I Want Every Game To Be Made In The Sui Generis Engine

Yeah, and they're terrible at screenshots.

I’m not sure how that will work for driving games or football managers, but people can work around it. I was being very impressed by the lovely-looking Miasmata earlier, there with an engine written by just one guy. And the same is true for Sui Generis. But this engine looks like it might be the sort of thing licensed by anyone wanting to make an RPG in the near future. Using technological magics, Madoc Evans has created a physical system that allows 3D objects to automatically behave according to their shape – complex objects can be entangled with each other seemingly without clipping, or revealing invisible limits. Then with dynamic lighting, on-the-fly terrain modelling and procedural generation, there’s a combat system of the likes I’ve never seen.

Of course, I’m no pioneer. Adam mentioned this one in last weekend’s Katchup, but you try finally getting around to watching the video, having a website, and not posting about it. You can see the Kickstarter here.

There are five people in the team – Bare Mettle Entertainment – with a big project in front of them. And their Kickstarter certainly isn’t racing toward its big goal of £150,000. After over a week they’ve etched out nearly £20k, and it will take a significant amount more attention for this to get funded in time. However, as I watch this I do wonder if they might not find themselves making some money either way. What’s not adequately shown in the video is the game they want to make. We can see how they want to make it, and it’s damned, damned impressive. But it’s unclear what vision and passion for an RPG tale they have. That might not be helping them here. But if this technology is easily used by someone else, then surely – SURELY – a big name RPG maker is going to want this tech?

However, they clearly do have a game in mind, and that’s been appearing in their updates. Although it doesn’t sound, perhaps, all too original. You’re some ordinary joe, exiled from your village into the mysterious world, where you find you’re not so ordinary after all. There’s a big bad evil escalating, and you’re the only one with the power to stop it. So it’s an RPG then. They promise the story will be dynamic, with multiple outcomes, and… well, yeah. That’s of concern, I think – that in response to requests for details, we have instead had the most generic outline for all RPGs that have ever existed. It doesn’t mean they don’t have a brilliant story to tell, of course. But that they’ve not been able to articulate it at this point isn’t a strong point.

There’s some “pre-alpha” footage to be seen here, but again it only shows the tech looking amazing, rather than anything that you could identify as of their game.

But dammit, I want to play a game in this technology. Whether it’s theirs, or whether an Obsidian or a CD Projekt comes along and says, “Here, have all our money,” I’m not too fussed at this point. I just want to throw tables at benches, and have sword fights with that skellington.


  1. Brosepholis says:

    You people are astonishingly easy to impress!

    ‘Accurate collision detection’: It’s mesh-mesh! Or convex decomposition! Either way, seriously not rocket science! Good luck getting any sort of performance with complex meshes, assuming you’re going to have any!
    ‘Realtime lighting’: Whoop de doo, just like every other game then!
    ‘Physics based animation’: Yeah, it looks exactly like Sumotori Dreams. Wanna know why? Because that’s all you can get out of simple procedural muscle control without cheating. And these guys are certainly cheating.
    ‘Procedural terrain’: Commercial games don’t use it because it creates boring envionments, not because of an activision conspiracy.
    ‘Weather generation system’: If this is the sort of feature creep this guy can get up to without any funding, imagine the irrelevant stuff he’ll add once he takes your money!

    All in all, it’s a guy who can program, and a few of his buddies who can sort-of half do some other stuff. Did you listen to the music? Or actually look closely at the models? That’s the sort of quality you’re going to be paying for.

    • tro says:

      what does that mean when u say cheating? i don’t know much technically about games and making them, so genuine question.
      from context, i’m understanding u mean that there are probably extra factors coded in to guide behaviours and outcomes, outside of the physics/muscle engine.
      is that what u meant? cuz it seems like u would want that in a game. to reflect various skill levels at different things. or is the cheating more complicated, or would affect actual gameplay in ways that i don’t understand?

      i think u are being a little harsh with them tho, the video doesn’t seem to be claiming that this is gonna be the salvation of RPG’s or anything. it seems like he’s more trying to show the mindset that generated the engine he’s been working on. which, to me comes off as trying to enable the creation of an open world, where u can set certain restraints or important pieces and then the engine will take care of how those begin to interact with each other. from reading the other material on their page too, they seem to be guided by a desire to reclaim a time, mostly in pen and paper, when RPG’s were more small scale, and handled their fantasy elements ‘realistically’. in that danger was more real and combat and consequences were a bigger deal. u didn’t storm some cave and kill 100 orcs and 10 minotaurs and 13 beholders and get 5 magic swords, 32,000 GP and 2 bags of holding. you know?
      it’s not a new idea, but the thought that a bad ass fight should be 5 thugs, a couple dogs and an evil priest. which personally, i think works better in a turn based combat system where tactics are emphasized. but again, no one in the video seems to be claiming that they are reinventing any wheels and i think it’s unfair to be so cynical about what they’ve come up with so far.
      i really do think the video is very impressive, but i think like many other people who also do, we are imaging what kind of rad things we could do if we were allowed to control this engine with mods or something. i don’t understand the implications of things u mentioned like mesh/mesh convex decomposing over my models, but if this stuff is so easy and irrelevant, why haven’t we seen it more?
      isn’t there anything that strikes u as pretty wicked about that giant whacking that guy with his ball and chain?

      • DeathCarrot says:

        I think it was more of a critique on the article’s enthusiasm. The video concentrates on the engine because that’s all there is at the moment, and that’s fine, just the article makes it out to be something it’s not.

        • MordeaniisChaos says:

          And I would heartily agree with the original critique in question. This is mostly stuff that is done already, or is done very poorly here. Accurate and dynamic indirect lighting in real time is about the only thing impressing me, and I’m surprised other isometric or similarly constrained 3D environments haven’t pushed for such lighting features.

      • Brosepholis says:

        When I say they’re cheating, I mean that they can’t be holding the character up entirely with its own muscles, because that’s essentially impossible to do in real time (and is a current research topic for a lot of people). They’re cheating by using external forces to keep the characters held up, probably applied at the hip joint, which is why the hips appear to stay at the same height all the time. That’s also the reason they look like string-puppets, because injecting external forces is pretty much the same as using invisible strings to control your character.

        Mesh-Mesh collision detection just tests all the triangles in one mesh against the triangles in the other mesh, which works perfectly well but gets really slow in any real game environment, which is why it’s never used in commercial games (to my knowledge). Their problem is, by doing this they are linking the complexity of their collision representation to the fidelity of their graphics, which is a very bad place to go.

        My comment was mostly expressing disbelief at the RPS Hivemind’s credulity when it comes to guys saying their tech is world-changing. The exact same thing happened with the Unlimited Detail business (Kickstarter incoming in 5…4…3…). At least get an actual game dev to verify these guys’ claims, there are enough of us on here.

        As for the game, this engine would work fine for a simple physics puzzle/platformer, but these guys want to make a rich procedurally generated RPG without having any actual artists. I wish them the best of luck, but if/when a finished game crawls out of this mess, it’ll be a great deal less ambitious than this kickstarter leads you to believe.

        • Madoc says:

          You’re almost right about the characters, of course we “cheat”, not as crudely you describe but we definitely cheat. We are animating characters by simulated muscular contractions and they remain very reactive to external forces but we don’t claim they can stand on their own feet as such, the balancer has artificial elements. We do hope to improve on the quality of both the balancer and the animations (which are not entirely procedrual and basically just terrible because we don’t have an animator, yet).

          Regarding the collisions you’re essentially wrong. We’re not limiting ourselves in any way, we can make our models as graphically complex as can be expected in any real time game without any perfomance issues for collision detection. Fact.

          As for lighting it is far more dynamic than the vast majority of engines that feature advanced lighting. Fact. And our lighting is advanced, our material system is definitely different from those of any current major engine and stupidly feature rich.

          We say we have procedural methods for terrain. It should be rather evident that the focus here is hand sculpting. Either way this is an iso game, terrain is generally not a prominent feature in such games.

          My only response to the sky comment is that we won’t be featuring any trolls in our game. You have absolutely no idea why I implemented such a feature in the first place. I do this for a living.

          Fortunately you’re not even aware what other technologies this “guy who can program” has developed. I would be interested to see what you can do in comparison seeing as you present yourself as such an authority.

        • tro says:

          thanks for the response, i think i understand more where u were coming from. it was obvious u were coming from a point of view inside the game developing industry, but that didn’t register with me for some reason. i’m personally pretty naive about technical details regarding all that, so, to me, it all looks pretty amazing and exciting. but if i think about stuff that i do know quite a bit about, i admit i also can find myself feeling similarly sometimes. and i understand now more that your comment was not directed at the game or it’s developers as much as towards a certain perceived culture maybe reflected in the comments thread or even the article itself.
          anyways, i still think it looks rad, but thanks for explaining the things i was wondering about.
          for someone like me tho, i don’t see the character’s as being particularly clunky in a bad way, or really notice things like lighting or even the supposed levels of details in graphics textures. this looks great to me, if i was the boss of it, i’d say ‘graphics look great, glad that’s done, let’s work on the game now.’ i’m looking at any game like this as a way to make an RPG playable on your computer. and not as a game, exactly. you know what i mean? i think that might have been a really terrible explanation.
          i think a lot of the market for a game like this, tho, is people similar to me, that maybe don’t care so much about how modern a thing may look, as long as it better captures what we think of as an RPG.

  2. The Random One says:

    Fact: Having completely realistic body movement controlled by clicking a mouse makes you character fight like a spaz.

    In fact, between your character fighting like a spaz and all objects interacting physically, this looks like an excellent engine for a Penumbra style game.

  3. Takshaka says:

    Backed, I really don’t care if this guy’s game is any good or not I just want to give him my money.

    My jaw hasn’t dropped so much at a tech demo since I saw a demonstration of Myth: The Fallen Lords. In the video the dwarf unit threw a grenade that bounced off the swinging sword of another unit. The grenade flew back at the dwarf and blew up at his feet. The best part is the guys from Bungie demoing the game responded like “wow we didn’t know that was possible”. Keep in mind this demo was some time around 96/97.

    • ffordesoon says:

      A new Myth in this engine (assuming such large-scale battles are possible with that level of fidelity, which they probably aren’t) would be a dream I didn’t know I had come true.

      Or a Warhammer 40K game in the style of Myth. A bunch of Space Marine miniatures you can custom-paint in an editor come to life and sawing Orks in half, blood and bodyparts splattering everywhere in an orgy of brilliantly stupid violence. God, it would be glorious.

  4. jkz says:

    Looks really good, I might actually have to put my hand in my pocket for this. Yes there are no story details yet, but it is early in development and the possibilities to build an interesting world are good.

    Most fantasy games aren’t exactly good stories anyway *ducks*

  5. mandrill says:

    They shouldn’t be being so ambitious with their RPG concept. I would happily back something a little less grand, a Soul Calibur style fighting game maybe? That combat looks amazing.

  6. MordeaniisChaos says:

    This looks neat at best. The most interesting and applicable thing I saw in this whole thing was the dynamic indirect lighting. The animations look awful, and the combat looks clumsy. It looks like a really really really shitty fan rendition of GTA IV drunken stumbles. I wish people would stop trying to make games with “hyper realist purely physics based animations!” because they look so incredibly and LITERALLY robotic. Not like, fake C-3P0, but like actual robots, the ones that can kinda walk around on their own. Only clumsier still. Most of the rest of the engine and project looked pretty typical to me. I applaud the ambition, but don’t see a hell of a lot of actual potential.

    • jkz says:

      I thought the combat looked good, the weapons had weight and power. So many games aren’t realistic, it’s a change to have some that are. It seems different anyway.

  7. Droopy The Dog says:

    Welp, that looks like pretty compelling evidence that dual wielding + real physics = uncoordinated flailing.

    Seriously, that combat makes everyone look endearingly clumsy, it’s a refreshing change from all the physically impossible choreography.

  8. Synesthesia says:

    haha it reminds me of drunk walking on gta 4. Damn, that euphoria engine was good huh.

  9. veremor says:

    That’s kind of how I would have imagined Dragon Age (together with big cities and reactive world politics) after the first project descriptions.

  10. jryan says:

    Skip the story. Take this beautiful engine and turn it into an online F2P melee extravaganza. I want to fight 1×1, 2×2, 10×10 gladiatorial battles using this engine soooooo badly. All you need is the meta game where you spend yer killbucks on weapons, armor, traits and spells and let nature take it’s course.

    Once you have shown how amazingly awesome your engine is you can start licensing it to others and use the money to make the story game you always wanted.

  11. jryan says:

    OPINION #2: I see where a physics heavy engine like this will be hell on online play given lag and latency…. so use an online gaming service to it’s fullest with this game. Onlive and other services like that have opened a window to these physics heavy games to make some epic multiplayer games.

  12. Megakoresh says:

    This is geniously simple! Material-based collisions without physics boxes and muscle contractions? Why haven’t people thought of it already?

    The graphics is great as well. I wonder if it has DX11 feature support.

    I have to though, that the game is significantly let down by the fact that they chose the top-down perspective. Behind-the-shoulder view would benefit a lot, as it stands I can barely tell how realistic the movement and collisions are because the camera is so far away.

    If they consider adding a behind the shoulder view I am definitely going to support the game.