I'd say the bigger problem with VVVVVV is that the TAS playthrough merely looks like perfect human play. There aren't really any surprises to be had. The run does obviously plays things very close to death, but there's no risk of actually dying. There's no unexpected TAS-only strats. The game isn't really broken.
I mean, as someone who appreciates a good no-glitch TAS, it's still a great piece of work. But for a GDQ event, I think you need to be striving to showcase runs that really set TASing apart from normal speedruns. I'd say that means that TASes will have to have either at least some degree of game breakage, or they will have to use strategies that are blatantly and obviously inaccessible to humans.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Wow, that's ridiculously close. I agree with your stylistic opinions, but it does seem like doing Red second makes for some slower platforming later. A shame.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Are there any known exploits for 3D consoles that doesn't require a modified save?
Symphony of the Night has ACE, but from what I understand it requires a fairly lengthy setup that's all menuing (compare the first ACE runs on the Gameboy Pokemon games). Also SotN isn't really a 3D game. Though I would dearly love to see a payload that lets you play as NES-era Simon Belmont in SotN. :)
I'm not aware of any other ACE exploits on post-SNES-era consoles.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Sorry about that Spikestuff... I wasn't quite sure what that "blank input" meant. Is it simply a matter of adding some input up until the last frame? Or should I not include the credits and such? Thanks Nécroyeur for fixing that.
You should end input as soon as the game is finished so that it will proceed to the "ending" of the game. There is no need to extend input to cover non-gameplay stuff like credits. The encode that the publisher makes of the run will include these anyway. Adding blank input makes your movie longer, because TASVideos judges movie length based on the length of input. Hence why its time shows as 10:06 in this thread's subject.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Looking great! Keep it up. :)
I'm surprised by how much the game allows you to get shot even on the hardest difficulty...but I guess it was balanced around regenerating health, so maybe that makes sense.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
I believe the rule is that the superior version of the game is preferred. "Superior" would generally be measured in terms of lag frames, loading times, texture quality, etc.
But I suspect that if someone made a new TAS on a non-preferred platform that was significantly faster gameplay-wise (i.e. not just because of platform choice) than an old TAS of the same game on a different platform, there'd be a ton of complaining and arguing, but the TAS would eventually be accepted. That's just my estimation of how the community would react.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Every time this subject gets brought up in our little corner of the internet, it very quickly devolves into a discussion about quantum physics. What does that have to do with the question of "free will," whatever that means? The points of debate appear to come down to, "The universe is completely deterministic," verses, "The universe has randomness to it," but both seem to be two sides of the, "Our actions are beyond our control," coin.
Not presuming to speak for others here, but the problem we usually run into on this front is that we take as a baseline assumption the fact that humans are simply very complex biological machines. Under such an assumption, it seems reasonable to then assume that any decision made by a human would simply be the expected output of that machine given its current inputs (its environment and current "system state"). Both assumptions appear to be consistent with what we've learned so far of biology, physics, psychology, etc. However, they also lead to the inevitable conclusion that humans are deterministic, and a deterministic entity cannot have free will (the ability to decide things for itself) because its decisions are inevitable and predictable. A clock cannot decide when one second ends and the next begins; the transition is an inevitable fallout from the way the clock is put together. So it is with humans, under these assumptions.
Thus, "free will" in such a system would require some kind of self/identity/soul that is independent of the biological self (not constrained by its rules) but nonetheless able to make decisions on its behalf. We have found no evidence of such a thing yet, which suggests that if it exists, it must "reside" in some part of the universe's rules that we haven't yet satisfactorily plumbed. Since its duty would be to free the biological machine from its deterministic nature, it seems reasonable to look to quantum mechanics, which is by its very nature nondeterministic.
Mind, even if you accept that quantum mechanics is real and that the universe is nondeterministic, that doesn't mean that free will exists. It just means that at a fundamental level, things are unpredictable.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Then please explain why you have any reason to expect that physics be deterministic at any level.
Because based on everything I've ever seen and experimented with myself: If all input variables are known and the algorithm is known, then the exact output is always calculable there is no room for "surprising" results.
So in other words, your personal experience is what leads you to extrapolate that the world is deterministic. To rephrase pithily, because classical physics is deterministic, all physics is presumed to be deterministic.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Correct me if I'm wrong, but it sounds like your argument can be boiled down to "because classical physics is deterministic, all physics must be deterministic." Is that accurate?
No, not at all.
Then please explain why you have any reason to expect that physics be deterministic at any level.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
That's the crux of my argument though, the idea of determinism vs. non-determinism specifically depends on knowledge. Unless we are certain we know everything (unlikely), then we cannot know it is non-deterministic. This is also why I said determinism is non-falsifiable.
Derakon wrote:
so by Occam's Razor there's no reason to assume that one exists.
Yet learning from mistakes of history, there's reason to assume one does exist.
Correct me if I'm wrong, but it sounds like your argument can be boiled down to "because classical physics is deterministic, all physics must be deterministic." Is that accurate?
If so, I don't find it be very convincing. There's no particular reason why we should assume that things behave sensibly when you look at them very closely.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
I'm not saying that QM as a whole must be wrong, I'm saying that it comes off as hubris to claim that things are non-deterministic simply because we haven't been able to yet explain random phenomenon.
Things are non-deterministic to the best of our knowledge. That "to the best of our knowledge" disclaimer is attached to literally every statement of fact in scientific inquiries, and is thus assumed to be implicit because otherwise writing things gets really repetitive, really fast.
Why should we not think there is a level underneath though? At some point scientists were saying the smallest element was the molecule. As science progressed, whatever was thought to be the end all later on was proven untrue and there was a level beyond. Just because the current extent of our knowledge that we're researching is QM, we should honestly think that is the last step? Eventually I guess there must be a last step, but unless we have proof that it is, why should we *believe* that? Such logic is probably just repeating the same mistakes of the past.
You are of course free to seek out some underlying level. There's nothing inherent about QM (as far as I'm aware) that means that it must be the most fundamental level. However, we have no evidence for a deeper level, so by Occam's Razor there's no reason to assume that one exists. Until such a level is found, "the universe is nondeterministic" is the best (most accurately predictive) physics model we have.
We can *blindly* trust the current knowledge at saying there is randomness, but I find that hard to swallow.
Well, you're letting your personal biases influence your beliefs. At least you're in good company; Einstein famously had trouble swallowing QM as well.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
It is wishful thinking to assume everything is known, when there is no proof for that.
This is a vacuous statement, is the thing. Your argument devolves into "you can't prove you know everything", which is true (and literally always will be true, because e.g. we can never prove that the universe isn't being manually manipulated by an omnipotent sentient amoeba who merely chooses in each instant to pretend that everything follows physical laws). But it's also useless, because it doesn't tell us anything about the universe, which is the fundamental rule we use to decide whether to trust a proposed theorem.
The fact is that the current understanding of QM is a better predictor of reality than any other theoretical framework we've achieved thus far. Does that mean that it is a perfectly accurate description of reality? That is, as noted above, impossible to declare with absolute certainty. But QM is the best model we have right now. Arguing that it might be incomplete due to [insert handwaving here] (as opposed to due to some legitimate predictive failure) amounts to philosophical wankery; it's not a serious argument.
Put another way, unless you can demonstrate some way to test your proposed hypothesis that things really are deterministic at some level lying "underneath" QM, said hypothesis is pointless.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Thanks for the explanation, Bobo. It's interesting to hear that this is still an open problem, so to speak. I guess bridging theoretical math and "applied math" (physics) is a tricky job. :)
Nach wrote:
On a side point, say the simulation theory was true, and those ruining the simulation just wanted to "mess with" those running tests to confuse the results and hide their influence. In this case, the simulators are the determining factor, yet every test is rigged. To put it differently, only looking for some hidden "variable" as it were is shortsighted.
That starts getting into "does it really matter if functionally the result is the same?" regions. I mean, barring some security exploit that allows us to jailbreak the simulation, someone mucking with the numbers to make our tests nondeterministic is identical (from any perspective we can achieve) to a nonsimulated reality in which things really are nondeterministic.
If we were able to actually determine that we were in a simulation, and that was why our test results seemed nondeterministic, then that would be a different matter. Likewise if we were able to achieve some deeper understanding of (a non-simulated) reality that explained the seeming nondeterminism.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Is it possible that nondeterminism is fallout from some equivalent to Goedel's Incompleteness Theorem as applied to reality?
I don't honestly know if what I just said is complete gibberish or a legitimate question. It does kind of feel like "it is impossible to make a consistent Theory of Everything for physics" is a statement of similar strength as "it is impossible to make a consistent, complete system of axioms".
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Disclaimer: arguing about this kind of thing right before going to bed is liable to make one incoherent. Apologies for badly-stated arguments below.
Put another way, you appear to be asking where a person's identity comes from. What makes you "you"? Barring mysticism (souls, etc.) your identity must derive from your physical self (because there's nothing else it can derive from), though I see no reason to place a separation between your brain and the rest of your body. Anyone who's ever been hungry, tired, excited, or afraid knows that the body and the mind are deeply tied together. A brain in a jar might be capable of cognition, but it would not have the same identity that that brain in its original body would have had.
Personally I don't think there's anything particularly special about consciousness or the act of experiencing things. Even very simple organisms are capable of experiencing trauma and learning to avoid it (or vice versa, experiencing pleasant things and learning to seek them). What's more unusual, as far as we can tell, is the ability to analyze ourselves and wonder why we experience things (i.e. the ability to perform meta-analyses). But while that might make us more complex than most creatures, it doesn't make us fundamentally different.
In brief, my stance is that we're all just collections of atoms and molecules which interact in interesting ways.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
1. I mean, why do people always work so hard on proving that everything in the universe is a slave of physics?
Because people seek understanding. There's a fundamental underlying assumption here, that things can be explained, that there are regular rules that govern how things work. So far that assumption has usually worked out pretty well, though there are holes in our knowledge and our rules get awfully complicated sometimes.
It doesn't hurt that the more refined our descriptions of those rules become, the more capable we become in our ability to manipulate the universe.
Getting back to the existence of free will: I'm not convinced that it matters one way or another. People behave as if they have free will, and that's what matters. Whether it's actually some non-physical (or not bound by our current understanding of physics) entity making a conscious choice, or the inevitable outcome of a complicated biological computer, the outcome is the same.
In other words, from a consequentialist perspective, the question is moot. And on a related note, P-zombies are people, and so are true AIs, even though the former has (by definition) no true consciousness and the latter may be entirely deterministic.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
It sounds like your overhead is most likely in simply running the emulator as fast as possible, which is not something you're likely to be able to improve very much except by getting faster hardware to run it on. Thus, I'd suggest going the simulator route instead. A simulator should be able to run thousands of battles in a second with ease, since it only needs to care about what "actually happens" in the fight (i.e. resolving attacks, HP, etc.) as opposed to deciding what to draw, what sound effects to play, and all the other stuff that an emulator has to do even if you've muted the computer and turned the display off.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
I suggest instead for screenshot one of the frames where Chewy is looking at the audience while sitting on that discus thingy.
Nice run! I didn't realize how big the AT-ATs are.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Of course, there is a huge difference between writing something purely in Lua (including heavyweight aspects like rendering, vector calculations, etc.), and writing just the high-level logic in Lua while farming out the time-intensive aspects to libraries compiled into machine code, or even to hardware such as graphics cards.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
