Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
Let me ccorrect myself here: Nemesis was working on a microcode-accurate 68k core, but due to issues with available free time, decided to leave it for after Exodus is open sourced, and is working on getting 1.1 out of the door first. So those numbers are for the old 68k core.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
Nemesis has stated in the past that 1.1 has a micro-opcode accurate 68k core and correctly emulates bus access timings; I would think that these results already account for that, but it isn't explicit.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
For starters, there is evidence that Nemesis has come around as far as licensing goes. As for the performance, Nemesis has posted this two days ago:
Nemesis wrote:
What consisted this internal change?
Complete restructure of how debug windows are implemented and managed, expansion of the abilities of extension plugins, and better separation between the emulation platform itself and the GUI. This model gives better separation between components and make everything more extensible.
Will the next release be improved in term of speed?
Yes, in a major way! I've heavily profiled and aggressively optimized the major bottlenecks in the emulator, as well as restructuring some critical areas of the bus system to reduce overhead. I've also made some critical improvements to the threading model which give a significant performance boost, the effects of which are magnified on dual-core systems. Here are some comparisons I just took between the 1.0 release version of Exodus and the current Exodus 1.1 development build:
That's an average 40% memory usage decrease, with a 90% performance increase for the Core 2 E7400, a 40% performance increase for the i7 920, and a 30% performance increase for the i7 3840QM. These ratings were taken on Sonic 2 using the first level for comparison. Note that this makes what is a 6 year old processor be able to run Sonic 2 (one of the more demanding Mega Drive games) at 85FPS. It might also be possible for a dual-core system to achieve a full framerate too. The E7400 gets 83% of full framerate. The Core 2 Duo E8600 might be up to the task. At the very least, it's now usable on a dual-core system for development and debugging.
These numbers may change before release. I've got another critical change to the bus system that could introduce more overhead, but it's important in order to be able to make it flexible enough to apply to all systems.
As for "slightly better" emulation, I can't comment much; Exodus will be much more accurate (Nemesis is using several detailed documents which he obtained and hasn't released yet), but for how long remains to be seen (as it is likely that many emulation improvements can simply be ported over to Gens+GX).
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
The centrifugal force is one of the many so called inertial forces: that is, forces that only appear in non-inertial frames (they have been historically misnamed as "fictitious forces" or "pseudo forces"). In one of the formulations of Mach's principle, inertia only exists in relation to other things; if there is nothing you are rotating relative to, you are not rotating at all; there is no rotational inertia, just as there would be no translational inertia. So the rest of the universe must be non-empty so that your inertia exists, and hence so that you can feel an inertial force such as the centrifugal force. Put in another way, all frames are inertial if the universe is empty, so you can't change to a reference frame where you have an inertial force.
Of course, Mach's principle is a philosophical one -- Mach was a philosopher, not a physicist -- so it makes no attempt at explaining how the contents of the universe would go about generating inertia. GR does give a possible mechanism -- the contents of the universe define the geometry of the universe, and geometry defines inertia -- but in such a way that gives GR a non-Machian character, depending on how you formulate the principle.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
Warp wrote:
Do I understand correctly that it's a physical-philosophical question of how do we know that an object is "motionless"? That if we define it as "motionless" by comparing it to the average distribution of mass in the entire universe, this opens up many difficult questions?
It is a bit mote than that: Mach's principle is basicaly the notion that you can't even talk about motiin unless it is in relation to something, hence inertia onky exists in relation to other stuff.
Warp wrote:
For example, how do we know that an isolated object is "rotating" if we don't have anything to compare it to? But this part sounds strange to me because rotation can be measured by measuring the centrifugal force that's applying to different parts of the object.
In a fully Machian theory, the centrifugal forces would only exist because of the distribution of mass in the rest of the universe; if there were nothing else in the universe but the "spinning" object, a Machian theory would say that it feels no centrifugal force. In GR, you can actually have a rotating universe (Gödel's solution), so it fails the Mach test in this front; but:
Warp wrote:
(And in GR it even causes frame dragging, which I also assume is measurable.)
The Lense-Thirring effect is actually expected in a Machian theory: a sufficiently heavy rotating "bucket" would affect the inertia of the other things in the universe.
Warp wrote:
On the other hand, an object orbiting another object (such as a planet) is also "rotating" but experiences no centrifugal forces... (but isn't it actually in inertial motion according to GR?)
It is inertial motion if the orbit is solely due to gravity, yeah.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
So I went doing some refresher on this, and I remembered this book: specifically, pages 75 and 76. If you can't view it, let me know and I will summarize it; I had to login to be able to see the pages in question.
Bottom line is: Mach's principle is very vague; depending on how you state it, GR is Machia or non-Machian, or both in varying degrees. He goes through 8 versions of the principle in the linked parts, but there are many more variants.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
Oh, no argument re: normalization; I was quibbling with the quoted bit.
Now, you may complain (not without merit) that the Schwarzschild solution is not a vacuum solution; technically, it it the same solution you would get in a space-time minus one point. So it can be a full vacuum solution depending on your spacetime.
Leaving technicalities aside, Mach's principle, in a nutshell, relates the inertia to the distribution of masses; you can't have inertia without having a distribution of masses, and this distribution determines inertia for each constituent. Since GR couples inertia and geometry, you can state this as 'no inertia = no geometry'.
Thus, the notion that a single point mass can determine inertia (it determines a full spacerime!) is distinctly non-Machian — there are no other masses relative to which you can use to determine inertia for the singularity, but it is there anyway (as the parameter M).
Moreover, the very fact that you can get a vacuum solution (special relativiy) makes GR a non-Machian theory.
But lets persevere: lets examine the multiple possible solutions you can get. Any solution is fully determined by appropriate boundary conditions*; but the fact that you can have two solutions that differ by gravitational waves is a blow against Mach's principle: here you have distortions in inertia propagating around. You can even have an otherwise vacuum solution with gravitational waves, where you have distortioins in inertia propagating in a geometry that is determined by no matter at all.
So yeah, while GR was derived with Machian thinking, it ended up being distinctly non-Machian. Einstein realized this the moment he saw Schwarzschild's solution.
* Indeed it is only by selecting vey special boundary conditions one can really attain any "Machianess" in GR.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
WST wrote:
Who told you that, can you please tell me?
That guy certainly never TASed Sonic.
QFT.
Quoted For Truth, not Quantum Field Theory
Indeed, the major reason why I only did one run without one glitch (level wrapping) is the sheer difficulty of making a glitchless run; no level wrapping already meant a huge amount of work, removing zips would make things vastly harder, and the thought of removing other glitches makes me give up a run I wasn't even really thinking of doing :-p
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
p4wn3r wrote:
Eventually we'll have to take this paradox seriously and resolve it, and it all comes down to a question Einstein asked way back. "Is spacetime a well-defined physical entity?" It looks like philosophical gibberish, but it's more fundamental to science than the formulas. If spacetime is well defined, then QFT is right, it makes sense to talk about a large spacetime and a small one and it's OK to vary the constants, let's reformulate GR. If it is not, then GR is right, we can only talk about spacetime because there's matter in it, and QFT should be changed instead.
Just a quibble here: if GR were a fully Machian theory, this would be correct; but it is not -- you have well defined spacetimes in GR in the complete absence of matter in all (or essentially all) points (see special relativity and Schwarzschild solutions). Push comes to the shove, GR is not a very Machian theory at all (because of vacuum solutions). And in fact, depending on what you mean by "well defined", GR gives an arguably more well-defined spacetime that the options -- because it is even more deterministic that Newton's laws.
Rather, the problem with GR and QFT is that the space-time given by GR is not defined in the context of QFT because the stress-energy-momentum tensor is only defined in probabilistic terms; so you don't have a well defined GR-spacetime in the context of QFT.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
Tub wrote:
How do you even define "traveling in time"? Movement is a change of position over time, so how does that work when you're talking about the position in the time dimension?
Simple: in relativity, you travel through paths in space-time, not merely in space; speed is the rate of change in space-time relative to your proper time and, in fact, everyone is already travelling in time to the future.
Time travel to the past, well, that is another thing, but the basic idea is the same: the difference being that your speed through time is oriented towards the past for at least a portion of the journey. Alternatively, you can say that you have time traveled to the past if, at any point, you are inside the past light-cone of a point you started at. Finally, another way is to talk about closed timelike curves: your trajectory intersects with itself at some point with an earlier proper time. The first technically fails in non-orientable manifolds; the middle one can say you are time traveling even if you are not doing anything special if you are in a weird enough space-time (your example with the photon, for example); but the latter will always be unambiguous.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
For Sonic 2, this is quite doable; for S3&K, they are quite an undertaking -- and I am speaking from experience here, even a run with zips in S3&K is an awful lot of work.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
GoddessMaria15 wrote:
Quite a nifty test run, WST. Is there any sort of lua or anything to keep track of Sonic's speed and positioning? I might want to mess with this game someday.
Here is the "TAS pack" for this game. I need to update the emulators to more recent SVN. Or maybe make aa BizHawk version if the script, assuming I can ever get it to work on Linux.
Edit: if yiu need help with any of the tricks and glitches, feel free to ask; I am the walking encyclopedia for Sonic games.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
That is not what I said; but then again, I said it in a very inaccurate way. What I meant to say was this: for all we know, the sum-over-all-possibilities aspect of QFTs could just be a way of approximating the effects of gravity at the quantum scale, and a full theory of quantum gravity would explain why this process works (and why renormalization is needed in QFTs), and give a whole different interpretation of the whole process. There has been some research on that front, but it is not yet "mainstream" physics.
But don't pay too much attention to me in this; particle physics and QFTs are not my field, I studied GR far more. That is, before I decided to leave physics for computer science.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
For S+T, Knuckles' area does not fully load and the camera locks too high; they can't go there. In other zones, things are better, but AIZ2 is not helpful.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
TheYogWog wrote:
Of course that's assuming Sonic can even make it to the zipline without losing his shield to the puddle underneath the first giant ring.
I can think of two ways: (1) jumping as the ledge breaks so you can double-jump before the pool and avoid touching the water; (2) having Tails down there to catch Sonic before the puddle and jumping right. In both cases, you won't lose the shield until your center point is submerged (same for fire shield), at which point you are "underwater".
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
The original idea was this: if you want the particle to bump into its past self in such a way that its trajectory will be changed enough that it won't bump into itself in the future (thus creating a paradox), you find out that you can't -- if the particle bumped into its past self, it will do so in a manner that will deflect it to bump into its future self in exactly that way (thus preventing a paradox).
In a sense, yes, it would be an ontological paradox; Novikov's principle is exactly that: the only paradoxes you can get in a time travel scenario are non-contradictory ontological paradoxes.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
There is Novikov's self-consistency principle/conjecture, which he formulated after trying (and failing) to set up a theoretical particle version of the grandfather paradox and always coming up with no paradox. The Wikipedia page sums it up nicely. So that particle hitting itself would probably just deflect its past self in such a manner that, when it bumped into itself in the future, would cause the same deflection it had experienced.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
tensor wrote:
I also have a question for everyone. I noticed while doing the test run that, as I mentioned, a respawned Tails can be controlled after he gets close enough to Sonic. I tried to search for information on it but wasn't successful. Are the exact conditions for it to happen known? Also, where exactly does he respawn?
If you look at my hud script, you can see timers for when Tails when despawn (the icon with his face in a red circle with a red slash through) and when he will respawn (just his face). You can prevent the despawn timer if you hurt Tails; getting him killed or jumping on a solid or platform object that is just offscreen will cause an immediate despawn. After Tails has despawned, the respawn timer will be initialized and will start running. While the respawn timer is going, you can press any button to force a respawn before the timer expires. Tails always respawns at with the same x cordinate as Sonic, and he spawns 192 pixels above Sonic (or 192 pixels below Sonic if anti-gravity is on).
When he is flying to Sonic, he aims for Sonic's position of 2 frames ago. For Tails to fall back down, he must be at that exact y coordinate and Sonic must not be in a hurt state or dying; moreover, Tails must also be either:
at the exact X position Sonic was 2 frames ago; or
the absolute value of the distance between Tails and Sonic's old position divided by 16 (and limited to $C00), plus the absolute value of Sonic's current speed divided by 256, plus 1 must be smaler than the absolute value of the distance between Tails and Sonic's old position.
And that is it — all the conditions required for Tails to fall down.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
Re: time travelling backwards in time in relativity: technically, you can do it. Well, not you: it is possible that some things do travel backwards in time, the equations of relativity (special or general) don't prevent that. What they do prevent is for something that was going one direction in time to turn around and go the other direction (although some of the weirder geometries in general relativity allow just that by moving in a specific way).
Re: ghost particles: personally, I think that the usual interpretation is a load of crap, an attempt to make sense of a mathematical trick that gives the correct results. Especially since gravity is nowhere to be found in QFTs because no one figured out how to do it yet*. Yes, it gives correct results; for all we know, it might be because QFTs give a way of approximating the effects of gravity in the quantum scale.
* And yelling "but gravity is too weak!" is a lame excuse -- general relativity is nonlinear, meaning gravity can become exceedingly strong in short range depending on the distribution of stress-energy-momentum, or even cause some other weird stuff such as flipping the sign of eletromagnetic fields at close range, as happens for a spinning charged blackhole.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
In the context of relativity, especially general relativity, time acquires a character no less real than that of space dimensions: relativistic physics becomes geometry in a 4-dimensional non-Euclidean space, and this space is curved. What is more, if you consider only curvature of space, instead of space-time, you get several results wrong: for example, you get half the value of light deflection by gravity that you observe with modern experiments. So in this sense it can be said to exist.
Moreover, traditional "classic" quantum mechanics has long since been replaced by quantum field theory, which basically can be summed up as quantum mechanics in the spacetime of special relativity (although this is a gross simplification). Since quantum field theory can give some pretty accurate results in general, one can infer that time as an axis exists here as well.
Fact of the matter is that, if you really want to, you can describe all of physics without time as a "true" axis. Your equations will be hideously complex, much more so than the alternative; and so will interpretations of what you come out with, because you will have to add a load of novel effects to explain the appearance of relativistic effects. In the end, it is simpler to just use relativity. Scientists, being ultimately pragmatic, takes this approach and assume that the theory means time exists as a separate axis.
But you ought to know that the creator of any theory or hypothesis in science is not the ultimate arbiter about it -- physics has moved on from Einstein's original views quite a bit, so what he believed about relativity is essentially irrelevant today except for a historical perspective. Physicists today work with time as a separate axis as a matter of fact.
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
Ragnarokia wrote:
Sadly your 3:55.7 time even for a single level is over 3 times as long as the time this level took when glitching. Looking at Nitsuja and Upthorn's TAS of this they took 1:11 game time to beat Act 2 of Flying Battery Zone. You may personally think glitches are boring, but without them it is clearly FAR slower.
Is his time 3:55.7 for the in-game time or for everything in the run? From the submission title, I think it is for the whole movie, but since BizHawk doesn't work on Linux (and the Linux version from the thread crashes on start), I have no way of knowing.
Edit: Actually, there was -- I opened the input file in a text editor and saw that 3:55.7 is for the whole movie. There is also evidence of losing time when charging spindashes, and the level select activation seems sloppy too. Does not inspire much confidence...
Experienced Forum User, Published Author, Experienced player
(751)
Joined: 9/29/2008
Posts: 964
Location: 🇫🇷 France
I will quibble with the 'skipped' in MGZ2 and ICZ2; there are title cards for both levels and the timer runs for a bit in both of them. But it is looking solid.
Out of curiosity, is this Amy solo or Amy+Tails?
