Talk:Physics engine

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Physics is also about optics, so why is it not called game mechanics ?Arnero 20:06, 1 November 2006 (UTC)[reply]

Because 'Physics Engine' is what 100% of the people in the business call these software libraries - they might well have chosen the name poorly - but that doesn't in any way alter the fact that 'Physics Engine' is what they are called. If there is ambiguity with something else that's also called 'a physics engine' then create a disambig page - but calling this by a name that nobody in the field uses would be ridiculous. The term 'game mechanics' has a completely different meaning to computer games people - it means (roughly) 'the rules by which the game is played'. Things like: Do you get to pick up more ammunition for your gun? Do you win by killing all of the enemy or by rescuing the princess? So calling this article "Game mechanics" would be terminally confusing for all involved. Nope - the name of the article is 100% correct. Sorry, SteveBaker 14:48, 2 November 2006 (UTC)[reply]

This needs more cites, but it's a difficult topic to cite[edit]

Other than Second Life and Gary's Mod for Half Life 2, there aren't many easily accessible ways for non-programmers to experiment with physics engines.

Meanwhile, the documentation of how physics works in Second Life isn't written down anywhere in an official manual, but if you start experimenting with physics you will discover these operational conditions sooner or later.

I might eventually provide other video demonstrations like my SL exploding wheeled cart, since about the only way to show how it works is to actually do it and then record the results.

DMahalko 22:25, 27 April 2007 (UTC)[reply]

There are plenty of citations available. Look in SIGGRAPH papers from 1995-2005, especially those of Baraff. He did the impulse/constraint/linear complementarity approach to game physics. That's widely used because, even though it assumes idealized collisions and handles friction badly, it can be computed in nearly constant time for any given scene. Spring/damper methods are more realistic but sometimes slower. --John Nagle 00:06, 25 May 2007 (UTC)[reply]

Paradigms paradigm is wrong...[edit]

The list of "paradigms" is not an established list. In fact, most if not all engines use "constraints" to model jointed/constrained mechanical systems. These constraints can be implemented as soft or hard potentials. A soft potential is often referred to as a penalty method and the most common model is a spring-damper model. This model runs into stability problems for stiff springs. In the limit of hard potentials the model goes into an implicit-or semi-implicit scheme that requires that a system of equations is solved. This can be solved using either direct or iterative solvers. The use of propagational iterative solvers (e.g. Gauss-Seidel) is what is nick-named "impulsive model" since the successive iterates update the velocities of the system in a way that reminds of impulses. A more correct notation would therefore be:

  • Soft penalty potentials, e.g. springs.
  • Stiff potentials formulated as constraints solved using e.g.
    • Sparse direct methods
    • Sparse iterative and propagational methods (somtimesnick named "impulse based methods"). — Preceding unsigned comment added by 212.217.204.225 (talk) 16:52, 22 November 2011 (UTC)[reply]

Bouding Box?[edit]

In the first paragraph under the Game Engines, it says "The simplified mesh used for physics processing is often referred to as the bounding box." I believe a bounding box is not just any simplified mesh, but as its name suggests, is a box that bounds the object and is used for a preliminary test for collision or for the actual collision test itself. —Preceding unsigned comment added by 69.109.209.33 (talkcontribs)

Your're right. The correct term is "collision geometry". The collision geometry may be a simple box, but the state of the art is well past that. Fixed article. --John Nagle 04:21, 8 November 2007 (UTC)[reply]

Fair use rationale for Image:Kis2.jpg[edit]

Image:Kis2.jpg is being used on this article. I notice the image page specifies that the image is being used under fair use but there is no explanation or rationale as to why its use in this Wikipedia article constitutes fair use. In addition to the boilerplate fair use template, you must also write out on the image description page a specific explanation or rationale for why using this image in each article is consistent with fair use.

Please go to the image description page and edit it to include a fair use rationale. Using one of the templates at Wikipedia:Fair use rationale guideline is an easy way to insure that your image is in compliance with Wikipedia policy, but remember that you must complete the template. Do not simply insert a blank template on an image page.

If there is other fair use media, consider checking that you have specified the fair use rationale on the other images used on this page. Note that any fair use images uploaded after 4 May, 2006, and lacking such an explanation will be deleted one week after they have been uploaded, as described on criteria for speedy deletion. If you have any questions please ask them at the Media copyright questions page. Thank you.

BetacommandBot 21:45, 6 November 2007 (UTC)[reply]

nVidia buys Ageia...[edit]

http://www.nvidia.com/object/io_1202161567170.html

That's interesting - it means that the patent issues that are thought to have clobbered the NV Physics SDK will magically "go away", so nVidia would be free to do Physics on the GPU much more cleanly. Hardware physics actually stands a chance of happening now! SteveBaker (talk) 21:19, 5 February 2008 (UTC)[reply]

I added Phun tot the list[edit]

I added Phun to the closed source/limited free distribution list. Phun is a 2D physics sandbox from Algoryx. I might add an article about it, as most things are well documented on our own wiki anyway.

~link0007 —Preceding unsigned comment added by 84.25.229.124 (talk) 16:16, 2 July 2008 (UTC)[reply]

Policy for listing engines and software?[edit]

I'd like to know the policy for referencing physics engines here. Apparently, several engines were deleted from the lists, as well as their entire entries in wikipedia - while some other engines are apparently still listed, and also still have their own entries in wikipedia intact. The policy for deletion appears to be pretty random, and this strikes at the quality of the entire topic.Kenbon (talk) 15:12, 8 October 2008 (UTC)[reply]

I doubt there is a well-defined policy, but my view is that if there is an article on a particular physics engine it should be linked to from here or (if the list gets too long) from a separate List of physics engines article. If engines currently linked to from this article are non-notable, their articles can be nominated for deletion. If the nomination is successful, then the resulting red-links should be removed from the list here. I am against the idea of simply listing every single engine that anyone has ever heard of with nothing more than a link to a website for that engine. That would turn Wikipedia into a directory rather than an encyclopedia. The downside of that is that the actual encyclopedic content gets lost in the mush.
In general, this article should not be a battleground for what is and is not a notable or important phsyics engine. If it is to list physics engines at all (which I think it should) it should list all of them that are worthy of articles in Wikipedia at large. No more, no less. GDallimore (Talk) 10:58, 9 October 2008 (UTC)[reply]
Well, I think the randomness of entries deleted for non-notability, and total lack of interest from editors to act consistently is a serious problem. Very, very few of the entries with their own pages have any sort of independent referencing, so if anyone suggested them for deletion, they would go away. However, the "suggest for deletion" process is totally random. I start to understand the all the criticism on wikipedia that exists. What do you think should be required specifically for a physics engine in terms of independent referencing to make a physics engine "notable"? —Preceding unsigned comment added by Kenbon (talkcontribs) 05:37, 10 October 2008 (UTC)[reply]

I'd like to know who the heck mentioned working model as 'high preicision'. I just spent a semester in a class learning how to use that thing, and it's horrendous - you can't even define reference geometry! Every time you want a reference item, you have to define an item, then tell it not to collide, and change its mass to zero, and then make it invisible after you connect things to it. Then, when you click run, unless you set your frames to be ridiculously close together, it will come up with some problem about how an object passed entirely through another since the last frame. —Preceding unsigned comment added by 24.121.161.224 (talk) 03:54, 24 December 2008 (UTC)[reply]

Engine missing from the list[edit]

just mengioning there is Digital molecular matter missing from the engine list :D —Preceding unsigned comment added by 90.191.39.41 (talk) 02:17, 15 April 2009 (UTC)[reply]

Another engine missing in the list of proprietary engines is AGX Multiphysics by Algoryx Simulation. It is the engine used e.g. in the Algodoo physics sandbox, which has had about two million downloads, and is used by several hundred thousand users in 57 countries, and it is also used by numerous industry users for engineering simulation. It was once listed but apparently removed by "Dantexyz" with the somewhat astonishing argument that AGX is "not a popular physics engine". — Preceding unsigned comment added by DrProfAlb (talkcontribs) 15:13, 23 February 2011 (UTC)[reply]

A proposal for the comparison of physics engines Wikipedia article[edit]

Initial information can be found here 188.16.99.208 (talk) 12:38, 7 April 2010 (UTC)[reply]

Brownian motion[edit]

After the intro sentence in the "Brownian Motion" section, the rest of the section is unrelated to brownian motion. A changed heading would be appropriate, but I'm a physicist, not a game guy, so I don't know what you'd actually call the algorithmic stuff it's referring to. —Preceding unsigned comment added by Horn.imh (talkcontribs) 23:03, 1 May 2011 (UTC)[reply]

The 'algorithmic stuff' related to the idea of 'Sleep' is directly related to the idea of Brownian Motion as it is the often used method of reducing Jitter as eluded to in the second sentence. The second paragraph was completely unrelated to the subsection and I chose to move it into the section introduction. John Six (talk) 15:25, 2 October 2014 (UTC)[reply]

Added another physics engine[edit]

Rubikon is Valve's in house physics engine for Source 2. — Preceding unsigned comment added by 74.96.229.39 (talk) 14:39, 21 October 2020 (UTC)[reply]

Wiki Education assignment: 4A Wikipedia Assignment[edit]

This article was the subject of a Wiki Education Foundation-supported course assignment, between 21 August 2023 and 16 December 2023. Further details are available on the course page. Student editor(s): Suidrew (article contribs). Peer reviewers: Timbenuka, Kmijares.

— Assignment last updated by Kmijares (talk) 22:40, 15 November 2023 (UTC)[reply]