Loading ...
Sorry, an error occurred while loading the content.

Future games, architectures and acceleration

Expand Messages
  • Ahmet
    I am currently working with a venture capital company interested in the game market. I wanted to get your opinions about the recent physics accelerators. I am
    Message 1 of 21 , Apr 19, 2006
    • 0 Attachment
      I am currently working with a venture capital company interested in
      the game market. I wanted to get your opinions about the recent
      physics accelerators. I am a bit puzzled about this. It seems the CPU
      and GPUs are getting faster all the time, and with the difficulty of
      programming to new platforms, I wonder if developer community will
      embrace these "accelerator" technologies.

      Any thoughts?
    • Brandon J. Van Every
      ... Here s one article talking about physics accelerators, dated March 22, 2006: http://gear.ign.com/articles/697/697757p1.html My first reaction, as an indie
      Message 2 of 21 , Apr 19, 2006
      • 0 Attachment
        Ahmet wrote:
        > I am currently working with a venture capital company interested in
        > the game market. I wanted to get your opinions about the recent
        > physics accelerators. I am a bit puzzled about this. It seems the CPU
        > and GPUs are getting faster all the time, and with the difficulty of
        > programming to new platforms, I wonder if developer community will
        > embrace these "accelerator" technologies.
        >
        Here's one article talking about physics accelerators, dated March 22, 2006:
        http://gear.ign.com/articles/697/697757p1.html

        My first reaction, as an indie game designer, is that this is inevitably
        a big budget, high programming expense market. Frankly I've been
        avoiding mere pixel shaders for many years now, because there has been
        an incredible amount of churn in the shader standards. My philosophy
        has been "wake me when it's over" and I've instead concentrated on more
        strategic factors. I don't have an army of programmers to sit around
        dealing with all the platform specific minutiae. Outfits that are
        forking over for an Unreal engine can afford to, whether for pixel
        shaders or even more newfangled physics accelerators. It may also be a
        good problem to outsource to the Third World. Tedious, canned,
        potentially limited in scope compared to the rest of the game.

        I'm not a Luddite; I'm investing heavily in the Chicken Scheme
        programming language, which nobody else uses. But if I'm going to
        invest in technical R&D, it has to be something of general value where
        I'm calling the shots. Not something that's going to be churned
        according to someone else's business model. I expect film licensees to
        drive this. Physics acceleration is part of the quest for visual
        realism and higher production values. The actual gameplay value of
        physics is limited. Especially, you don't need particularly fancy "game
        physics" to get the low hanging fruit from a game designer's standpoint.

        I expect that physics accelerators will ultimately take hold in the
        marketplace. Sure, a platform like the PS3 Cell architecture isn't
        going to need it, but PCs will. There will be battles about who
        supports what architecture. Look to the early history of 3D
        accelerators to see what will happen. Some HW architectures will take
        root, someone will become de facto. My attitude towards all of this is
        wake me when it's over. A venture capitalist's attitude should be,
        concentrate on execution. Lotsa guys are gonna try, someone's gonna
        succeed, many will lose. Consider all the 3D cards that used to exist,
        and now for gaming purposes we're down to ATI and NVIDIA.

        I think the best business models will push towards cinematic realism,
        not gameplay. Gameplay is too easy to do - better - without an accelerator.

        I'd also say that physics accelerators are much more straightforward
        than AI accelerators. I don't see the latter taking hold anytime soon.
        Software will remain where it's at. Physics is more of a canned
        problem, more amenable to committment to HW.


        Cheers,
        Brandon Van Every
      • waghdude
        ... I doubt that the mainstream computer user and game player will open up his or her computer just to add this one card, for this one particular purpose. The
        Message 3 of 21 , Apr 21, 2006
        • 0 Attachment
          --- In gamedesign-l@yahoogroups.com, "Ahmet" <takeaction@...> wrote:
          >
          > I am currently working with a venture capital company interested in
          > the game market. I wanted to get your opinions about the recent
          > physics accelerators. I am a bit puzzled about this. It seems the CPU
          > and GPUs are getting faster all the time, and with the difficulty of
          > programming to new platforms, I wonder if developer community will
          > embrace these "accelerator" technologies.
          >
          > Any thoughts?
          >

          I doubt that the mainstream computer user and game player will open up
          his or her computer just to add this one card, for this one particular
          purpose. The msrp on the PhysX is in the hundreds.

          A physics chip would be much more likely to be adopted if included as
          part of a more common gaming related upgrade: a video card. Video
          cards, themselves, contain a lot of number-crunching power, and I
          wouldn't be surprised if a future video card simply provided some
          physics-related features, whether through the Ageia chip or on their own.

          Would the developer community embrace this kind of acceleration? It
          probably means most to those who are making games where realistic
          physics is of the utmost importance, and who also feel the additional
          expense of assigning a programmer or two on this task is warranted.
          These developers (and their customers) would have to be utterly
          disgusted with the lack of realism that's currently available.

          If the support can be transparent in some way, i.e. the engine you're
          using does it for you versus your code invoking those features
          directly, then it'd be less of a gamble to give the specialized
          hardware a try.


          +pravin
        • Brandon J. Van Every
          ... Cost aside, it is worth noting that the majority of mainstream computer users won t open up their boxes for any reason whatsoever. Whatever 3D card is
          Message 4 of 21 , Apr 21, 2006
          • 0 Attachment
            waghdude wrote:
            > --- In gamedesign-l@yahoogroups.com, "Ahmet" <takeaction@...> wrote:
            >
            >> I am currently working with a venture capital company interested in
            >> the game market. I wanted to get your opinions about the recent
            >> physics accelerators. I am a bit puzzled about this. It seems the CPU
            >> and GPUs are getting faster all the time, and with the difficulty of
            >> programming to new platforms, I wonder if developer community will
            >> embrace these "accelerator" technologies.
            >>
            >> Any thoughts?
            >>
            >>
            >
            > I doubt that the mainstream computer user and game player will open up
            > his or her computer just to add this one card, for this one particular
            > purpose. The msrp on the PhysX is in the hundreds.
            >
            Cost aside, it is worth noting that the majority of mainstream computer
            users won't open up their boxes for any reason whatsoever. Whatever 3D
            card is sold with the system, that is typically what the user stays
            with. Users also don't update drivers for the most part. The PC is a
            decided PITA in this respect.

            Conversely, if you can figure out what segment of the market actually
            opens up and installs new 3D cards, you can probably get 'em to stick
            other cards in their boxes some % of the time as well.
            > A physics chip would be much more likely to be adopted if included as
            > part of a more common gaming related upgrade: a video card. Video
            > cards, themselves, contain a lot of number-crunching power, and I
            > wouldn't be surprised if a future video card simply provided some
            > physics-related features, whether through the Ageia chip or on their own.
            >
            Agreed. It will all inevitably be consolidated onto 1 card. AI chips
            as well, someday, but those are going to take much longer.

            > Would the developer community embrace this kind of acceleration? It
            > probably means most to those who are making games where realistic
            > physics is of the utmost importance, and who also feel the additional
            > expense of assigning a programmer or two on this task is warranted.
            >
            Yep. And those development studios are out there.

            > These developers (and their customers) would have to be utterly
            > disgusted with the lack of realism that's currently available.
            >
            I doubt it. They'd just have to think they could make money on an early
            adopter thingamabob.

            > If the support can be transparent in some way, i.e. the engine you're
            > using does it for you versus your code invoking those features
            > directly, then it'd be less of a gamble to give the specialized
            > hardware a try.
            >
            >
            I don't see how an engine is going to do physics for you, any more than
            it's going to make a game design for you. It could certainly streamline
            the amount of work you have to do to put physics into your game, though.


            Cheers,
            Brandon Van Every



            [Non-text portions of this message have been removed]
          • waghdude
            ... streamline ... When you create 3D models for a game, lighting is not something that you have to concern yourself with for rendering each and every model.
            Message 5 of 21 , Apr 22, 2006
            • 0 Attachment
              --- In gamedesign-l@yahoogroups.com, "Brandon J. Van Every"
              <bvanevery@...> wrote:
              :
              > > If the support can be transparent in some way, i.e. the engine you're
              > > using does it for you versus your code invoking those features
              > > directly, then it'd be less of a gamble to give the specialized
              > > hardware a try.
              > >
              > >
              > I don't see how an engine is going to do physics for you, any more than
              > it's going to make a game design for you. It could certainly
              streamline
              > the amount of work you have to do to put physics into your game, though.

              When you create 3D models for a game, lighting is not something that
              you have to concern yourself with for rendering each and every model.
              You set up some parameters for the lighting sources in the world and
              the rendering system automatically applies those parameters to the
              objects you wish to render. Shadows, motion blurs, and other effects
              are also included for free once you've set up those features into the
              rendering pipeline.

              In a very similar way, once you have assigned all of the physics
              parameters for the objects in your world, the physics engine that's
              part of your engine (and whatever extensions and special features your
              engine supports) become available to you for free.

              That's how the engine does the physics for you.

              It's not "free" in terms of performance, and the point of physics
              acceleration hardware is to unburden the main CPU so that you might be
              much more generous in enabling physics for most of the objects in your
              world.

              Right now, it's more likely that the heavy duty physics are enabled
              for the playing character, characters in his party if there is such a
              concept in the game, some types of enemies, and selected objects in
              the world that the player is likely to interact with (by that I mean
              destroy).


              +pravin
            • David Haley
              ... But are the mainstream computer users the market for such cards to begin with? Probably not. Maybe you meant mainstream computer _gamers_, and by then
              Message 6 of 21 , Apr 22, 2006
              • 0 Attachment
                On this day of 4-21-2006 8:57 PM, Brandon J. Van Every saw fit to scribe:
                > waghdude wrote:
                >
                >> --- In gamedesign-l@yahoogroups.com, "Ahmet" <takeaction@...> wrote:
                >>
                >>
                >>> I am currently working with a venture capital company interested in
                >>> the game market. I wanted to get your opinions about the recent
                >>> physics accelerators. I am a bit puzzled about this. It seems the CPU
                >>> and GPUs are getting faster all the time, and with the difficulty of
                >>> programming to new platforms, I wonder if developer community will
                >>> embrace these "accelerator" technologies.
                >>>
                >>> Any thoughts?
                >>>
                >>>
                >>>
                >> I doubt that the mainstream computer user and game player will open up
                >> his or her computer just to add this one card, for this one particular
                >> purpose. The msrp on the PhysX is in the hundreds.
                >>
                >>
                > Cost aside, it is worth noting that the majority of mainstream computer
                > users won't open up their boxes for any reason whatsoever. Whatever 3D
                > card is sold with the system, that is typically what the user stays
                > with. Users also don't update drivers for the most part. The PC is a
                > decided PITA in this respect.
                >
                > Conversely, if you can figure out what segment of the market actually
                > opens up and installs new 3D cards, you can probably get 'em to stick
                > other cards in their boxes some % of the time as well.
                >

                But are the mainstream computer users the market for such cards to begin
                with? Probably not. Maybe you meant mainstream computer _gamers_, and by
                then you've already upped the percentage several notches from users. But
                even then, the real market for such independent physics cards is still
                not the casual gamer.

                >> A physics chip would be much more likely to be adopted if included as
                >> part of a more common gaming related upgrade: a video card. Video
                >> cards, themselves, contain a lot of number-crunching power, and I
                >> wouldn't be surprised if a future video card simply provided some
                >> physics-related features, whether through the Ageia chip or on their own.
                >>
                >>
                > Agreed. It will all inevitably be consolidated onto 1 card. AI chips
                > as well, someday, but those are going to take much longer.
                >

                What would an 'AI chip' do? You could have it take care of search
                algorithms, I suppose, or depending on what your AI was doing, computer
                vision code, etc. But I'm curious what you have in mind when you speak
                of an 'AI chip'.

                >> Would the developer community embrace this kind of acceleration? It
                >> probably means most to those who are making games where realistic
                >> physics is of the utmost importance, and who also feel the additional
                >> expense of assigning a programmer or two on this task is warranted.
                >>
                >>
                > Yep. And those development studios are out there.
                >

                I don't think they're as rare or unusual as you make them out to be.
                Look at Elder Scrolls IV: Oblivion. That game has an amazing physics
                model, and you can tell that just by looking at the preview videos where
                the character interacts with the chain and lets it flop around. Physics
                simulations have been used in several games for ragdoll deaths of
                characters. Realistic physics are not just for, say, a car or plane game
                where physics are truly important for the gameplay; physics are highly
                important for decorative/immersive purposes and that should _never_ be
                underestimated.

                >> If the support can be transparent in some way, i.e. the engine you're
                >> using does it for you versus your code invoking those features
                >> directly, then it'd be less of a gamble to give the specialized
                >> hardware a try.
                >>
                >>
                >>
                > I don't see how an engine is going to do physics for you, any more than
                > it's going to make a game design for you. It could certainly streamline
                > the amount of work you have to do to put physics into your game, though.
                >

                I'm a little surprised to see you make a statement like this. Of course
                the engine can do it for you. Waghdude gave a good explanation of how it
                would work. Taking care of things like physics, lighting, collision etc.
                is pretty much the whole point of a game engine. Were you referring to
                something else?

                - David

                --
                ~David-Haley
                http://david.the-haleys.org
              • Brandon J. Van Every
                ... Don t ask me. I think it s a complete load of hooey. There s an AI Standards committee at the IGDA, and I think it s a joke. Give it another 20 years
                Message 7 of 21 , Apr 22, 2006
                • 0 Attachment
                  David Haley wrote:
                  >
                  > What would an 'AI chip' do? You could have it take care of search
                  > algorithms, I suppose, or depending on what your AI was doing, computer
                  > vision code, etc. But I'm curious what you have in mind when you speak
                  > of an 'AI chip'.
                  >
                  Don't ask me. I think it's a complete load of hooey. There's an "AI
                  Standards" committee at the IGDA, and I think it's a joke. Give it
                  another 20 years and maybe something will be sufficiently repetitive in
                  games that lotsa people can agree on how to do it. Right now, I think
                  AI is way too flexible to commit to hardware. People make these false
                  analogies to 3D graphics HW all the time. They fail to notice the long
                  history of 3D HW, how long 3D has been a canned problem, etc. They just
                  think they can put it on HW because they want to; it's silly.

                  > physics are highly
                  > important for decorative/immersive purposes and that should _never_ be
                  > underestimated.
                  >
                  Your other points may be great / fine. Here, I think some debunking is
                  in order. First off, we've gotten this far in game industry history
                  without it. That should tell you something. Second off, walk out your
                  front door. The physics is all there, and it's dead boring. Don't
                  forget about game design, it's more important than physics.

                  I see physics as a specialty area that certain studios could grow and
                  make money in. We're a ways off before it becomes some kind of checkbox
                  requirement for all games. Even good art assets aren't 100% to that
                  point, and they're proven bank.

                  >
                  >> I don't see how an engine is going to do physics for you, any more than
                  >> it's going to make a game design for you. It could certainly streamline
                  >> the amount of work you have to do to put physics into your game, though.
                  >>
                  >>
                  >
                  > I'm a little surprised to see you make a statement like this. Of course
                  > the engine can do it for you. Waghdude gave a good explanation of how it
                  > would work. Taking care of things like physics, lighting, collision etc.
                  > is pretty much the whole point of a game engine. Were you referring to
                  > something else?
                  >
                  You still gotta design the game and figure out why the physics is in
                  there. Is your game design going to be driven by what your physics
                  engine can do? That's the tail wagging the dog. Depending on what
                  you're trying to achieve, the physics could introduce a lot of
                  programming complications that don't really help you with anything
                  important. If your physics API requires you to program to a 100
                  parameter interface, then it's not doing a lot for you, it's creating
                  work for you. So the real question is to what degree a physics engine
                  can offer "canned solutions," vs. how much it's just another general
                  purpose API layer that you have to write a whole bunch of stuff on top
                  of. Is it a value add, or is it a gratuitous support burden, driven by
                  a perceived need for realism?

                  Similarly, a 3D engine is not a game engine. You end up writing all
                  that stuff on top of the game engine. Or, some companies do offer game
                  engine middleware that does get you farther towards an actual game. But
                  those are typically genre specific, i.e. a MMOG engine.


                  Cheers,
                  Brandon Van Every



                  [Non-text portions of this message have been removed]
                • Todd Zircher
                  ... I ve written a couple of tech demos using Newton Game Dynamics and physics programming does not work that way. It s not a free ride, but it is not a
                  Message 8 of 21 , Apr 23, 2006
                  • 0 Attachment
                    Brandon J. Van Every wrote:
                    >
                    > Depending on what you're trying to achieve, the physics could
                    > introduce a lot of programming complications that don't really
                    > help you with anything important. If your physics API requires
                    > you to program to a 100 parameter interface, then it's not doing
                    > a lot for you, it's creating work for you.

                    I've written a couple of tech demos using Newton Game Dynamics and
                    physics programming does not work that way. It's not a free ride,
                    but it is not a crushing burden either.

                    > So the real question is to what degree a physics engine can
                    > offer "canned solutions" vs. how much it's just another general
                    > purpose API layer that you have to write a whole bunch of stuff
                    > on top of. Is it a value add, or is it a gratuitous support
                    > burden, driven by a perceived need for realism?

                    Game physics is very much a canned solution just like using
                    triangles to represent faces on a model rather then solid geometry.
                    There are fundamental similarities that all physics models and
                    simulations have.

                    Adding physics is similar to setting up a 3D model's lighting
                    properties. You create the object, set up its physical properties,
                    apply materials (rough steel, flesh, water, slippery glass, etc.)
                    and then you simply call an update function each game loop to let
                    the physics library update the interactions in the world. It's
                    actually very easy once you wrap your brain around the creation of
                    materials and applying them. [Remember when you had to learn about
                    material lighting properties: ambient and difuse colors,
                    specularity, normals, and the like?]

                    There is one huge change though in the way you design your games.
                    Beyond initial creation of the models, you don't (or should not)
                    directly manipulate them. Instead, you apply forces to the model to
                    handle acceleration and rotation. That requires a good
                    understanding of how physics works in general and how it applies to
                    the library you are using. About the only exception to that
                    guideline is teleportation or applying special powers that defy the
                    laws of physics.

                    Something to remember about the physics cards, there is always a
                    software solution too. The card is pure gravey and just a selling
                    point for your game. IF you're going to have any physics at all,
                    you might as well add a little extra and leverage that to your
                    advantage.
                    --
                    TAZ
                  • David Haley
                    ... Haven t you been following which games have the most commercial success? Game design helps but fancy decorative/immersive graphics are nearly constantly
                    Message 9 of 21 , Apr 25, 2006
                    • 0 Attachment
                      On this day of 4-22-2006 11:18 PM, Brandon J. Van Every saw fit to scribe:
                      >> physics are highly
                      >> important for decorative/immersive purposes and that should _never_ be
                      >> underestimated.
                      >>
                      >>
                      > Your other points may be great / fine. Here, I think some debunking is
                      > in order. First off, we've gotten this far in game industry history
                      > without it. That should tell you something. Second off, walk out your
                      > front door. The physics is all there, and it's dead boring. Don't
                      > forget about game design, it's more important than physics.
                      >

                      Haven't you been following which games have the most commercial success?
                      Game design helps but fancy decorative/immersive graphics are nearly
                      constantly present in highly successful games.

                      When 3d graphics came along, we'd gotten along just fine without them
                      till then. When pixel shaders came along, we'd gotten along just fine
                      without them till then. When surround sound / EAX in games came along,
                      we'd gotten along just fine without them till then. Etc. The point is
                      not that physics is "necessary", anymore than *any* of those
                      technologies are "necessary"; the point is that it makes the experience
                      that much better for some people.

                      > I see physics as a specialty area that certain studios could grow and
                      > make money in. We're a ways off before it becomes some kind of checkbox
                      > requirement for all games. Even good art assets aren't 100% to that
                      > point, and they're proven bank.
                      >

                      As I mentioned, more and more FPS games are using physics for even just
                      ragdoll deaths of characters. Not to mention fancy explosions and
                      whatnot. And then you have car games and of course flight/space
                      simulators that use physics quite extensively for more than decoration.
                      I'm not sure it's quite that much of a niche as you make it out to be.


                      >>> I don't see how an engine is going to do physics for you, any more than
                      >>> it's going to make a game design for you. It could certainly streamline
                      >>> the amount of work you have to do to put physics into your game, though.
                      >>>
                      >>>
                      >>>
                      >> I'm a little surprised to see you make a statement like this. Of course
                      >> the engine can do it for you. Waghdude gave a good explanation of how it
                      >> would work. Taking care of things like physics, lighting, collision etc.
                      >> is pretty much the whole point of a game engine. Were you referring to
                      >> something else?
                      >>
                      >>
                      > You still gotta design the game and figure out why the physics is in
                      > there. Is your game design going to be driven by what your physics
                      > engine can do? That's the tail wagging the dog. Depending on what
                      > you're trying to achieve, the physics could introduce a lot of
                      > programming complications that don't really help you with anything
                      > important.

                      I think you're making an unfair comparison. It's not as if by throwing
                      in _any_ feature you don't have to design the game anymore. That being
                      said, sometimes your platform _does_ not drive but at least constrain
                      what your game design will have; that's a necessary feature. Designing
                      for the perfect, ideal and inexistent game platform isn't terribly
                      productive.

                      > If your physics API requires you to program to a 100
                      > parameter interface, then it's not doing a lot for you, it's creating
                      > work for you. So the real question is to what degree a physics engine
                      > can offer "canned solutions," vs. how much it's just another general
                      > purpose API layer that you have to write a whole bunch of stuff on top
                      > of. Is it a value add, or is it a gratuitous support burden, driven by
                      > a perceived need for realism?
                      >

                      Todd Zircher explained this well; physics engines don't work the way you
                      seem to think they do.

                      > Similarly, a 3D engine is not a game engine. You end up writing all
                      > that stuff on top of the game engine. Or, some companies do offer game
                      > engine middleware that does get you farther towards an actual game. But
                      > those are typically genre specific, i.e. a MMOG engine.
                      >

                      A 3D engine can be part of a game engine. It depends on how you define
                      "game engine". The Quake2 game engine, going back a bit, included
                      graphics with its "physics" (in quotes because they weren't proper
                      physics, just the physics of the world like collisions etc.) Lots of
                      games actually build a game engine on top of the 3d engine, not the
                      other way around. Look at 3D engines like Ogre3D (excellent product, by
                      the way); using its structure, you build the game engine on top.

                      You seem to have a particular meaning of "game engine" in mind; could
                      you expand on that a bit?

                      Cheers,
                      - David

                      --
                      ~David-Haley
                      http://david.the-haleys.org
                    • Brandon J. Van Every
                      ... Yes, and many of em don t have real physics. ... 2D is alive and well. 3D of course proved to be a growth market. ... Actually we re still continuing to,
                      Message 10 of 21 , Apr 25, 2006
                      • 0 Attachment
                        David Haley wrote:
                        > On this day of 4-22-2006 11:18 PM, Brandon J. Van Every saw fit to scribe:
                        >
                        >>> physics are highly
                        >>> important for decorative/immersive purposes and that should _never_ be
                        >>> underestimated.
                        >>>
                        >>>
                        >>>
                        >> Your other points may be great / fine. Here, I think some debunking is
                        >> in order. First off, we've gotten this far in game industry history
                        >> without it. That should tell you something. Second off, walk out your
                        >> front door. The physics is all there, and it's dead boring. Don't
                        >> forget about game design, it's more important than physics.
                        >>
                        >>
                        >
                        > Haven't you been following which games have the most commercial success?
                        > Game design helps but fancy decorative/immersive graphics are nearly
                        > constantly present in highly successful games.
                        >
                        Yes, and many of 'em don't have real physics.

                        > When 3d graphics came along, we'd gotten along just fine without them
                        > till then.
                        2D is alive and well. 3D of course proved to be a growth market.

                        > When pixel shaders came along, we'd gotten along just fine
                        > without them till then.
                        Actually we're still continuing to, to some degree.

                        > When surround sound / EAX in games came along,
                        > we'd gotten along just fine without them till then.
                        Here you're stretching. A lot of people have cheesy sound systems not
                        worthy of anything great. Sure it's a value add to some, but it's
                        hardly a checkbox requirement. Titles do get along without the world's
                        best audio, both in terms of technical and artistic quality.

                        > Etc. The point is
                        > not that physics is "necessary", anymore than *any* of those
                        > technologies are "necessary"; the point is that it makes the experience
                        > that much better for some people.
                        >
                        I'm not convinced, perhaps because I haven't played a game whose physics
                        really blew me away in terms of game design. What title do you think
                        showcases the be-all end-all of physics today?

                        >
                        >> I see physics as a specialty area that certain studios could grow and
                        >> make money in. We're a ways off before it becomes some kind of checkbox
                        >> requirement for all games. Even good art assets aren't 100% to that
                        >> point, and they're proven bank.
                        >>
                        >>
                        >
                        > As I mentioned, more and more FPS games are using physics for even just
                        > ragdoll deaths of characters. Not to mention fancy explosions and
                        > whatnot.
                        Yep, FPS are those specialists that spring to mind.

                        > And then you have car games
                        Which is really a UI problem rather than a physics problem. Some
                        physics middleware actually recognizes this, providing parameters to
                        make the game "driveable" rather than realistic.

                        > and of course flight/space
                        > simulators that use physics quite extensively for more than decoration.
                        > I'm not sure it's quite that much of a niche as you make it out to be.
                        >
                        Simulators are certainly niche. Actually I wonder if flight simulators
                        deal with the same physics as FPS. Ergo, will a physics card help?

                        >
                        >> Similarly, a 3D engine is not a game engine. You end up writing all
                        >> that stuff on top of the game engine. Or, some companies do offer game
                        >> engine middleware that does get you farther towards an actual game. But
                        >> those are typically genre specific, i.e. a MMOG engine.
                        >>
                        >>
                        >
                        > A 3D engine can be part of a game engine. It depends on how you define
                        > "game engine". The Quake2 game engine, going back a bit, included
                        > graphics with its "physics" (in quotes because they weren't proper
                        > physics, just the physics of the world like collisions etc.) Lots of
                        > games actually build a game engine on top of the 3d engine, not the
                        > other way around. Look at 3D engines like Ogre3D (excellent product, by
                        > the way); using its structure, you build the game engine on top.
                        >
                        > You seem to have a particular meaning of "game engine" in mind; could
                        > you expand on that a bit?
                        >
                        You have defined the industry standard meaning of "game engine," and
                        that's what I was talking about. Game engines are more inclusive than
                        3D engines and typically written on top of them.


                        Cheers,
                        Brandon Van Every



                        [Non-text portions of this message have been removed]
                      • David Haley
                        ... So? I didn t say they did. We were talking about the importance of decorative effects; you were saying that they weren t terribly important and I was
                        Message 11 of 21 , Apr 25, 2006
                        • 0 Attachment
                          On this day of 4-25-2006 11:45 AM, Brandon J. Van Every saw fit to scribe:
                          > David Haley wrote:
                          >
                          >> Haven't you been following which games have the most commercial success?
                          >> Game design helps but fancy decorative/immersive graphics are nearly
                          >> constantly present in highly successful games.
                          >>
                          >>
                          > Yes, and many of 'em don't have real physics.
                          >

                          So? I didn't say they did. We were talking about the importance of
                          decorative effects; you were saying that they weren't terribly important
                          and I was arguing that they were. Real, complete physics engines are
                          relatively new, so it's not surprising that the majority of games don't
                          have physics engines.


                          >> When 3d graphics came along, we'd gotten along just fine without them
                          >> till then.
                          >>
                          > 2D is alive and well. 3D of course proved to be a growth market.
                          >

                          How many modern games that are commercial successes are 2D? Calling 3D
                          merely a "growth market" seems to be quite an understatement.

                          >> When pixel shaders came along, we'd gotten along just fine
                          >> without them till then.
                          >>
                          > Actually we're still continuing to, to some degree.
                          >

                          Then why do all modern 3D games use them? Who is this mythical "we" you
                          keep talking about?

                          >> When surround sound / EAX in games came along,
                          >> we'd gotten along just fine without them till then.
                          >>
                          > Here you're stretching. A lot of people have cheesy sound systems not
                          > worthy of anything great. Sure it's a value add to some, but it's
                          > hardly a checkbox requirement. Titles do get along without the world's
                          > best audio, both in terms of technical and artistic quality.
                          >

                          That wasn't the point, though. The point is that they add value to a
                          title, for those who have the sound card.

                          >> Etc. The point is
                          >> not that physics is "necessary", anymore than *any* of those
                          >> technologies are "necessary"; the point is that it makes the experience
                          >> that much better for some people.
                          >>
                          >>
                          > I'm not convinced, perhaps because I haven't played a game whose physics
                          > really blew me away in terms of game design. What title do you think
                          > showcases the be-all end-all of physics today?
                          >

                          As I previously mentioned, the Oblivion physics are very impressive. I'm
                          not sure if I'd call it the "be-all end-all" of physics, but download
                          the preview videos and see for yourself.

                          And you keep talking about it in terms of game design, yet I've been
                          claiming not that they are critical to game design but that they add to
                          the decorative, immersive environment of a game, and that environment is
                          important _in addition_ to the game design.

                          >> As I mentioned, more and more FPS games are using physics for even just
                          >> ragdoll deaths of characters. Not to mention fancy explosions and
                          >> whatnot.
                          >>
                          > Yep, FPS are those specialists that spring to mind.
                          >

                          Hardly a small group of "specialists", though. Calling them
                          "specialists" make it sound as if it's a relatively small group.

                          >> And then you have car games
                          >>
                          > Which is really a UI problem rather than a physics problem. Some
                          > physics middleware actually recognizes this, providing parameters to
                          > make the game "driveable" rather than realistic.
                          >

                          A UI problem?? A realistic car game needs to have realistic physics. The
                          UI problem is in addition to that.

                          >> and of course flight/space
                          >> simulators that use physics quite extensively for more than decoration.
                          >> I'm not sure it's quite that much of a niche as you make it out to be.
                          >>
                          >>
                          > Simulators are certainly niche. Actually I wonder if flight simulators
                          > deal with the same physics as FPS. Ergo, will a physics card help?
                          >

                          Simulators are niche, but my point was not based solely on simulators
                          but also on everything else preceding.

                          >>> Similarly, a 3D engine is not a game engine. You end up writing all
                          >>> that stuff on top of the game engine. Or, some companies do offer game
                          >>> engine middleware that does get you farther towards an actual game. But
                          >>> those are typically genre specific, i.e. a MMOG engine.
                          >>>
                          >>>
                          >>>
                          >> A 3D engine can be part of a game engine. It depends on how you define
                          >> "game engine". The Quake2 game engine, going back a bit, included
                          >> graphics with its "physics" (in quotes because they weren't proper
                          >> physics, just the physics of the world like collisions etc.) Lots of
                          >> games actually build a game engine on top of the 3d engine, not the
                          >> other way around. Look at 3D engines like Ogre3D (excellent product, by
                          >> the way); using its structure, you build the game engine on top.
                          >>
                          >> You seem to have a particular meaning of "game engine" in mind; could
                          >> you expand on that a bit?
                          >>
                          >>
                          > You have defined the industry standard meaning of "game engine," and
                          > that's what I was talking about. Game engines are more inclusive than
                          > 3D engines and typically written on top of them.
                          >

                          I didn't say that game engines are solely defined as 3D engines. Maybe
                          we're not agreeing on what these terms mean, which is why I asked you to
                          clarify your terms. When I say "3D engine", I don't mean anything
                          having to do necessarily with games. I mean something that takes 3d
                          models and their arrangements, and takes care of computing lighting,
                          shadows, reflections, bla bla bla. OpenInventor is an example of a 3D
                          engine that is often used for scientific visualization, not games.

                          A game engine is something used to drive a game's fundamentals and as
                          such a graphics engine is typically a part of it.


                          --
                          ~David-Haley
                          http://david.the-haleys.org
                        • Brandon J. Van Every
                          ... Again, we ve gotten this far without em. There is also the Uncanny Valley to reckon with. The more you push for realism, the more the human brain
                          Message 12 of 21 , Apr 25, 2006
                          • 0 Attachment
                            David Haley wrote:
                            > On this day of 4-25-2006 11:45 AM, Brandon J. Van Every saw fit to scribe:
                            >
                            >> David Haley wrote:
                            >>
                            >>
                            >>> Haven't you been following which games have the most commercial success?
                            >>> Game design helps but fancy decorative/immersive graphics are nearly
                            >>> constantly present in highly successful games.
                            >>>
                            >>>
                            >>>
                            >> Yes, and many of 'em don't have real physics.
                            >>
                            >>
                            >
                            > So? I didn't say they did. We were talking about the importance of
                            > decorative effects; you were saying that they weren't terribly important
                            > and I was arguing that they were. Real, complete physics engines are
                            > relatively new, so it's not surprising that the majority of games don't
                            > have physics engines.
                            >
                            Again, we've gotten this far without 'em. There is also the Uncanny
                            Valley to reckon with. The more you push for realism, the more the
                            human brain notices that something is wrong.

                            >
                            > How many modern games that are commercial successes are 2D? Calling 3D
                            > merely a "growth market" seems to be quite an understatement.
                            >

                            Well most of the casual games if you want to pick nits, but I was really
                            thinking about the number of times 2D games did in fact make the Top Ten
                            in recent years. Enough to debunk the idea that 2D games can't make
                            it. Maybe it's starting to fade now, I haven't really been keeping
                            track. Anyone wanna recommend a Top Ten site with a decent archive? I
                            seem to keep dancing around the problem on Google.


                            >>> When pixel shaders came along, we'd gotten along just fine
                            >>> without them till then.
                            >>>
                            >>>
                            >> Actually we're still continuing to, to some degree.
                            >>
                            >>
                            >
                            > Then why do all modern 3D games use them?
                            They don't, actually. But you did use the word "all," rather than a
                            qualifier like "to some degree."

                            >>> When surround sound / EAX in games came along,
                            >>> we'd gotten along just fine without them till then.
                            >>>
                            >>>
                            >> Here you're stretching. A lot of people have cheesy sound systems not
                            >> worthy of anything great. Sure it's a value add to some, but it's
                            >> hardly a checkbox requirement. Titles do get along without the world's
                            >> best audio, both in terms of technical and artistic quality.
                            >>
                            >>
                            >
                            > That wasn't the point, though. The point is that they add value to a
                            > title, for those who have the sound card.
                            >
                            Sure it's the point. Most titles aren't going to need the best physics
                            either. Nor the best AI.

                            >
                            >> I'm not convinced, perhaps because I haven't played a game whose physics
                            >> really blew me away in terms of game design. What title do you think
                            >> showcases the be-all end-all of physics today?
                            >>
                            >>
                            >
                            > As I previously mentioned, the Oblivion physics are very impressive. I'm
                            > not sure if I'd call it the "be-all end-all" of physics, but download
                            > the preview videos and see for yourself.
                            >
                            It seems Oblivion has its physics detractors. Recall the Uncanny Valley.
                            http://games.slashdot.org/article.pl?sid=06/04/17/1437229&from=rss
                            Lotta game design discussion in that thread. Well, at least if people
                            do get all hot and bothered in the future about physics in games, the
                            game designers can do something with the effects. On that point, some
                            people think Oblivion didn't:
                            http://www.joystiq.com/2006/04/19/oblivions-missing-physics-acceleration/
                            Which speaks to my earlier point, that physics accelerators don't
                            magically do the work of game design for you.

                            > And you keep talking about it in terms of game design, yet I've been
                            > claiming not that they are critical to game design but that they add to
                            > the decorative, immersive environment of a game, and that environment is
                            > important _in addition_ to the game design.
                            >
                            I'm not convinced it's that important. I think once you've achieved a
                            certain level of production values, dominated primarily by visual stuff,
                            you get diminishing returns.

                            >>> As I mentioned, more and more FPS games are using physics for even just
                            >>> ragdoll deaths of characters. Not to mention fancy explosions and
                            >>> whatnot.
                            >>>
                            >>>
                            >> Yep, FPS are those specialists that spring to mind.
                            >>
                            >>
                            >
                            > Hardly a small group of "specialists", though. Calling them
                            > "specialists" make it sound as if it's a relatively small group.
                            >
                            Well you know there's a lot of ways to use a FPS engine. They don't
                            have to all be Half-Life 2.

                            >
                            >>> And then you have car games
                            >>>
                            >>>
                            >> Which is really a UI problem rather than a physics problem. Some
                            >> physics middleware actually recognizes this, providing parameters to
                            >> make the game "driveable" rather than realistic.
                            >>
                            >>
                            >
                            > A UI problem?? A realistic car game needs to have realistic physics. The
                            > UI problem is in addition to that.
                            >
                            It doesn't need realistic physics. It needs a steering interface you
                            can bloody control.

                            >
                            > I didn't say that game engines are solely defined as 3D engines. Maybe
                            > we're not agreeing on what these terms mean,
                            Maybe we are talking past each other to no purpose. I don't see why you
                            say there's a disagreement.


                            Cheers,
                            Brandon Van Every


                            [Non-text portions of this message have been removed]
                          • Wayne Imlach
                            While Oblivion doesn t require the use of advanced physics accelerators, the physics it does implement are put to use in primarily two ways - ragdoll for
                            Message 13 of 21 , Apr 25, 2006
                            • 0 Attachment
                              While Oblivion doesn't require the use of advanced physics accelerators, the
                              physics it does implement are put to use in primarily two ways - ragdoll for
                              opponent deaths (which primarily increases immersion) and for physical
                              'traps' that can be sprung on hapless victims (including enemy creatures).
                              The traps are pretty good, as the player can dodge or avoid the various
                              swinging spiked balls, rolling logs and tumbling rocks if they are alert and
                              quick enough on thier feet (or should I say joypad). Impacts from heavy
                              objects will knock the character off their feet, and in some cases over the
                              edge of a precipice, with appropriate damage from both the initial hit and
                              the subsequent fall.

                              It's not perfect by any means, but still being put to good use as a game
                              mechanic as well as immersive component.
                            • Brandon J. Van Every
                              ... Such traps don t require realistic physics to implement though. It ll be interesting to see if any physics packages take hold as standards, and whether
                              Message 14 of 21 , Apr 25, 2006
                              • 0 Attachment
                                Wayne Imlach wrote:
                                > While Oblivion doesn't require the use of advanced physics accelerators, the
                                > physics it does implement are put to use in primarily two ways - ragdoll for
                                > opponent deaths (which primarily increases immersion) and for physical
                                > 'traps' that can be sprung on hapless victims (including enemy creatures).
                                > The traps are pretty good, as the player can dodge or avoid the various
                                > swinging spiked balls, rolling logs and tumbling rocks if they are alert and
                                > quick enough on thier feet (or should I say joypad). Impacts from heavy
                                > objects will knock the character off their feet, and in some cases over the
                                > edge of a precipice, with appropriate damage from both the initial hit and
                                > the subsequent fall.
                                >
                                > It's not perfect by any means, but still being put to good use as a game
                                > mechanic as well as immersive component.
                                >
                                >
                                Such traps don't require realistic physics to implement though. It'll
                                be interesting to see if any physics packages take hold as standards,
                                and whether players will come to expect a specific timing from physical
                                events based on a package. Or, it could remain entirely within the
                                realm of game design. "The trap sprung too fast, I was robbed!"


                                Cheers,
                                Brandon Van Every
                              • David Haley
                                ... How could you do this kind of very physical trap without some kind of physics simulation? Have you seen the kind of traps the game has? They re all very
                                Message 15 of 21 , Apr 25, 2006
                                • 0 Attachment
                                  On this day of 4-25-2006 7:00 PM, Brandon J. Van Every saw fit to scribe:
                                  > Wayne Imlach wrote:
                                  >
                                  >> While Oblivion doesn't require the use of advanced physics accelerators, the
                                  >> physics it does implement are put to use in primarily two ways - ragdoll for
                                  >> opponent deaths (which primarily increases immersion) and for physical
                                  >> 'traps' that can be sprung on hapless victims (including enemy creatures).
                                  >> The traps are pretty good, as the player can dodge or avoid the various
                                  >> swinging spiked balls, rolling logs and tumbling rocks if they are alert and
                                  >> quick enough on thier feet (or should I say joypad). Impacts from heavy
                                  >> objects will knock the character off their feet, and in some cases over the
                                  >> edge of a precipice, with appropriate damage from both the initial hit and
                                  >> the subsequent fall.
                                  >>
                                  >> It's not perfect by any means, but still being put to good use as a game
                                  >> mechanic as well as immersive component.
                                  >>
                                  >>
                                  >>
                                  > Such traps don't require realistic physics to implement though. It'll
                                  > be interesting to see if any physics packages take hold as standards,
                                  > and whether players will come to expect a specific timing from physical
                                  > events based on a package. Or, it could remain entirely within the
                                  > realm of game design. "The trap sprung too fast, I was robbed!"
                                  >

                                  How could you do this kind of very physical trap without some kind of
                                  physics simulation? Have you seen the kind of traps the game has?
                                  They're all very involved physically; you can also do things like throw
                                  objects to distract monsters and that also uses physics to properly
                                  model the trajectory.

                                  I'm not sure how you'd do all this without some kind of physics engine.

                                  --
                                  ~David-Haley
                                  http://david.the-haleys.org
                                • Brandon J. Van Every
                                  ... How physical is it to notice that something is swinging towards you and you need to dodge? Not terribly. Platformers have been doing this kind of thing
                                  Message 16 of 21 , Apr 25, 2006
                                  • 0 Attachment
                                    David Haley wrote:
                                    > On this day of 4-25-2006 7:00 PM, Brandon J. Van Every saw fit to scribe:
                                    >
                                    >> Such traps don't require realistic physics to implement though. It'll
                                    >> be interesting to see if any physics packages take hold as standards,
                                    >> and whether players will come to expect a specific timing from physical
                                    >> events based on a package. Or, it could remain entirely within the
                                    >> realm of game design. "The trap sprung too fast, I was robbed!"
                                    >>
                                    >>
                                    >
                                    > How could you do this kind of very physical trap without some kind of
                                    > physics simulation?
                                    How physical is it to notice that something is swinging towards you and
                                    you need to dodge? Not terribly. Platformers have been doing this kind
                                    of thing forever, no physics required.

                                    > Have you seen the kind of traps the game has?
                                    >
                                    No I haven't.

                                    > They're all very involved physically;
                                    Well I'll see if those videos show me anything more, but the examples
                                    supplied so far do not sway me.

                                    > you can also do things like throw
                                    > objects to distract monsters and that also uses physics to properly
                                    > model the trajectory.
                                    >
                                    I know you don't need physics for that. Thief did it just fine
                                    without. It was a core mechanic. All you need for throwing stuff is
                                    just for it to bounce off of surfaces somehow. Totally irrelevant if it
                                    bounces realistically, it just has to bounce. If it's difficult to
                                    predict the accuracy of where it lands, no big deal, give the player 5
                                    grenades instead of 2. Actually, when a quarter rolls away from me in
                                    real life, it's very hard to predict where it will land. I've learned
                                    to look a lot farther away than where I think it should be, because they
                                    travel. So I don't see any inherent "consistency of interface" value
                                    for physics here. The unrealistic interfaces are more consistent and
                                    perfectly adequate in game design terms.


                                    Cheers,
                                    Brandon Van Every



                                    [Non-text portions of this message have been removed]
                                  • David Haley
                                    ... Then I think that s why you don t appreciate what these traps do and why physics simulation is important for them. It s a lot more than the mere swinging
                                    Message 17 of 21 , Apr 25, 2006
                                    • 0 Attachment
                                      On this day of 4-25-2006 7:38 PM, Brandon J. Van Every saw fit to scribe:
                                      > David Haley wrote:
                                      >
                                      >> On this day of 4-25-2006 7:00 PM, Brandon J. Van Every saw fit to scribe:
                                      >>
                                      >>
                                      >>> Such traps don't require realistic physics to implement though. It'll
                                      >>> be interesting to see if any physics packages take hold as standards,
                                      >>> and whether players will come to expect a specific timing from physical
                                      >>> events based on a package. Or, it could remain entirely within the
                                      >>> realm of game design. "The trap sprung too fast, I was robbed!"
                                      >>>
                                      >>>
                                      >>>
                                      >> How could you do this kind of very physical trap without some kind of
                                      >> physics simulation?
                                      >>
                                      > How physical is it to notice that something is swinging towards you and
                                      > you need to dodge? Not terribly. Platformers have been doing this kind
                                      > of thing forever, no physics required.
                                      >
                                      >
                                      >> Have you seen the kind of traps the game has?
                                      >>
                                      >>
                                      > No I haven't.
                                      >

                                      Then I think that's why you don't appreciate what these traps do and why
                                      physics simulation is important for them. It's a lot more than the mere
                                      swinging of a pendulum.


                                      >> you can also do things like throw
                                      >> objects to distract monsters and that also uses physics to properly
                                      >> model the trajectory.
                                      >>
                                      >>
                                      > I know you don't need physics for that. Thief did it just fine
                                      > without. It was a core mechanic. All you need for throwing stuff is
                                      > just for it to bounce off of surfaces somehow. Totally irrelevant if it
                                      > bounces realistically, it just has to bounce. If it's difficult to
                                      > predict the accuracy of where it lands, no big deal, give the player 5
                                      > grenades instead of 2. Actually, when a quarter rolls away from me in
                                      > real life, it's very hard to predict where it will land. I've learned
                                      > to look a lot farther away than where I think it should be, because they
                                      > travel. So I don't see any inherent "consistency of interface" value
                                      > for physics here. The unrealistic interfaces are more consistent and
                                      > perfectly adequate in game design terms.
                                      >

                                      More consistent? A physics engine is quite consistent. Not sure what you
                                      mean, unless you're using consistent to mean predictable.

                                      I'm not sure why you're showing such gut opposition to using physics in
                                      a game. If physics can solve the problem of "making stuff bounce", in
                                      addition to driving rag-doll effects, particle explosions, arrow
                                      trajectories, bla bla bla, then it seems useful. Physics *chips* might
                                      not be terribly useful at-this-point-right-now, but that doesn't mean
                                      that physics in general is a waste of time. If it were, then there
                                      wouldn't be such effort going in to projects like ODE.

                                      You might be interested in this:
                                      http://www.ode.org/users.html

                                      It lists some products that use the ODE physics engine. Note that there
                                      are several kinds of car games (and not just "UI problems"),
                                      first-person shooters, block-playing games, flight/space simulation,
                                      tools for designing animation, and much more.

                                      --
                                      ~David-Haley
                                      http://david.the-haleys.org
                                    • David Haley
                                      ... To elaborate slightly, the physics used in car games are not just for driving the car. They re used in rally-type games, where physics control such things
                                      Message 18 of 21 , Apr 25, 2006
                                      • 0 Attachment
                                        On this day of 4-25-2006 7:56 PM, David Haley saw fit to scribe:
                                        > It lists some products that use the ODE physics engine. Note that there
                                        > are several kinds of car games (and not just "UI problems"),
                                        >

                                        To elaborate slightly, the physics used in car games are not just for
                                        driving the car. They're used in rally-type games, where physics control
                                        such things as how the car collides, bounces, falls, goes off of ramps,
                                        etc. It's a lot more than just driving the car (steering, etc.); it's
                                        the entire environment that is simulated using the physics engine. In
                                        fact, some of these games basically *depend* on having physics in order
                                        to make the game even slightly convincing.

                                        --
                                        ~David-Haley
                                        http://david.the-haleys.org
                                      • Brandon J. Van Every
                                        ... I disagree. You adjusted a mouse for constant linear vs. accelerated motion lately? A physics engine has potentially the same problem. On the positive
                                        Message 19 of 21 , Apr 25, 2006
                                        • 0 Attachment
                                          David Haley wrote:
                                          >
                                          > More consistent? A physics engine is quite consistent.
                                          I disagree. You adjusted a mouse for constant linear vs. accelerated
                                          motion lately? A physics engine has potentially the same problem. On
                                          the positive side, at least there's something for a designer to do.

                                          >
                                          > I'm not sure why you're showing such gut opposition to using physics in
                                          > a game.
                                          Because it's API churn. In a lot of cases it doesn't add value in terms
                                          of game design. It's the same reason that a number of game designers
                                          object to overweening 3D production values. I'm not interested in games
                                          that require a bazillion new technologies so that the budgets are
                                          multi-million and only big publishers have the bucks for it and
                                          therefore they make exceedingly conservative business decisions because
                                          there's too much money on the table. I want game designers to get
                                          better at shipping original games, with more creative control for the
                                          designers.

                                          > If physics can solve the problem of "making stuff bounce",
                                          The point is it's not a problem.

                                          > in addition to driving rag-doll effects,
                                          Ragdoll is a somewhat cool visual. But that's all it is. Game
                                          mechanically it doesn't mean anything.

                                          > particle explosions,
                                          Systems with medium (100) to large (1000) numbers of moving entities are
                                          interesting, but they don't need physics to be interesting. Actually a
                                          big game design question is what you're going to do with all that
                                          onscreen visual complexity, if it's meant to be more than illustration.
                                          How do you control 1000 particles as a player?

                                          > arrow trajectories,
                                          There's nothing wrong with the bow in Thief 1. Normal arrows drop in
                                          flight. Magic arrows travel in an exact straight line, giving an aiming
                                          advantage. You have to pull your bowstring back to get enough power for
                                          your shot. If you release too soon, you have no power. If you hold too
                                          long, your hand starts shaking and then you can't hold it any longer.
                                          None of this is realistic physics. It's all just UI and simple game
                                          pseudo-physics. I doubt there's anything more complicated for the arrow
                                          than a high school algebra trajectory, if that.

                                          > bla bla bla, then it seems useful. Physics *chips* might
                                          > not be terribly useful at-this-point-right-now, but that doesn't mean
                                          > that physics in general is a waste of time. If it were, then there
                                          > wouldn't be such effort going in to projects like ODE.
                                          >
                                          Flawed reasoning. Open source people put effort into everything. A lot
                                          of technologists just think "Physics is kewl! Physics is really gonna
                                          buy me something!" They don't think like game designers. There's piles
                                          and piles of technologists out there who spend their careers churning
                                          out technology. It turns them on.

                                          > You might be interested in this:
                                          > http://www.ode.org/users.html
                                          >
                                          > It lists some products that use the ODE physics engine. Note that there
                                          > are several kinds of car games (and not just "UI problems"),
                                          > first-person shooters, block-playing games, flight/space simulation,
                                          > tools for designing animation, and much more.
                                          >
                                          >
                                          Now of all these, which would you point out as compelling games? As
                                          opposed to physics tech demos? That's the rub. You show me ODE as the
                                          basis of a physics game, and I really like that game "because of the
                                          physics," then I'll change my tune.


                                          Cheers,
                                          Brandon Van Every




                                          [Non-text portions of this message have been removed]
                                        • Brandon J. Van Every
                                          ... I m not interested in convincing. I m interested in amusing. Realism is not the be-all end-all of game design. So if you can point out a truly amusing
                                          Message 20 of 21 , Apr 25, 2006
                                          • 0 Attachment
                                            David Haley wrote:
                                            > On this day of 4-25-2006 7:56 PM, David Haley saw fit to scribe:
                                            >
                                            >> It lists some products that use the ODE physics engine. Note that there
                                            >> are several kinds of car games (and not just "UI problems"),
                                            >>
                                            >>
                                            >
                                            > To elaborate slightly, the physics used in car games are not just for
                                            > driving the car. They're used in rally-type games, where physics control
                                            > such things as how the car collides, bounces, falls, goes off of ramps,
                                            > etc. It's a lot more than just driving the car (steering, etc.); it's
                                            > the entire environment that is simulated using the physics engine. In
                                            > fact, some of these games basically *depend* on having physics in order
                                            > to make the game even slightly convincing.
                                            >
                                            >
                                            I'm not interested in convincing. I'm interested in amusing. Realism
                                            is not the be-all end-all of game design. So if you can point out a
                                            truly amusing ODE car game that has a demo, I'll look at it.


                                            Cheers,
                                            Brandon Van Every



                                            [Non-text portions of this message have been removed]
                                          • Joel Davis
                                            ... Here s one. Truck Dismount (Not ODE but same idea.): http://jet.ro/dismount/ You can t tell me that game is not amusing. And it s certainly not realistic.
                                            Message 21 of 21 , Apr 25, 2006
                                            • 0 Attachment
                                              > I'm not interested in convincing. I'm interested in amusing. Realism
                                              > is not the be-all end-all of game design. So if you can point out a
                                              > truly amusing ODE car game that has a demo, I'll look at it.

                                              Here's one. Truck Dismount (Not ODE but same idea.):
                                              http://jet.ro/dismount/

                                              You can't tell me that game is not amusing. And it's certainly not realistic.

                                              I think what you're trying to say is that physics engines have no place in a 4X turn based strategy game, so you're not interested in them. Which is fair enough. But to say that they have no purpose in FPS or RPG gameplay ignores Half-Life2 and many games that has come since (including Oblivion). Maybe some of the traps in HL2 or Oblivion could be scripted, but they wouldn't feel the same, and many of the traps couldn't be scripted. I'd say the Gravity Gun is a pretty central gameplay mechanic to HL2, and that would be absolutely impossible without a robust physics engine. And it's a really, really fun gameplay mechanic.

                                              joel

                                              ----- Original Message ----
                                              From: Brandon J. Van Every <bvanevery@...>
                                              To: gamedesign-l@yahoogroups.com
                                              Sent: Tuesday, April 25, 2006 9:35:13 PM
                                              Subject: Re: [gamedesign-l] Re: Future games, architectures and acceleration
                                              I'm not interested in convincing. I'm interested in amusing. Realism
                                              is not the be-all end-all of game design. So if you can point out a
                                              truly amusing ODE car game that has a demo, I'll look at it.

                                              [Non-text portions of this message have been removed]
                                            Your message has been successfully submitted and would be delivered to recipients shortly.