Posts for Randomno

1 2
8 9
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
distanceRan is a double. It's divided by 40 to give the score shown on screen. I haven't actually done the maths on when this would reach max value of 1.79*10^308, the 17 million years comes from here. I had already checked that the counter doesn't stop at 99999, but I looked more closely.
Language: js

distance = this.getActualDistance(distance); // Score has gone beyond the initial digit count. if (distance > this.maxScore && this.maxScoreUnits == this.config.MAX_DISTANCE_UNITS) { this.maxScoreUnits++; this.maxScore = parseInt(this.maxScore + '9', 10); } else { this.distance = 0; }
MAX_DISTANCE_UNITS is always 5 (the number of digits in the distance meter). So this runs once when you reach 99999, but after that maxScoreUnits is 6, so it never runs again. I don't know why they did this, possibly because at 7+ digits it gets cut off the screen, because the X position of the distance meter isn't recalculated. Therefore once you pass 999999, the score only displays the last 6 digits. I looked at the code history and this behaviour was added in Feb 2015. Before then, passing 99999 just reset the score to 0. In the current version, once your distance passes 4*10^22, JS starts representing the score in scientific notation, which breaks the display somewhat. Once you reach the max of 1.79*10^308 JS represents it as "Infinity" which fully breaks the score display (it's just blank). All this being said, reaching 999999 takes a bit under 12 hours. It would be quite impressive to bot this with RAM values, considering how many layers you have to go through. And the severe lag probably makes it infeasible through recognising shapes on the screen.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Yes, it reaches the max speed of 13
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
The max score in this version takes about 17 million years
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
This game has no submissions but many existing TASs, such as this Platinum medal run by taityo: Link to video Recently there has been an effort to find a perfect break in Billiards. Through a significant amount of reverse engineering and brute force this was found on Jun 1st: Link to video It should be noted that real-time medal speedruns accumulate high scores in the minigames beforehand. This means when you earn each medal during the run, you skip the high score animation and save a few seconds each time. Setting this up for a TAS could be quite time-consuming, particularly with Billiards and Charge.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Been curious about this game for a while, despite not playing it. I'm making this thread now since there's finally a "non-joke" speedrun which retires as Governor. All runs before this and the existing submission retire immediately (though, that submission would be acceptable nowadays). There's a massive gamefaqs guide for the game. The highest rank you can retire at is King's Advisor (95-100 pts). I have no idea what the route would look like.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Platform should be Windows?
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Made a copy of this where games are linked to existing game entries on the site. There are probably still some mismatches though. 421/714. https://docs.google.com/spreadsheets/d/1bFkvpAxpZ8kKOVKNKMXejBp5sth1jhsJX3d4hvCDS0o/edit#gid=164004752
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Not to be confused with Escape the Office. Link to video Is the Ruffle version in your recent submissions correct? Not 2024?
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Please can the rules for non-Arcade MAME systems be clarified. CD-i was accepted despite being listed as Imperfect for both Graphics and Sound. It seems reasonable to allow anything listed as Working, but I don't know how exactly MAME classifies these things.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Mikewillplays wrote:
They don't, it's just the mouse's polling rate on most monitors from what I understand
This is what I think as well, seems like a misunderstanding (or bad wording) by the author.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
That submission/judgment implies that Impossible Quiz demo speedrunners are editing the SWF's framerate, which doesn't look to be the case.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Ah there it is. I wasn't aware that Flash would even process inputs above the SWF's framerate, but I guess it differs for mouse and keyboard. Where does your understanding about arbitrary framerate increases come from, is it a previous judgment?
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Despite the previous post being deleted, I'm still curious why q104 is significantly slower than RTA.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Makes sense, could Delay it then.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
feos wrote:
Does it sync on latest libTAS interim tho?
I made it on the latest interim at the time. It syncs on current latest.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
To save someone else the headache: Mr. Tuff (Europe) (En,Fr,De) (Proto) (1994-07-12).sfc - 58840C2CE4F68E30F3D815E9774A1FD3D0D25BD6 Mr. Tuff (Europe) (En,Fr,De,Es,It) (Proto) (1994-12-05) [b].sfc - 52E7D817807A2488555E68F4273AA859C07EFD87 Mr. Tuff (Europe) (En,Fr,De,Es,It) (Proto) (1994-12-05).sfc - 17EFA7BF116A462ABEDD3EA2CB59DDD1B8F9C327 SHA-1 associated with this movie is the first, 58840C2CE4F68E30F3D815E9774A1FD3D0D25BD6. This is NOT the version used for the final movie. The TAS desyncs immediately on this version, as there is no copyright splash screen. This movie syncs on 52E7D817807A2488555E68F4273AA859C07EFD87. This is a hacked ROM dumped a long time ago. The high score screen has been edited (among other things?). This movie does not sync on 17EFA7BF116A462ABEDD3EA2CB59DDD1B8F9C327, the good dump. It desyncs a few stages in.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
SuperSqank wrote:
I’m cool with looking at it. I was actually interested in TASing the game myself, willing to remake the whole thing. I will see what I can do and I will look into sending you an improved movie when I’m done. Could you provide a BK2 or TASPROJ file for the run?
You may as well use GBAHawk, it's a fork of BizHawk and has the same TAStudio, and a more accurate GBA core.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Entertainment aside, I'm not sure that this constitutes completing the game. After finishing 8-4 you go straight into 1★1. It's an optimised movie though, and the quirks noted in the submission and by Niftski sound like they make it an interesting game to run, even though that doesn't translate over to a TAS. Link to video
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Sticky wrote:
So is this something that can be manually added to the catalog so the site knows where to put it? Obviously this is not a pressing issue, or an issue within the site's code whatsoever, I didn't realize there were catalogues that would automatically populate different aspects of the site. I assure you I am not pressing for this to be fixed, you guys are wrong, etc. I'm more interested in the new site workings. Is there a forum post or wiki that explains these changes when they were brand new? I would like to catch up on the workings. Keep up the good work, folks. I appreciate you.
When a submission is made, the author enters the name of the game, but it's not initially linked to any Game entry on the site. It tends to be linked by the publisher, or it may end up sitting on the todo list if not published. The system/console for the submission is determined by the movie file. When a game entry is made, it's not linked to any system, since the same game can be released for multiple systems. The editor has to add a Game Version, including the system, region, ROM name, and checksum. In this case, the game entry was made and linked to the submission, but no game versions were added. I'm not sure if there's a summary of the new site; it's been completely rewritten but maintains a lot of the old features and compatibility. It's open-source and was announced here, and launched Jan 2022. Welcome back by the way.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
It's this I haven't done any stuff with games pages for a while, but I don't know why this happened. Edit: Oh, it's in every System list. That's why.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
You may want to look into FCEUX's built-in AVI/WAV recorder. Your screen recording drops two thirds of the game's frames.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Analyzing the assembly, writing out the logic and adding comments I assume. I don't think these games were even written in C.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
There will be other people more knowledgeable, but since no one has else responded. There's no simple or automatic way to get a game's source code. Technically it's impossible unless it gets leaked or released, but even taking a compiled executable and turning it into readable code can take a lot of effort. In Mega Man's case, the executable was packed with EXEPACK, so you'd probably want to start with an unpacker. There are pages like ScummVM's reverse engineering howto with further resources.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
Link to video A stupid exercise but I may as well document it. I built a pre-1.0 version that targets .NET Framework 3.5. The only issue is the SlimDX library. This in turn uses the Effects 11 library, which was built with SSE2 instructions. So I replaced the SlimDX.dll with an older version and built it again and it worked. What will it take to run BizHawk on Windows 95? I don't know.
Editor, Experienced Forum User, Published Author, Active player (404)
Joined: 2/11/2018
Posts: 220
I tried porting with AssetRipper, using Deepest Sword again. All I had to do after decompiling and opening in Unity was to install Unity UI package, and it built. It was fairly broken though. Had the idea to port the Windows version the "normal" way (btw, unify is really useful here) and then zombie it together with the recompiled version until it worked. I found that globalgamemanagers determines what rendering engine to use (DirectX, Vulkan, etc.). Replacing this file makes the game launch but with the classic broken-shaders pink screen. Replacing Resources/unity_builtin_extra fixes the graphics. However, text is still broken. This game uses TextMeshPro which the decompiler doesn't like, and I don't know how it works or how to deal with it. If anyone has more experience with Unity or wants to TAS a Unity game, I suggest trying this method and see what you can do. Of course, still need the ported game to work in libTAS. I don't know if there's any ways to make it more compatible when building.
1 2
8 9