Curious about what kind of limits exist for TASing games.
The only real limitations I can really think of is that the entire system would need to be:
1) Deterministic.
2) Be able to be Emulated.
What kind of systems or games have proven or are expected to be impossible to TAS for whatever reason?
Any game that is multi-threaded (or distributed across multiple processes, or worse, across multiple computers) would require deterministic-ifying much more of reality than TAS tool creators may ever be willing to do.
But we can get worse: Imagine a speedrun that, to emulate, would require emulating the external effects of reality on a processor's hardware. Now stop imagining because this is reality:
Link to video
This will probably remain out of reach of TASing indefinitely because how TF would you emulate that, add a temperature variable to the input file??
There exists a cursed game that has no speedrun in existence.
X-COM: Apocalypse: https://en.wikipedia.org/wiki/X-COM:_Apocalypse
Here is a full playthrough with hardest difficulty: https://www.youtube.com/watch?v=lEiFwKfcn5g&list=PLqyBKiHbM__EBJpNQandsqtwD7tFkwFTM
I do not understand the details of the unspeedrunnability of it, so please explain why no speedrun exists!? XD
Maybe it has to do with DOS/Windows CD-ROM instability which causes random crashing? TAS should be able to avoid crashing?
The game itself is quite good and provides a great challenge for speedrun/TAS. If you understand anything about DOS/WIN games, please investigate and make a TAS of it! :-D
The tools for TASing DOS and Windows games exist, but are a lot harder to work with than TASing e.g. a NES game where you can use all of Bizhawk and TASEditor tools. And yeah, I'd be less certain an arbitrary DOS/Windows game works in current TASing emulators than I would those of other consoles. Could check if you want.
In the beginning of the video below the player says: "Okay, good, it doesn't look like I lost anything with that crash."
https://www.youtube.com/watch?v=bWyvjyOiMAI
Which makes me think the issue is random crashing and data loss.
This game may indeed be "speedrun proof" because nobody can finish a run without runkiller crashes.
But I think it is not "TAS proof" because TAS player can avoid crashes by playing differently. That is, if you can set it up.
I think the game may be nearly TAS proof because even if you can avoid crashes it is a big demotivator to retry random solutions just to avoid arbitrary crash. As for the game itself, it is difficult to find good random seed for procedural city generation because you do not know the long game without testing different setups. The game may not be "absolutely TAS proof" but just very very hard, which is practically the same thing, and the reason why I hope somebody will try to TAS it! :-D
Joined: 8/30/2020
Posts: 112
Location: Sydney, Australia
MMOs. Even if you could emulate the network, for a growing number of them, the server software is lost media.
I contribute to BizHawk as Linux/cross-platform lead, testing and automation lead, and UI designer. This year, I'm experimenting with streaming BizHawk development on Twitch. nope
Links to find me elsewhere and to some of my side projects are on my personal site. I will respond on Discord faster than to PMs on this site.
Hey look buddy, I'm an engineer. That means I solve problems. Not problems like "What is software," because that would fall within the purview of your conundrums of philosophy. I solve practical problems. For instance, how am I gonna stop some high-wattage thread-ripping monster of a CPU dead in its tracks? The answer: use code. And if that don't work? Use more code.
"Round 42" I am leaning more towards tas possible until shown otherwise.
Penn & Teller's Smoke and Mirrors with that Impossible Level is impossible to get a proper ending, but it feels more like a gag ending. There is a whole forced frozen cutscene where the game essentially ends since it is impossible to continue by design. Not sure about this one.
Patashu wrote:
Any game that is multi-threaded (or distributed across multiple processes, or worse, across multiple computers) would require deterministic-ifying much more of reality than TAS tool creators may ever be willing to do.
But we can get worse: Imagine a speedrun that, to emulate, would require emulating the external effects of reality on a processor's hardware. Now stop imagining because this is reality:
Link to video
This will probably remain out of reach of TASing indefinitely because how TF would you emulate that, add a temperature variable to the input file??
Despite being multi threaded or distributed it would be possible if the entire system is emulated. Not possible now for complex system afaik though, but not quite "impossible". A small example of a "distributed" system would be those pokemon tases that use more than one console and communicate together.
If there was a game that required the hot plate technique to run correctly that would be pretty crazy. To replicate those hot plate speedruns might be very difficult, but probably measurable and reproducible somewhat? Considering the runners are using that over and over again to get a consistent result.
YoshiRulz wrote:
MMOs. Even if you could emulate the network, for a growing number of them, the server software is lost media.
This one I agree with the most. If the server side media is lost then it is truly impossible to beat those games. Probably huge numbers of MMOs and mobile games are already lost causes due to this. Hopefully those lost game media turn up in the future and are preserved.
