Search Unity

  1. Megacity Metro Demo now available. Download now.
    Dismiss Notice
  2. Unity support for visionOS is now available. Learn more in our blog post.
    Dismiss Notice

Large Scale RTS Battles

Discussion in 'Editor & General Support' started by GerhardMedia, Aug 18, 2013.

  1. GerhardMedia

    GerhardMedia

    Joined:
    Dec 20, 2012
    Posts:
    20
    Hello!
    I had a different kind of question about unity
    Can it handle large scale rts battles in one map like lets say i have 1500 soldiers in army i created and my friend has 2000

     
    blox5000 likes this.
  2. cannon

    cannon

    Joined:
    Jun 5, 2009
    Posts:
    751
    Yes, but then it's no longer a question of can Unity do it, but more how good you are at optimization as those numbers will bring any engine to its knees without a great deal of skill available.
     
    blox5000 likes this.
  3. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    My understanding is that Unity cannot handle high quantity simulations, at decent frame rates as it does not support:

    Hardware instancing, a way to reduce draw calls when lots of animated characters are used together or have native support for Imposters, a technique used where a 2d sprites are generated from models and used to represent characters in the far distance.
    Multithreading of gameObjects to allow the soldiers to be run controlled by separate cores and increase the simulations speed.

    Just try it for yourself with an army of cubes how many can you run on a terrain at 60fps.
     
    blox5000 likes this.
  4. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    Made a very quick test, using unity terrain, cubes with sphere + colliders + rigidbodies.

    Two armies of 1000 cubes marching towards each other and as you can see Unity gets 2.0 fps on my PC.

    $unityMarchingCubeArmyTest.jpg
     
    blox5000 likes this.
  5. brilliantgames

    brilliantgames

    Joined:
    Jan 7, 2012
    Posts:
    1,937
    Man. You wouldnt use rigid bodys for characters on a scale like that.. The best solution would be your own physics system. Unity has nothing to do with any of this.. An example of how you could easily get a large amount of characters on their feet running around would be to have each character do a raycast downward and position the character at the hitpoint of the raycast+the height of the character. But point is, of course unity doesn't have systems in place for large scale armys, Total War has been building their engine for this exact purpose for a very long time.

    But, nothing is impossible, when you take matter into your own hands, and don't rely on an engine to have built in features for every game type and scenario.
     
    blueivy, SparrowGS, gesicht and 2 others like this.
  6. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    On the plus side the same simulation with two armies of 500 or 1000 units total runs at 60fps, but once you start adding particles, models animations, ranged units, units of different types, and this number will drop if your aiming for 60fps.

    So the simple answer is No, Unity cannot handle > 1000 units moving over terrain using it's physics system.

    But as brilliantgames points out if you can write your own "physics" system then you could have more units moving over terrain.
     
  7. DallonF

    DallonF

    Joined:
    Nov 12, 2009
    Posts:
    620
    Don't feel the need to defend Unity on this. No, Unity can't handle 3500 units onscreen at once, because it's primarily meant for small developers who have the common sense to know that a massive-scale RTS isn't within their budget. Even if you were good enough at programming to optimize for this sort of thing, Unity would only be getting in your way and you'd want to roll a custom engine.
     
    blox5000 likes this.
  8. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    OK, Changed the test over to Kinematic Rigidbodies that follow the terrains height (terrain.SampleHeight(pos)) and have managed to get 2000 units marching into each other at 40-50 fps and still detecting collisions (in webplayer/pc build).

    And on a PC standalone build got 2 x 2000 units charging at each other at about 25 fps.

    I think there are still some optimisations that could be made to this, so it might be just about possible to hit 4000 units in combat at 30 fps, but you are going to have to do some serious optimisation work to get this type of game out of Unity.

    Link to 4000 units webplayer simulation
     
  9. jc_lvngstn

    jc_lvngstn

    Joined:
    Jul 19, 2006
    Posts:
    1,508
    Interesting. Just some thoughts...

    1. How often are you checking each unit to see if it needs to fight/interact with another nearby unit? Every frame, once every second, etc?
    2. I wonder if using Graphics.DrawMesh would be a little faster than letting Unity handle them as GameObjects. Probably would just remove the instantiate cost, but still might be worth looking at.
    3. I'll bet LOD groups would help a lot here.


    I guess I'm curious about how much goes into just drawing the units themselves, and how much goes into calculating their activity.
     
  10. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    Using just cubes, unity manages to get the whole scene rendered in 20 draw calls, the physics engine is checking for collisions and is running at the default physics update frequency.

    Looking at the profiler the scripts that march and terrain follow for each unit are called 4000 times each and take up a good chunk of the frame rate.

    So having a single march/terrain follow script or two could greatly reduce overheads.
     
  11. cannon

    cannon

    Joined:
    Jun 5, 2009
    Posts:
    751
    An RTS, especially a multiplayer on like in the OPs example has to be lockstep, you can't use the Physics subsystem at all for anything but client-side visuals.

    It's a given that you'll need to make design trade-offs to get those numbers:

    - flat terrain or a fast heightmap lookup so you don't do a raycast per unit to get correct terrain hugging
    - tons of atlased imposters
    - grouped AI so you treat globs of AI as one unit where feasible
    - per-unit AI will of course have only minimal intelligence
     
  12. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    Updated above weblplayer version hit's about 40fps (for me)
    Now has a central single update function to move and terrain follow the units.

    Link to 4000 units webplayer simulation

    Standalone version only manages 50fps.

    @Cannon - Yes I think a faster terrain lookup could boost this even more possibly taking it to 60fps.
     
    Last edited: Aug 19, 2013
  13. cannon

    cannon

    Joined:
    Jun 5, 2009
    Posts:
    751
    Yeah, it probably would take it past 60.
    There's also a couple of tricks too to get around the skinning issue for Shogun style games (tricks with swapped rigid meshes come to mind).

    An extreme upper bound for a Unity game is probably AI Wars, which goes up to 30,000 units at one time. There's a lot of fun trade-offs between quality and quantity that can be made to scale the game from 10 units to 30,000.
     
  14. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,614
    While I can share no details as it was a professional gig, not that long ago I wrote a simulation with 30,000+ agents running around which ran at 30+ FPS on an old Pentium 4 machine we had at the office.

    Like brilliantgames says, yes Unity can do it, you just don't want to use their off-the-shelf general-purpose solutions to do the whole thing - this kind of scale is not what those general purpose tools are for.
     
    GarBenjamin and zombiegorilla like this.
  15. KarelA

    KarelA

    Joined:
    Dec 30, 2008
    Posts:
    422
  16. DallonF

    DallonF

    Joined:
    Nov 12, 2009
    Posts:
    620
    An RTS doesn't have to be lockstep, it just so happens that most of them are. Planetary Annihilation is using an asynchronous client-server model with great success, especially for an alpha build.
     
  17. cannon

    cannon

    Joined:
    Jun 5, 2009
    Posts:
    751
    Yes, I've overgeneralized, that's true.

    PA is rather unique though. From what I've heard their design does impose limits on what they can do with gameplay, and I highly suspect that they still do run a deterministic Physics simulation (which Unity does not; not even trig functions are the same between platforms let alone PhysX) just to limit the amount of synchronization that they have to do. These are just my guesses though; I'm intensely curious as to how they do the synchronization with no unit cap.
     
  18. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    Had a play with using the array data from the terrain to set the unit heights but it did not easily follow the terrain, as using integers to look up the position and no interpolation between surrounding points. Units were slightly out of sync with the terrain. Climbing before the terrain appears and descending under the terrain.

    So in conclusion, out of the box, Unity's generic components do not currently allow for massive RTS battle simulations at 60fps.

    You would probably have to build your own engine to power an RTS type game, or there may be other engines that are built to provide the features you need.

    Or you could limit your unit counts to hundreds of units and make a reasonable RTS simulation system with Unity.

    If Unity's core engine was multi-threaded with load balancing would that improve things?
     
  19. Hikiko66

    Hikiko66

    Joined:
    May 5, 2013
    Posts:
    1,304
    If you want that kind of scale you probably wouldn't let a battle "play out" half as much as you'd orchestrate it.

    A troop doesn't fall after being hit by a stray bullet, the engine decides that troop is going to fall and then it fires the bullet at him.
    You wouldn't need to test for collisions if you're explicitly creating them.

    I'm speaking as a complete noob though.
     
    Last edited: Aug 19, 2013
  20. GibTreaty

    GibTreaty

    Joined:
    Aug 25, 2010
    Posts:
    792
  21. Gigiwoo

    Gigiwoo

    Joined:
    Mar 16, 2011
    Posts:
    2,981
    The webplayer ran a consistent 46 FPS for me. 4000 units, with physics. Well done arrow!
    Gigi
     
  22. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    In large scale historic battles where blocks of troops charge into each other you are probably better off with some kind of simple obstacle avoidance and flocking behaviour.

    You could then have a group or troop of men as a single Navmesh agent, but there would probably be lots of tweaks and heuristics needed to manage this well. E.g. troops cross narrow bridge or climb ladder. Widening of bridges to stop troops falling off or code to allow troops to fall into moats and drown.

    NA-RA-KU yes, or at least troops just going through the shooting animations and the battle engine calculating effects on enemy units.
     
  23. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    Just tried a test with Graphics.DrawMesh and was amazed to find although faster at rendering without collision detection, when I implemented a simple array based collision detection system (each unit represented by a binary mask in an int array) it ran slower. About 25fps in webplayer and 30fps in standalone build.
     
  24. chanfort

    chanfort

    Joined:
    Dec 15, 2013
    Posts:
    641
    I am also creating RTS game and facing this problem. When I looked to how trees are placed on terrain (as it can be placed hundreds thousand of trees), I found that trees are putted into arrays:
    http://forum.unity3d.com/threads/180569-Remove-trees-from-the-terrain-during-runtime

    This could increase performance if it would be possible to store objects into array and instead of calling Update() or FixedUpdate() thousands of times it could be possible to call these functions once and modify arrays. My idea is to test how such ways compare:
    GameObject1
    Update()
    GameObject2
    Update()
    ...

    with this one:
    Update()
    GameObject1
    GameObject2
    ...

    For example if you put army of knights into one array, on each update they would behave like a large group and as a single game object. But if you want to rearrange order of this army, you could probably rearrange positions inside this array.

    I know that arrays could increase performance in programming compared with recalling functions many times. But here the question could be: how to work with such arrays. Do anyone have any experience with that and maybe somebody have an answer how this could affect performance?

    I will be looking forward to write very simple scripts just to put many objects into array with different coordinates and just check how it will work...
     
    Last edited: Dec 18, 2013
  25. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,614
    Arrays don't magically make things faster. Working effectively with memory access is what matters. Though you clearly won't be doing the same kind of optimisation these guys are doing if you're working in Unity, this blog post does a pretty good job of explaining why, specifically the Memory Access section.
     
    Last edited: Dec 18, 2013
  26. Velo222

    Velo222

    Joined:
    Apr 29, 2012
    Posts:
    1,437
    Honestly, the biggest bottleneck for me is actually animations, or at least it was until the mecanim game object optimization option, which was just introduced in 4.3 I believe.

    The next biggest hindrance is the physics system. Having simple capsule colliders on all my units slowly adds up. That being said, I've mainly only been testing 100 vs. 100 battles (each unit has it's own ai, capsulte collider, and of course animations running), and I'm getting about 50 fps during the battles with everything running hard.

    I might be able to manage 200 vs. 200 at some point, but that's my goal. So I'm mainly mentioning this to say it's really hard with Unity to achieve a "Supreme Commander" unit count, and still have playable framerate. But then again, Supreme Commander units were very simple, with very low or few animations.

    I am currently working on an RTS as well, and my dream is to have 400 vs. 400 battles that can all be on-screen at once, not even thousands. But I'm still a long way off. The "optimize game object" option has helped a lot though. At least they're making progress.
     
    Last edited: Dec 18, 2013
  27. MikaelTroc

    MikaelTroc

    Joined:
    Nov 2, 2012
    Posts:
    33
  28. Doddler

    Doddler

    Joined:
    Jul 12, 2011
    Posts:
    269
    There's a good number of optimizations you could to to make this possible. As mentioned before, AI Wars allows for ~30000 units and is done in unity. x4000, the developer, mentioned he abandoned most of the unity scene logic and rather than have each unit as a separate entity in the hierarchy, he handles units internally and displays them by generating the quads that the units are displayed on in code into one or more giant meshes. He has no collisions or pathing to worry about though, so that simplifies his situation.

    Using built in physics is overkill for this kind of game, and definitely won't work smoothly with that many units. For one, if the units don't leave the ground, it's effectively a 2D game, and would be better to treat it that way. You probably don't need to use physics at all, and can get away with direct distance comparisons between units, though maybe you could get unity 2d physics to work for you. To align the units to the ground, a raycast might work but could be overkill, you could just use the terrain heightmap as a lookup to figure out how high a unit should be. If you were clever, you might be able to pass that heavy lifting over to the GPU with clever use of the vertex shader.

    To make everything run smoothy, you could have all the game logic run in it's own thread. The logic thread could run infrequently, as low as 4-5 times a second. The logic would run ahead of real time, and determine what will happen in the next real time block. So the logic determines what happens for each unit over the next say 200-300ms, and then main thread would simply execute those pre-determined actions. So while the game is displaying what was previously determined to occur, the logic is busy processing the next time step. That causes a delay in input to action, but as long as you give immediate feedback that input was received most players won't even notice (as an example Supreme Commander had a 300ms delay between internal updates but most users don't notice any delay).

    Displaying that many units could be a problem though. You could go like the old total war games and use 2d quads on 3d terrain, that would likely work pretty well. If you used 3d models, you'd probably have to use a lot of weird optimizations to make things fast enough. Skinned meshes with bone weights won't do on the scale of 1000 objects, it's too computationally intensive. I'd say your best bet would be to throw memory at the problem by using (very) low poly meshes and storing the vertex data for each frame of animation. If I had to guess I'd say total war and other similar games use (or used in the past) a similar approach because it's pretty clear that unit animations run at a lower framerate than the game itself without any interpolation. You could then batch groups of units and position these objects in screen space.

    If you had projectiles like arrows, you'd handle the physics yourself. When a projectile is fired, the logic can calculate where and when it will land. An arrow won't collide with anything until it hits the ground, so you don't have to do any collision checks until it reaches that point. Once it does land, you only have to do a 2d distance check to see if it hits anyone. That's still a large number of comparisons to be made, but likely your units will be grouped up into groups or formations anyways, so you could use bounding boxes for groups to easily rule out large number of units at once.

    I haven't done any of this myself so it might not be the best approach, but thinking about it, it's probably the right way to go to speed things up.
     
    Last edited: Dec 18, 2013
    xiaohai_pan likes this.
  29. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    Have you seen the Mantle API demo from Oxide.



    In the demo about 26+ minutes in they are throwing around thousands of ships and ten times that in missiles and blaster bolts.

    Maybe UT could adopt the Mantle API?
     
    Last edited: Jul 9, 2017
  30. BrUnO-XaVIeR

    BrUnO-XaVIeR

    Joined:
    Dec 6, 2010
    Posts:
    1,687
    You guys are looking for a concept called "Position Based Phisycs"; Forget about PhysX for something like this.
     
  31. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,614
    This is a similar approach to mine. I do have collisions and pathfinding, but likewise I don't use the built in stuff for that. Since this is a pretty niche application I built a custom solution for it, though it should be pointed out that doing that isn't as hardcore as it sounds.

    All software is built based on assumptions. The assumptions Unity have made for their general purpose tools aren't necessarily the best assumptions for more niche scenarios. What assumptions can you make or apply to what you're doing to minimise the computational work required to run your scenario?
     
  32. Velo222

    Velo222

    Joined:
    Apr 29, 2012
    Posts:
    1,437

    So I think the biggest help for a person like myself is are there any tutorials or videos (either free or payed) that could help someone learn how to optimize for an RTS (i.e. large amounts of units as the original poster stated)? Practical Unity-specific lessons would be the best, in Unityscript or C#.

    My other thought is: angrypenguin are you available for teaching lessons? lol :) I'm only half-joking there.


    @Arowx: Mantle sounds awesome. Even if it limits your customer/player-base to only AMD users. But then again, I don't have to worry about profits as most of these larger companies have to. So I'd be all for Mantle support in Unity. That being said, I don't see it happening. Unity tends not to make high-risk moves without a proven track record of working with a broad-base of platforms and with an eye towards mobile devices.
     
    Last edited: Dec 19, 2013
  33. BFGames

    BFGames

    Joined:
    Oct 2, 2012
    Posts:
    1,543
    Partition the space is very important, then write your own simple AABB collisions system, multi-thread it.
    Would also not use a "normal" A* approach.
     
  34. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,614
    I don't think you can really make tutorials specifically about this, because each solution needs to be tailored to the style of scenario being created and the method of presentation. As for stuff teaching you how to work efficiently with large sets of data, I already linked to a series of good optimisation tutes. Even at a higher level than the specifics of what the Overbyte guys talk about, the essentials are understanding what your CPU and memory are doing, and creating algorithms and data structures that suit that. (This is the opposite to what most universities and courses teach, which is to design things in the way that the human brain works because it's easier and cheaper to implement and in most cases performance doesn't matter. Low-level game programming, however, is not one of those cases.)

    Even if I were available to teach I'm not experienced enough to do that. I know enough to spot the most obvious places my code and designs suck, and to experiment to find ways to make them "good enough", but that's all. Code is pretty rarely a bottleneck in the projects I work on.
     
  35. BFGames

    BFGames

    Joined:
    Oct 2, 2012
    Posts:
    1,543
  36. Velo222

    Velo222

    Joined:
    Apr 29, 2012
    Posts:
    1,437
    Hmm that looks like a cool book. Thanks.
     
  37. chanfort

    chanfort

    Joined:
    Dec 15, 2013
    Posts:
    641
    It might be possible to use particle system for large armies to set particle positions if possible. I found this to access how to set individual particle positions:
    http://docs.unity3d.com/Documentation/ScriptReference/ParticleEmitter-particles.html

    I read that it can hold up to 100k - 150k of particles. I am not very sure if it would be possible to do everything what you can do with regular objects (for example to start one or another individual particle animation), but if it would hold, it would be good results.
     
  38. chanfort

    chanfort

    Joined:
    Dec 15, 2013
    Posts:
    641
  39. Windexglow2

    Windexglow2

    Joined:
    Sep 23, 2012
    Posts:
    90
    I'm making a modern rts with the idea of as many units as possible. Lot of fun hurdles to go around : guns, intelligent looking movement, reacting to enemy (movement, placement), injuries (bringing friendlies back) and so fourth. But it's fun as S***.

    Last week I tackled a problem I didn't even think about : Sound. Usually there's not enough going on to be bothered by sound. You shoot your gun? Play a sound clip. Turns out playing a sound clip each time hundreds of units shoot is difficult on the cpu. And I also learned sounds start acting up if you play too many of the same clip at once.

    The solution is pretty cool : I downloaded a 10 second loop of wind. I put it ingame with a volume of 0, pitch high, and looping. Every time a unit shoots it increases the volume. Every fixed update the volume decreases. It works incredibly well. I have a sound for the initial pop (fast fall off), a sound for echo (requires multiple shots quickly to hear it, but falls off slowly), and a sonic crack (super fast fall off, only works 10% of the time in front of a shooter). With dozens of units firing quickly there isn't any issues.

    It sounds like gunfire because the volume goes from 0.0 to 1.0 and quickly fades away. It takes a fraction of a second to fade, but it sounds great. But what happens if there's a gunshot every frame, or other frame? Than it starts to stay near 1.0 constantly and you stop hearing that pop, instead hearing what the looping sound is (wind). But at that point any sound is going to sound muffled.
     
  40. Vaupell

    Vaupell

    Joined:
    Dec 2, 2013
    Posts:
    302

    https://dl.dropboxusercontent.com/u/19148487/CubeWars/Web.html

    You webplayer link runs with

    Assault - steady at 53 Fps
    Fight - increasing slowly to 57
    After - steady at 62.7
     
  41. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    It is hardware dependent, still about the 45fps / 59fps / 60+fps for me.
     
  42. Ziron999

    Ziron999

    Joined:
    Jan 22, 2014
    Posts:
    282
    "Reviving old thread haha" The # of units is necessarily the problem, ya you can do 4000 units but....how are you going to target the nearest enemy during that in your web demo they just destroy themselves on impact no AI at all...
     
  43. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    The demo just demonstrates that you can move a lot of elements in Unity, and they can have simple interactions.

    The combat would depend upon the scenario you are trying to recreate, e.g. medieval combat is melee whereas modern combat is ranges. But this would probably be calculated using random probabilities and a combat sub system.

    The AI would need to focus on entire companies or groups of Units and giving them orders e.g. move to, attack, retreat, hold and therefore be a higher level system than the basic movement and combat system.

    Yes adding these systems could impact performance although you could run them on separate threads so they do not impact the main Unity thread.
     
  44. Neoku

    Neoku

    Joined:
    Oct 27, 2014
    Posts:
    261
    Nice demos but in a real game each unit are a detailed (low or medium poly textured and rigged model) then my question is if Unity was used in a batle game of this type?
     
  45. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,958
    The unit may appear detailed while zoomed in, but the quality will typically decrease the further the camera is zoomed away from the mesh. At some point it may even become 2D while appearing 3D.

    http://en.wikipedia.org/wiki/Level_of_detail
     
    angrypenguin and Kiwasi like this.
  46. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,614
    Yeah, heavy AI work shouldn't happen per-unit-per-frame. It should only have to do significant work when something changes, and the low-level actions that it triggers (move to location, play animation, whatever) should be cheap.
     
  47. SunnyChow

    SunnyChow

    Joined:
    Jun 6, 2013
    Posts:
    360
    I have no experience on RTS instead of tower defence type game. But If i have to make a RTS with massive units, this is what i am going to do:

    I will use ParticleSystem for the rendering. I feel that it's faster than dynamic batching.
    There will be no real physics. I will use a height map to adjust units' y position and something like an AI map to adjust their movement .
    there will be one main script to control all units instead for one component for each.
     
  48. ostrich160

    ostrich160

    Joined:
    Feb 28, 2012
    Posts:
    679
    So many factors to put into this, but in theory yes, with the right optimization
    But you gotta think, 2d or 3d. That makes a difference. How many poly's per unit, how would the AI work, animations (a 3d soldier running needs an animation, an 8bit flying spaceship does not), so much to consider. Long story short, its possible, but will require sacrifices
     
  49. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    You can mix 3d and 2d check out the game development technique Imposters, the idea is that you have the most distant Level of Detail objects as pre-rendered 2d images.
     
  50. chanfort

    chanfort

    Joined:
    Dec 15, 2013
    Posts:
    641
    I just made free edition of uRTS available for everyone:
    https://www.assetstore.unity3d.com/en/#!/content/30247

    It can handle battles up to several thousand units at runtime (check out demos in description). It also have more "clever" AI phase system, where units are searching for targets, moving towards them, attacking them and finally dying.

    I also work for a while with uRTS full version, which includes real models and sprites. Optimisation there works quite well for several thousand of units with all animations playing and no need of dll's. However, extra care needs to be taken, as each new implemented feature always affects performance of previously implemented features.

    Here are some screenshots how these battles looks like:
    big2.png Screen Shot 2015-02-12 at 23.35.36.png
     
    Last edited: Feb 22, 2015