I ever am able to finish this project.
Here is a complete breakdown of improvements so far. I was going to wait until I defeat the next boss before posting but the stupid bats in the last section keep turning around staying on screen much longer than in chupapi's run which adds in a ton of lag and I wanted to post in case I give up again for a while in case anyone from 3+ years ago is still curious about progress. Even copy-pasting in chupapi's input results the bats turning around and the excessive lag I am seeing. It appears to be based on the position of your character but not completely or else chupapi's input would work. I'm hoping to keep working on this but this part is really frustrating.
If anyone has any feedback, sees any issues with the run, has knowledge about how these bats work, or any other comments on the run feel free to let me know. Hopefully I have another update before 2019.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Thought about getting back into TASing and gave this one a shot. It is as boring as grassini implied. So far I have only done part of the first level (up to about 31 seconds in his stream) but up to that point I have managed to save ~6.2 seconds.
145 frames were saved by going through the intro screens faster by hitting start.
Another 227 frames were in the actual first level. These frames were saved in 2 ways. First I found that if you hold A down your max walking speed is almost double what your speed is without it. Holding A you follow a pattern of 111 subpixels/frame, 112 subpixels/frame and 113 subpixels/frame. Without holding A the max speed you achieve is 63 subpixels/frame, 64 subpixels/frame and 65 subpixels per frame. What is stupid about this system is that hitting A causes you to jump so the only way to sprint is to jump and then hold A after you have hit your maximum height.
The other speed improvement was taking advantage of the pause damage glitch used in grassini's run. The glitch is that when you receive any damage your lose all of your horizontal speed and go through an animation that freezes your character for a few frames causing you have to accelerate back up to normal speed. With the glitch you can pause and unpause right as you are damaged to prevent the damage animation and not lose any horizontal momentum. Sadly this glitch does not work with fall damage as seen in the previous run, but if you line up your jump so that when you land you take damage from an enemy and not from the fall you can avoid the freeze in many other situations.
If anyone has any input on the run let me know. I can't guarantee I will actually finish this one but I'm giving it a shot.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Mitjitsu wrote:
That jumping noise is really irritating. Is that necessary to go faster like in Super Mario World?
From my tests of this game like 2 years ago jumping makes no difference, you move the same speed regardless of if you are jumping or not.
On the run, I will say that it is a very boring, run right for justice run and that is why I stopped my attempt after barely doing research.
Unless I'm just dumb I can't find where you uploaded an actual WIP file so I can't compare the frames exactly but from the youtube video it appears that you have to stop on the toucan to clear the wall in level 1. If you time the ball throw properly you shouldn't have to halt your momentum. Like I said it is just the youtube video and not the actual file so I can't directly compare if it is slower or just a glitch of the video but something to keep in mind since there will be essentially 0 entertainment value so you will have to make sure it is completely technically sound to get it vaulted.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Asto Robo Sasa is an NES game released only in Japan (previously on the MSX) where the goal of the game is to make it through 16 stages and win the game. You accomplish this task by either shooting out or picking up with your character a group of energy pods spread out through the level. You can find the Game Center CX take on the game (where I discovered it from) here.
What is unique about this game is your characters' movement is controlled by firing your gun in the opposite direction and allowing the equal and opposite force to push you in the other direction. This leads to a game with difficult controls and makes decisions on when to shoot enemies and boundaries more difficult because it can affect your velocity and position.
The fastest TAS version I could find online was a single player run that ended in 11 minutes 31 seconds as seen here.
I am attempting a 2 player run and, while I am only on the 5th level I am already 7 seconds ahead with the major advantage of 2 players yet to come in the much more complex levels.
There may be improvements in what I have done so far but I am just trying to get a good first attempt generated. If anyone else has feedback or someone better at TASing than me wants to give this game an attempt I think it could end up being an interesting relatively quick run. As I said I am going to try to continue with this run and get it to a publishable state but work and family has severely cut into my previously plentiful TAS time so if no one else takes up the gauntlet it could be a bit of time before it is done.
RAM values I have found so far:
Player 1 X-Position: 0202
Player 1 Y-Position: 0201
Player 2 X-Position: 0209
Player 2 Y-Position: 020A
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
I agree with AnS, or at least how I interpret the meaning behind his posts.
All other movies on this site either can run by loading in the game rom and then the controller input directly, or requires extra config before you start entering the input and is thus marked as starting from an unaccounted for activity.
Just as movies that start from a special ram state can be considered legitimate but must be listed with the statement that abnormal initial conditions must be met for the movie to work I feel like this movie should list a similar statement and be judged under the same scrutiny and standards.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
I know I will be in the minority but I voted no. I really do not see the need for another for entertainment-based Mortal Kombat run on the site.
I am not really a fan of fighting games, pretty much only played Super Smash Bros. series out of enjoying the Nintendo characters, but I just don't see what this adds over any of the other runs on the site.
From my untrained eye while I know the fights are different that just turned into watching the same thing 24 times in a row.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
xxNKxx wrote:
they blew blew me :(
I know it is not good to laugh when it is ESL but still...
That's why I think if have good awards then pepole will want come to TAS more than. Like Academy/Oscar awards :D
About copyright I don't have ideas. But with awards, just give to good records, example year awards.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Even with these improvements this run can still be improved through simple menu navigation improvements.
You get to the first level on frame 830, Yunshui started on 779.
You finish the roller coaster level on frame 9656, Yunshui finishes on 9517.
You start the bumper car level on 10025, Yunshui starts the level on 9873.
I didn't go through the whole game but just up until the third mini-game you could save at least 160 frames. That is almost 3 seconds and there is more menus and things following. I know you do beat his time substantially but these simple things add up and losing 3 seconds to menu and text halfway through the run leads me to think there are other parts where it was not fully optimized.
Additionally there is another strategy to beating the game that (if my calculations are correct) could save an additional 37 seconds seconds off of your fastest time.
The strategy is that you can either beat all four levels or get enough points to acquire 50 tickets to face the main boss. To get enough points to get 50 tickets is the equivalent of beating the bumper cars 8.5 (round up to 9) times as it averages around 4.8 points per frame which greatly beats any of the other events. Beating the bumper cars 9 times and then trading the points in for tickets in total would take up up to roughly 35,500 frames. At this same point of entering the boss stage you are at 37,720 so even if the bumper cars took an extra second a piece than I am estimating it would still be over 30 seconds faster.
Now there are some drawbacks to this approach. First, as I have discovered in starting my own run, the bumper cars are by far the most random and annoying part of this game to beat. Second, it would require you to beat the same level 9 times which is not very entertaining.
My thoughts on those two drawbacks are first: TASes are supposed to be hard and frustrating, otherwise we would have all of the game already beat with ideal time. Second: as this is almost certainly going to be going into the vault entertainment shouldn't matter.
If you want you can give this another shot optimizing the menus and trying this strategy, otherwise whenever I finish this (assuming I do since I have had so much less time lately) it will be obsolete.
EDIT And I have the math if anyone wants to look into it themselves. The post was just getting long as it is.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
I don't know if this would actually improve anything but I think having a prompt appear with text about "Your First TAS Submission" to appear on screen when a user is submitting their first TAS to the site would be a nice addition to the site.
I know there is a box that says "Please read the Submission Instructions before submitting a movie" but this text is both not extremely noticeable on screen and the link is more about general submission practices rather then a simple list that new users should see.
There has been a large number of first time submitters sending in movies that they admit to being at WIP level or are simply an enhanced regular play that doesn't fit the format of the site. I would think if a small prompt appeared on the first submission by a user explaining that submissions are not the place for WIP and to make sure your run beats all known records before submitting it would help new users realize that the submission page is not the proper venue for the work and would direct them to the WIP thread where they could get the proper feedback for their run.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
This run syncs on FCEUX 2.2.1.
I am not sure why it would be made in fceu but if it syncs I guess it works.
It doesn't look optimal but it is a clunky game so it very well could just be that but hopefully someone with more experience at the game could look into it.
mklip2001 had mentioned that the published run was improvable so I don't know if this was as big of an improvement as he thought or not.
Regardless it does improve on the run so I vote yes for vault but no for moon.
I am actually surprised that the current run has a moon because in my opinion it is a pretty boring game.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
hegyak wrote:
Prion wrote:
What are the technical limitations preventing a run that syncs on hardware?
Patashu wrote:
There will never be a console accurate emulator (and thus never a console-verified TAS) for pokemon rby/gsc due to how insanely hardware-implementation-defined the RNG in those games is.
Could there eventually be a console-verified TAS for the Pokémon RBY games if you instead synced the run in GB Tower of Pokémon Stadium? I know for the Super Game Boy for the the SNES and for the Game Boy Player for the NGC the actual hardware of a Game Boy is contained so it would encounter the same issues as trying to sync on a real Game Boy. But, as far as I could find, the GB Tower in Pokémon Stadium is actually just an emulator that is contained on official hardware.
So, while I don't recommend making the run on an N64 emulator to play a GB emulator, could the run be re-synced on those emulators to allow for a console verified run on official Nintendo Hardware?
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Whew. I thought my, by far, least entertaining run was about to be obsolete. But as others have said this is not optimal and thus will have to be rejected.
On a sidenote I seriously doubt there is an improvement for my run. This game is extremely simplistic.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Voted no. Doesn't add anything to the any% run by Marx in my opinion and we already have two pacifist Contra runs published so it also lacks being a new concept which would make it interesting from that perspective.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
creaothceann wrote:
MESHUGGAH wrote:
Sure, there are a lot of emulators that has big flaws (FCEUX, VBA), but this doesn't mean that BizHawk or lsnes are the holy grail.
No, they are. The point of using the bsnes core, which is what is used in these emulators, is that it is so accurate that further accuracy improvements will have no impact on commercially released games. The timing has been 1. figured out and 2. emulated correctly. Once a TAS has been created for these emulators, it is extremely unlikely that it will need to be modified for playback on newer versions.
Just a side note. Bizhawk isn't even the most accurate NES emulator let alone a holy grail that will never be improved. FCEUX is far less accurate than Bizhawk but to say Bizhawk is as accurate on the NES as it is on the SNES is not true. Not to mention there are many FCEUX movies that have been verified themselves.
On the topic I disagree with the co-authorship but am fine with some sort of mention. I would be ticked if I spent a lot of time on a run and someone else got half the credit because they realized adding 2 frames between a level would make it sync on a newer emulator.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Masterjun wrote:
yeah so?
From the last submission
AnS wrote:
Input resyncing is not TASing. Ideally it should be done automatically, with minimum human involvement. In fact, in TAS Editor 1.0 you can already do it semi-automatically, so once it's ported to Bizhawk (TAStudio) you can expect an avalanche of dummy submissions, which take only a couple of hours to produce.
This submission definitely can not be "accepted as an improvement", because it fails to beat the old record. It only makes the old record sync with another emulator, similar to how the author of NESBot had to pad FM2 movies with blank frames in order to sync with the real NES.
However, such kind of submissions can be accepted as "movie conversions" (so that old authors remain, and the new author is only mentioned in the description).
Well, seeing as here the new author didn't simply copy the old Input but recreated the Input while trying to improve it (thus wasting more than ~2 hours) I wouldn't object against adding him to the end of the actual authors list as a consolation.
Just, when writing the judge decision, make it clear that you are not encouraging submissions that fail to improve.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
An improvement is an improvement so I don't see why this wouldn't be accepted. But I really wish that when people make small improvements to their own submissions they would write a full submission text instead of just reference the old comments. It is annoying when later on you want to read about how a run was made and you have to jump back through the previous few groups of submission comments to get a grasp. Even if it is just a copy-paste of the old text with the same note you have currently put first it helps.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
Sappharad wrote:
The source code to Project 64 1.7 is not available.
With the way that previous emulators (Gambatte, BSNES) have been integrated into BizHawk, it would not be possible to integrate that version of PJ64 without the source code.
Since Project64 recently released version 2.0 and actually went open source unlike the previous rumors (it is surprising how terrible a forum post announcing that a piece of software is now open source is)I was just wondering if there was any renewed interest in integrating it into Bizhawk? I know people have been hoping for it for a while so if it is as much better than mupen64 as it sounds that could mean good things for the state of N64 TASing.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
FractalFusion wrote:
Hm. While the concept of throwing 30720 punches to beat a game seems interesting, it seems like a cheat code of some kind. I don't think it should be accepted.
SmashManiac wrote:
I'd be more interested in a full warpless/glitchless category, even if it means waiting one hour on a blank screen.
I think the blank screen waiting can be reduced to 5 minutes, if I remember correctly.
I still am not sure if it is a cheat code or a legitimate way to beat the game but I am erring like you on the side of cheat code because it doesn't make sense. At the same time I am not sure what distinguishes something from a cheat code from a weird set of gameplay rules. I am still looking for a copy of the game manual to find out if it was ever mentioned or if it was just hinted at in interviews.
The blank screen can be reduced to only 5 minutes. I am looking into the route. It will probably be shorter than this run but only just. If this strategy does turn out to be legitimate there might be a chance it would be the shorter glitchless% time.
OmegaWatcher wrote:
Now that I read the submission, I'm speechless. Someone actually programmed this game to warp you to the last level if you punch the air 30,720 times. this is insane.
is it your discovery, goofydylan8?
It was not my discovery, no, but I might be the first to actually do it or prove it exists (I am not sure). I can't find a video of anyone accomplishing it but there are mentions of it online.
It is weird most of the mentions online say you have to punch 20,000 times, which is over 10,000 short of the actual number, and the vast majority of websites that mention it say that it is a hoax or rumor to create buzz about a game to make it a cult legend. As I said I am still trying to find a game manual (which I won't be able to read since I can't read Japanese) to see if they make any mention of this strategy. I am thinking they might because A) how would a rumor like this get created otherwise? B) from what I have seen it is a quite unconventional game and was marketed as such.
Experienced Forum User, Published Author, Experienced player
(508)
Joined: 11/2/2010
Posts: 359
On the one hand I was mentioned in the submission text and that always equals a yes vote in my mind. On the other hand you forgot the 8 in my username and that is something that cannot be forgiven.
No vote.Actually I thought it was a pretty interesting concept of a hack but didn't differ enough from a traditional Mega Man run to warrant a publication. A trend of hacks that allows flying in other games would be interesting and I would like to see what other new strategies could be made from it.
