CyberShadow had a bot search for better solutions and mentioned it in this thread. See my old post for a 2-2 tip. It might be exactly that position the screenshot is showing. Then going for the left 3x1 block will be even faster.
Good news. Switching the core to NEShawk didn't desync the movie at all. Continuing with new core. But if NEShawk is more accurate, why is QuickNES the default?
The published run takes the blue route, but it's faster to take the red route. Curiously, the 100% run on speedrun.com also makes this mistake.
I'll probably look over some of the other levels as well, but puzzle games generally aren't my thing, so if anyone else wants to jump in and look for other improvements feel free.
Hold the phone. There's code in Bizhawk to hook up the tilt sensor and the vampire sensor to the mgba core. Is that not working? If so, it's probably a 5 minute fix for me. Did someone report this? I don't see anything on the issue tracker with "tilt" in it.
oops
System.FormatException
.
.
.
in BizHawk.Client.Common.Bk2ControllerAdapter.SetControllersAsMnemoicI'm not sure why writing a new emulator is required to get the peripherals that mGBA already supports working in BizHawk. Wouldn't it make more sense to fix up the mGBA integration and hack on the already fairly mature mGBA?
Good luck with this project. However, I don't see the point since mgba already aims for accuracy. Can you elaborate what you mean by "for things like Yoshi Topsy Turvy"? Because I already started a WIP on the mGBA core which worked out fine (although a luascript is required to make tilting work). Link to the Yoshi Topsy Turvy thread
However, what do you plan to bring more than mGBA for example?
Good luck! BTW, do you have plans to edit the DualGB mode so that it can use GBHawk? Or maybe even make it support more than 2 games at the same time?
I would once again ask: This exact thing works at speedrun.com. Why wouldn't it work here? There, many games have many categories that have technically speaking "arbitrary" rules and limitations, but these have been decided and agreed by the speedrunners themselves. The list of limitations in some categories for some games can be quite extensive (and sometimes not without slight controversies). But nevertheless, it works there, and can result in very popular speedrunning categories (once again Ocarina of Time being the posterboy for this, although it's of course not the only example.)
However, there's still another issue. Basically, the game's audio is out of sync. Characters' mouth movements don't sync with their voices. This also happens in mednafen, no matter the dumped image. I once created an account in their forum to report the bug, here's the link: https://forum.fobby.net/index.php?t=msg&th=1571&start=0& According to the last post, this apparently has to do with CD-ROM seek time not being properly emulated. This could prove useful if you're willing to look into fixing stuff.
Console-verified using Game Boy Interface.
I was playing Holy Diver just now, and I got softlocked after the fifth boss. Normally, after the boss dies, the end of level music begins playing and the floor opens up beneath the guy on the cross and you jump down after him. However, in my situation, the music never starts up; the floor doesn't open, and I'm left trapped in the room. I've retried the fight a few more times, but I always get the same result. I'm using v2.2.2(64x), playing the game with the QuickNes core.
Despite showing up and showing the domain when in action, the code data logger doesn't appear to work with NESHawk. Is this a bug or was it just not fully implemented? Seems like it should be working with NESHawk of all cores. It only shows the "RAM" domain and despite causing the game to lag up (it's clearly doing something) it's not populating that domain.
Ok we can now tilt 'n tumble!
I think I got everything working with the mapper and EEPROM. Motion controls and centering work, saving works, and TASing is possible.
However, I've never had a GBC or played this game, so I don't really know if the controls feel right, or if it gives the proper playing experience. So, if you are interested in this game I would really appreciate some testing and feedback. (I know there are a few graphical bugs, I'll work those out over time, I'm primarily interested in how it controls.)
Some notes on controls:
You control the angle of tilt directly with the controls. This means there are some nonlinear effects. They aren't really that noticable in casual play though, since the angles are not that large.
The control scheme is based around rotating about the up-down tilt access first, then rotating around the left-right tilt second. However, both rotations are around body fixed axes. This means that left-right tilt is coupled to the up-down tilt. The effect of this is that left-right tilt is less effective as up-down tilt increases. This might be an important consideration for a TAS, but so far in casual play I didn't see any serious impact.
The current control scheme assumes you tilt slowly (basically the only accelerations on the system are from gravity.) There are no induced accelerations from fast motions. This is one area where there might be noticable differences between real play and emulator. I can simulate them if the feedback is that it's really needed.