Jump to content
  • Advertisement


  • Content Count

  • Joined

  • Last visited

Community Reputation

100 Neutral

About jovani

  • Rank
  1. Quote:Original post by Ravyne Quote:Original post by jovani It sounds like you are looking for a hardware solution to implement your own proprietary engine. If this is the case then the answered you are getting will not lead you anywhere. Havok does not have a networked solution for physics neither does Ageia. That has *nothing* to do with it. If you have the server side engine, and that is distributed over the network the way MMO servers are, the physics solution itself doesn't deal with the "distributed" aspect of the system. I am sorry it did not know you had a similar problem in the pass and you solved that way. However I stand by what I said, in my humble experience I had only seen physics in a networks game in the valve engine, which uses Havok the way you are saying. My impression from Telamon was that the physics solution itself was going to be networked, which is not possible with any physics engine to my knowledge, so recommending Havok, Ageai, Nvidia, Ati, Gpu hardware, or any other not tested solution is just fiction. Quote:Original post by Ravyne Now, the clearspeed cards are nice for many things, but the cost is prohibitive and its likely that they might even be overkill. Its no poin't putting a $10,000 card into a server that can handle 1000000 physics objects if the box's other resources can only handle 1000 players. It would be far more effective to slap a 8800GTX cards in for 1/16th the cost and adequate performance. --- Also, I just checked the numbers and the 8800GTX, which is little more than 128 single-precision FPUs, can sustain 330gflops. Granted the clearspeed card is double-precision, but even allowing generous overhead the nVidia card has at least 5 times the potential throughput at 1/16th the price. Well there you go maybe the G8800 is better than Clearspeed and may become a consumer add on card for more general programming but it is not a real solution yet. I was given a real and general hardware answer, I still do not know how a video card can be used for general physics programming other than cute particle effects demos with shaders. But you are right maybe the new G8800 and the promised CUDA language will be a solution in future.
  2. It sounds like you are looking for a hardware solution to implement your own proprietary engine. If this is the case then the answered you are getting will not lead you anywhere. Havok does not have a networked solution for physics neither does Ageia. There is a hardware solution you can try. http://www.clearspeed.com/ It is not cheap because it is not a consumer card but it can produce 25 gfloats in double precision of sustained throughput, and you can connect many of them into an array. It is priced at around $10000 depending on the options, but if you consider that a Havok license goes for over $100000, (not including gpu effects and extras), it is not that expensive is you already have your own software. You can achieve top of the line (blue gene kind of performance) with few of them depending on how much you are willing to invest. Other than that your only other solution would be a farm of xeons quad cores [Edited by - jovani on January 12, 2007 9:27:22 PM]
  3. jovani

    How to model Projectiles?

    Quote:Original post by jchmack Im making an fps using the ageia physics engine and im wondering if what im planning to do is a good idea to model my projectiles. If you are using Ageia why not using their functionality. If it can do this chances are that it will handle your fps too
  4. jovani

    inertia tensor and mass problem

    Quote:Original post by Airo Quote:Original post by remdul Okay, I'm stuck now on the integrator. I came across an alternative to RK4, which seems to be simpler, faster and more accurate than RK4, described here: http://www.brightland.com/Physics/index.htm So basicly this is it: v += a*dt x += v*dtBut I'm completely clueless as to what to do with this. What numbers go in, and what comes out? How can I turn this into a useful function? Example/source code would be greatly appreciated. That is not more accurate than RK4. You plug in the same numbers in that code as you would do when you use RK4. RK4 calculates the acceleration a four times and here it is only calculated once. I suggest you to read more literature about the subject, starting with baraff's siggraph lectures. Hallelujah, some body finally sees the light. About time misinformation like this was debunked, some people some how think changing the name of a first order euler integration will change the error from O(k1 * h^3) for euler to O(k4 * h^5) for rk4 However is important to state that to make a fair comparison the Euler integration would have to be evaluated 4 times for each RH4 evaluation. In these situations RK4 is still better because of the smoothing of the derivatives, but only marginally. Also given that one step of RK4 is more expensive than 4 steps of euler (both memory and operations), plus is more inconvenient, it is better to use Euler with smaller step. Not because is better, but because is more convenient.
  5. jovani

    PhysX - was Novodex, is free!

    Actually this is a typical marketing method used by big companies to drive the competitors out of business. Is called “Loss Leader Strategy”, it had been used for about hundred years on capitalistic markets and it is considered one of the more extreme aberration of capitalism, it is illegal in some countries. For this strategy to work the company actually has to have a profitable product, and in this case the strategy is backward, the profitable product is given at a lost with the hope to save the unprofitable product. This is not a strategy that comes without consequences. Only Behemoth companies like Wal-Mart, Microsoft, Sonic, can do this for an indefinite amount of time until the competitor go out of busisnes.
  6. Quote:Original post by jyk Just to avoid further confusion it should be pointed out that 'column major' is not the correct term here. 'Column major' refers to the matrices' layout in memory, which isn't under discussion. It's probably safe to say though that the OP is using column vectors. @The OP: You still haven't really provided enough info for us to guess at a solution. As has been suggested, if (AB)C and A(BC) produce different results, then there's a problem with your math code. If your math code is correct and neither ABC nor ACB gives you the results you're after, then there's a problem either with how you're conceptualizing the transforms, or with how the transforms are being constructed or applied I concede to your knowledge of algebra, I am not pretending to teach you anything I was trying to give an answer because I thought GCoda did provide enough information, so I do not know why spinning it with pedantic definitions. I always thought that column major and row major was the term used to express whether a vector was written as a column into a matrix or as a row into a matrix. I always though that associativity of matrix multiplication was still a valid law of linear algebra. It was based on that fact is that I suggested that, if GCoda is seen different results in the operations A * (B * C) and (A * B) * C then it may be that his matrixes multiply function has a bug. I may be wrong and you may be right, if so I apologize. It is just that I had never seen correct implemented linear algebra code leading to A * (B * C) != (A * B) * C Regardless of how the matrices are written be row or vector column. But like I said I may be wrong.
  7. The fact that you are concatenating the matrices from left to right, imply yours matrices must be written in column major order. That is, If you write M = A * B * C the template for the matrix format is: x11 x12 x13 x14 x21 x22 x23 x24 x31 x32 x33 x34 0.0 0.0 0.0 1.0 I would not make sense any other way, since matrices do not commute and since you stated that A is the world matrix. the only way this make sense is if vectors is placed on the right side of the equation, say Pw = A * B * C * Pl Where Pw is a point in world space and Pl is the same point in local space. The other thing is that, the order of multiplication is irrelevant A * (B * C) is always equal to (A * B) * C If you are getting different results then you may have a bug in the matrix multiply function itself
  8. Quote:Original post by jyk Quote:Original post by Rajveer Then you would do A * (B * C). A local transformation is on the right, and a world transformation is on the left...This depends on whether row or column vectors are being used, which the OP did not specify. I think hi did, Hi stated that A is the world transform and B is the object transform and he wrote the expression A * B So we must conclude that A * B * C is the correct order, as written by Rajveer
  9. jovani

    Black Holes

    Quote:Original post by taby It's not that you're wrong per se, it's just that you blindly trust the theories of others, and they are wrong. Ha thank you for the remark I will try to be more careful with my sources. I must say however that I am nobody to contradict 80 years of observations that seems to confirm over and over again Chandra Limit’s theory. It seems I am not the only one that blindly believe his theoretical deductions, since he was awarded a novel prize and NASA named their more advanced and powerful observatory satellite after him, the Chandra X-Ray Observatory. What I find estrange is that you accuse me of blindly trusting a theory of others, that you say it is wrong. Yet you are believe in “another theory of others” that for some reason you think is true. I think that if you are going to make such claim you need to produce more evidence than a TV news flash about an observation of one particular supernova as proof. I remind you that what Chandra limits states is that the critical value for a star to end its life as black hole is 1.4 solar masses, it does not say anything about more massive stars. I read the TV news report and here is a quote from it. Quote: The study's co-author, Mark Sullivan of the University of Toronto, dubbed it a "rogue supernova." The team that found the supernova has two theories about how the white dwarf got so big before it exploded. One is that the star was spinning at such a high speed that gravity couldn't pull it in and crush it at the normal Chandrasekhar limit.Another theory is that the observed supernova is the result of two white dwarfs merging and that the two stars together exceeded the limit only briefly before exploding. It seems to me that what is said is that they found a supernovae that originated from a larger than 1.4 solar masses white dwarf, and they are given two possible reasons for it. Basically they think some internal forces are working to prevent the implosion into a supernovae when the dwarf was 1.4 solar masses. It does not say that they found a supernova that originated from a smaller than 1.4 solar masses white dwarfs, which is what would be requiered to prove your theory of the density as the reason for black holes formations. ......... You insinuated that Arthur Eddington was right, and Chandra was wrong. Quote:Original post by taby Evidence to the contrary of the immutability of Chandrasekhar's limit has been found. Perhaps Eddington wasn't quick to judge after all: That was not the only thing Arthur Eddington was famous for; He did not believe in the decay of the electron either, in fact he is quite famous since he also claimed he had a secret formulae that can calculates the exact number of protons in the universe. Legend said that he was prompt to fabricate scientific facts in order to discredit other fellow researchers. I think if he was still around hi would had good chance at been hire by Gene Rodenberry as Star Trek writer. [Edited by - jovani on November 2, 2006 3:29:34 PM]
  10. jovani

    Black Holes

    Quote:Original post by Rockoon1 There is no rule that a black hole needs to be above a certain mass or size threshhold. Density is the determining factor, and as such its a combination of BOTH mass and size. If someone cares to try, plug in the theoretical mass of the universe into the event horizon equation and be shocked. Pretty big huh? Damn near the size of the visible universe! Maybe we know exactly what its like to be within one! My comments where based more on the work of theoretician like Chandrasekhar Who won a novel price for proving that no clump of mass smaller than 1.4 solar masses can became a black hole. Later that was confirmed to be about 3 to 4 solar masses for active starts with the potential to explode in supernovae. http://scienceworld.wolfram.com/physics/ChandrasekharLimit.html Maybe you are right and I am wrong, and there had been some new discovery probing the existence arbitrary small mass size black holes based only on the idea of density. Do you have some reference to some publications or some observations of any of these arbitrarily smalls black hole? I have not read any and i am an astromy reader buff ;)
  11. jovani

    Black Holes

    Quote:Original post by erissian A black hole can have any mass provided it's compact enough, although its lifetime is proportional to its mass. Interesting that you should mention iron though, it being the most prolific byproduct of the stellar fusion process. Could a clump of mass the size of the earth or the moon become a black hole? if so what natural or artificial phenomenon can produce that kind of object. Also why it is not happing to all the planets. It was my impression that what stop gravity from crunching every thing was the electrostatic force at planetary scale, and nuclear forces at the scale of the stars. Maybe I am wrong but I though that if a star, after it burned all its fuel, if it did not have enough mass to crunch the nuclear force between the neutrons, it became a cool object know as neutron star. I though that the critical mass for that to occur was calculated to four solar masses. Not that I am astronomer or astrophysicist I just like to read Astronomy magazine. I am just curios as how mini black hole form.
  12. jovani

    Black Holes

    Quote:Original post by erissian Black holes aren't all consuming beasts. Our own Sun could be replaced by a black hole of the same mass and we'd do just fine (except for the whole heat, light, and radiation thing) :) For the planets to do fine it would have to be a black hole of the same mass. I think I read somewhere that the minimum mass required to generate a gravity force strong enough to overcome the nuclear forces between protons and neutrons is about three to four solar masses. So even if the sun was made of iron it cannot turn into a back hole by natural causes.
  13. jovani

    Constant Angular Momentum

    I am not expert but if you take a look at the form of the expression for the derivative of angular momentum, you will see that the differential equation to calculate the angular velocity is highly non linear. Ix * W’x(t) = Iyz * Wy(t) * Wz(t) + Tx(t) Iy * W’y(t) = Izx * Wz(t) * Wx(t) + Ty(t) Iz * W’z(t) = Ixy * Wx(t) * Wy(t) + Tz(t) Where: Ix, … are the inertia around the local axis Ixy = Ix - Iy , …. are the inertia gyroscopic coupling Wx, …. are the local angular velocity Tx, … in the external net torque The only time the angular velocity coupling effect vanish is when the inertia tensor is spherical meaning Ix = Iy = Iz When W is very small product of Wxy is even smaller an the system become stable but even at moderate values of Ws, there are not known method of integration that can solve the above differential equation. Some people have naively used tricks of projecting the momentum back over the angular momentum space but that is quite a terrible heuristic with not theoretical ground, and it would only be valid on an environment devoided of torque and forces (not very practical). So while in theory angular momentum is conserved, in practice it is not possible to calculate it with any degree of accuracy numerically. What mainstream physics engine do to go around the integration problem is adding an attenuation term to the equation. Ix * W’x(t) = Iyz * Wy(t) * Wz(t) + Tx(t) – ks * Ix * Wx(t) Iy * W’y(t) = Izx * Wz(t) * Wx(t) + Ty(t) – ks * Iy * Wy(t) Iz * W’z(t) = Ixy * Wx(t) * Wy(t) + Tz(t) – ks * Iz * Wz(t) Where ks is a damping coefficient. This is what explain why spinning objects slowdown when simulated with realtime physics engines.
  14. jovani

    A question about CRC32

    But you just said that a collision did happened in a population of 24000 words. That means a non-zero probability and that rules CRC out. Had those two words be first and second in the array you have a big problem. The only zero probability of collision solution is the word itself.
  15. jovani

    A question about CRC32

    Why making it so hard? Why not using a variable size arithmetic or Huffman encoding? I can guarantee you than with 10000 words, any of the popularly CRC32 routines will generate at least one collision.
  • Advertisement

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!