Post subject: Medievil PSX: Soldier Despawn Bizhawk Crash
Player (196)
Joined: 4/27/2012
Posts: 109
Here's a video of the glitch crashing (Not my video) just as an example of what i'm talking about. Link to video The glitch works fine on Psxjin as seen in the previous tas. Not sure if an exception would be made to use jin for a tas of this game, but I can't get this to work on bizhawk 2.4 or 2.4.2.1 and it saves a decent chunk of time. My iso dump has a 99.9% accuracy. It's a shame because even if there is a solution with emulator modifications, which I have tried the tas would have to be redone :/. I did a trace logger and it happens that the game freezes when the ai is set to zero. I also found a ram address that allows the freeze to stop if you freeze the value at(0EDC38), which I assume is the cutscene count down timer.
I wumbo you wumbo he she me wumbo
suqd
He/Him
Player (25)
Joined: 3/19/2020
Posts: 4
Does the softlock happen on real console, or only on Bizhawk?
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2822
Location: Northern California
Judging from the title, the provided video appears to be on console, so it looks like it's being handled correctly on BizHawk. If the glitch only works without crashing on PSXjin, that'd be an emulation error, and it wouldn't be allowed. I imagine the only reason it was allowed then is due to the fact that PSXjin is the best we had at the time, and there wasn't enough research done to know whether or not that particular glitch worked the same way on console. It's unfortunate, but that's the way it goes with accuracy.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
Player (196)
Joined: 4/27/2012
Posts: 109
Samsara wrote:
Judging from the title, the provided video appears to be on console, so it looks like it's being handled correctly on BizHawk. If the glitch only works without crashing on PSXjin, that'd be an emulation error, and it wouldn't be allowed. I imagine the only reason it was allowed then is due to the fact that PSXjin is the best we had at the time, and there wasn't enough research done to know whether or not that particular glitch worked the same way on console. It's unfortunate, but that's the way it goes with accuracy.
I'm pretty sure it can't be done on original ps1. But this begs the question about the standard for future tases of this game. The glitch can be fudged a little bit by spawning 2 soldiers then having your men kill them, then spawn the other two but it's still not as fast as the current run. Now the question is, is the current run still acceptable knowing that it broke one of the rules, or will it be outdated regardless of time loss on the final battle?
I wumbo you wumbo he she me wumbo
Samsara
She/They
Senior Judge, Site Admin, Expert player (2241)
Joined: 11/13/2006
Posts: 2822
Location: Northern California
This kind of thing happens sometimes, where glitches in published runs are discovered to be emulation bugs . I believe Super Mario Land 2 had an issue like this as well? Since we don't unpublish runs for any reason, the only thing to do is to just submit an improvement run that doesn't use the glitch, or in your case I suppose it means using the glitch in a different way that doesn't crash on console. It won't have any impact on how the run is judged, don't worry. It's the same sort of thing as when we get new, more accurate emulators, and we end up getting improvement runs that are actually "slower" than published runs (note that the obsoleted run appears to be over 10 seconds faster, but this is due to the new run including the PS1 BIOS). The more console-accurate run will always be preferred, even if the accuracy makes it slower in the end. Hopefully that clears things up!
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family. Now infrequently posting on Bluesky
warmCabin wrote:
You shouldn't need a degree in computer science to get into this hobby.
Player (196)
Joined: 4/27/2012
Posts: 109
Samsara wrote:
This kind of thing happens sometimes, where glitches in published runs are discovered to be emulation bugs . I believe Super Mario Land 2 had an issue like this as well? Since we don't unpublish runs for any reason, the only thing to do is to just submit an improvement run that doesn't use the glitch, or in your case I suppose it means using the glitch in a different way that doesn't crash on console. It won't have any impact on how the run is judged, don't worry. It's the same sort of thing as when we get new, more accurate emulators, and we end up getting improvement runs that are actually "slower" than published runs (note that the obsoleted run appears to be over 10 seconds faster, but this is due to the new run including the PS1 BIOS). The more console-accurate run will always be preferred, even if the accuracy makes it slower in the end. Hopefully that clears things up!
Yeah it clears things up, the run will be 20 minutes faster regardless which is an upside. Thanks for clearing things up :D.
I wumbo you wumbo he she me wumbo
Site Admin, Skilled player (1255)
Joined: 4/17/2010
Posts: 11495
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
You can still try to get it on mednafen that PS our core originates from. Just to be sure.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.