Jump to content


Community Members
  • Posts

  • Joined

  • Last visited

ScionOfWar's Achievements


Discens (2/14)



  1. It's easy to prevent too much unbalance, at least when placing the units. But if they have flight or fight responses similar to real wildlife, then the game board can change in favor of a player even if every critter on the map is meticulously placed for balance. I worked on RMS scripts in AoM and it was possible to create clusters of animals distanced from other clusters and buildings and forests, ect.. However, in AoM, the animals just stood there, even if you slaughtered their buddy nearby. Theoretically, if the deer run from your army like they would in reality, then a clever player could corall all the huntable animals to his side of the map, effectively herding them away from their opponent. Though this might lead to one player having more food then the other, I find it perfectly reasonable and realistic. Food like deer *do* run. And if you're not careful, deer could be herded and/or stolen from your side of the map by your enemy, making it difficult to ever get them back. Guess you're stuck with farming! Personally I think balance should be sacrificed in favor of realism and strategy. This is one such instance
  2. If you need to, you can lower some settings that'll make the game run smoother. For my computer the fancywater simply doesn't work, and unless I need them for screenies I remove the shadows too for the sake of speed (doesn't slow me down too much but every bit helps!) Visit this folder: \binaries\data\config And make a local.cfg file with information like this: fancywater=false shadows = false default.cfg has more settings you could possibly tweak in your local.cfg file to increase performance.. It may just be that you're on 64 bit though. Then again keep in mind that some of the game is still in progress, in particular if you're running on the older sim system then pathfinding and such is very glitchy. There's a newer sim setup with better pathfinding but lol you'll need VC++ to build it
  3. You can test it even if you can't get Visual C++ to work or you don't have enough space: Try going to this folder: \binaries\system Then double click either pyrogenesis_dbg.exe, pyrogenesis.exe, or atlas.bat and you're set to go. I personally don't know yet about the debug tool.. Haven't played with the coding too much since I installed, heh I've been too distracted making models
  4. I agree on all those fronts and was actually thinking that myself concerning the over-saturated wood.. It shouldn't be too hard at all to change though That persian texture is fantastic and I'll definitely get some use out of it for the texturing of the pillar Right now I'm trying to use the texture to bring out some non-modelled details in the statue (such as the ears and legs of the bull) and once I get a good foundation I'll try mixing in the persian textures you offered I'm personally surprised there aren't more people jumping at the oppourtunity to help the game out. I don't think this reflects poorly on the game - just poorly on peoples abilities to download and compile source code. I think as the game gets more and more playable, more people will jump on board and supply stuff. I recently talked to my older brother and I'm trying to get him to join in and help too.. He's been playing with game engines for a few years now, and in fact we were once discussing working on a game design together (since I'm better at modeling and decent at coding, while he's better at coding and decent at modeling lol).. With any luck we'll hear from him soon. I'll try to keep my momentum in the meantime and pump a model out every day or so. First step: finish that persian pillar! I'm also going to make a variation of the pillar which holds the beam on its shoulders, similar to the reference I showed you, so that you could make some structures with it. Unfortunately I barely touched 500 polygons on that so it's not the lowest poly pillar in the game And that was after two hours of trying to reduce poly counts without loosing detail. I may try to make an even lower poly version just to see if it's possible without wrecking the shape, but I can't imagine taking out more then ten polygons from the bull portion..
  5. Actually my newest screenshot above fixes the cross problem (I think) Thanks though
  6. How's this? I know it looks kinda dwarfed compared to stonehenge but I made it narrower so the beams don't look quite as chubby. I also added blood to one of the textures as a variant. Also included in the shot is a stonehenge model, broken pillar, crossroads, greek style arch, and the most difficult one to make yet, a persian pillar (in progress): http://i146.photobucket.com/albums/r249/Sc...pg?t=1266664049 @Badass - I do believe the 'den hole' concept you posted is already in the game in the form of animal corrals.
  7. I was working on a persian pillar (based off this and other references - http://lh4.ggpht.com/_kpSrBksrlDc/Ry5zhcnU...Q/DSC03795.JPG). Still need to texture it quite a bit but the modeling is done. Also fixed the crosses: http://i146.photobucket.com/albums/r249/Sc...pg?t=1266664049
  8. lol I actually made a large wooden cross I was going to make one with a corpse but I need to figure out how to set up the body first heh http://i146.photobucket.com/albums/r249/Sc.../untitled-6.jpg Good suggestions though, until someone gives me something more important to work on I will try to knock out some more eyecandy
  9. And now more simple embellishment: signposts As a question, what would be the preferred naming conventions for these things? Right now I've just been naming them 'wrld_' + whatever their name is. But if you would prefer we could call them embell_ or something else o.O
  10. I have some crosses to show you guys One with the INRI inscription (IESVS NAZARENVS REX IVDÆORVM or in English "Jesus the Nazarene, King of the Jews") and one without. I was thinking about texturing some holes where the spikes went, and blood onto it as well. As you might guess these are extremely low poly since they're such simple geometry. Edit: Bob/Kimball suggested I make the beams thinner so I will fix them up and post a new screeny later I figure it doesn't get much closer to 0 A.D. then the death of Jesus! Considering it was his life and death which distinguishes A.D. from B.C. haha
  11. Thanks guys I was looking at that eyecandy thread and the stonehenge someone asked for was particularly inspiring.. I also gave a go at fixing the UVW maps on my mosque thing and texturing it and the results were decent (though the texture looks more celtish then turkish so it definitely has room for improvement) Basically I've made four models in this screenshot - mosque (like I said textures look kinda earthy until I fix it), greek arch, stonehenge arch, and a broken stonehenge pillar I made a fancy screenshot complete with bushes and textured roads! Check it out
  12. Hey, I wanted to prove my worth to you guys by modelling something. Problem is, I don't want to redundantly make a model that's already been made. So my question is this: Is there anywhere I can see a list of incomplete/unstarted models? In particular I would like to work on building design since I'm not confident enough in my skills at making low-poly organic characters. Ships, carts, siege engines, and stuff like that I can easily do though Even simple embellishment units, like ruins or bee hives or whatever.. I do have experience in 3ds Max modelling, UVW mapping, and burning AO to a texture, so these things shouldn't be a problem In fact I usually work from high-poly and then go down to low poly which usually results in more detailed AO textures Toss me some suggestions and I will get to work! The best result is you guys will get some nifty models and speed up game design, and the worst result is that I get a nice low-poly model to use in future renders, so it's a win-win? edit: Also, I tried importing a mesh into Atlas and was successful on everything but one small problem. When I try to rotate the model I made, Atlas freezes. Maybe this is a common problem? edit2: fixed that glitch.. Had something to do with how I set up the entity file I believe edit3: A very unartistic screenie for you guys of some experimenting I did with Persian architecture... I'm not done with the cathedral/mosque (need to re-UVW it cuz the seams are too obvious and then texture it.. the texture on it right now is just an AO render I did with mesh smooth+edit mesh modifiers) and I still want to tweak the arch and make complimenting ruins.. Also the scales aren't perfect yet heh http://i146.photobucket.com/albums/r249/Sc.../untitled-2.jpg
  13. It certainly does work! I love the ability to see the calculated path, that's a very nice feature for debugging and tweaking. I am going to play with this for a while and I'll probably pop back up asking questions again in no time I did notice a possible problem - when you right click on an impassible node, the script searches an enormous radius. Maybe something worth detecting because that could result in sizeable calc times if you select 30 units and right click a tree hehe.. though it may be possible to treat grouped armies as one unit, at least for the pathfinder (speculating lol) However the lag I mentioned seems remedied by the changes in the pathfinding, even with the debug visuals and the aforementioned problem. It looks pretty efficient too in terms of searching. The best I can do is maybe tweak and play with it and show you guys my results lol Thanks again for your help and keep up the good work. Hopefully as I get better I will be able to contribute
  14. lol I have to agree here. The basic logic behind A* seems simple enough, but as the third link I posted above kinda proves, there's endless ways to design the heuristic. I heard one person compare it to art haha! I'm going to definitely have to experiment with it, since it's so interesting (read: confusing as hell) I was wondering if something like this is feasible as a rough concept for how the heuristic would place the cost on open nodes. The differences could even be made smaller so it spends less time differentiating if it was a problem: Unfavorable terrains (marsh or forests) - 1.01 Grasses/Easily passable Terrains - 1.0 Favorable terrains such as roads- 0.99 Nodes containing human units, or trees, could be 'passable but highly unfavorable' (maybe a value like 2.0) to prevent the blockage we see sometimes between two units. This may make it possible for units to pile up on eachother, but it would potentially fix the chokehold problem of big armies (unless buildings enter the fray) Cliffs and occupied building slots would be closed completely, and might even alter the heuristic of nearby nodes to make them 'unfavorable'. This way it would choose any one of the three normally available nodes, but perhaps differentiate especially on longer paths so units might prefer a small detour around a forest over going through it. This could be especially increased when you're using armies, rather then individuals, since a platoon of 30 soldiers is clearly going to take more space so they detour around congested regions. Of course the numbers above are just to demonstrate the idea, the exact figures could be anything. For speeding it up I was reading some interesting things which we could maybe put into practice. One is to calculate long paths using a smaller grid system (like what you mentioned with the triangulation), then calculate the active 'branch' of the path only. This way you can spend less time calculating how to get from point A to B to C to D, and just focus on point A to B first, and then B to C, and finally C to D. Another point I noticed mentioned several times is that overestimating the heuristic is always better then underestimating. It may not guarantee the shortest path, but it can help to guarantee the faster calculation. But this may be easier said then done heh. I'm personally of the opinion that longer paths may sometimes be interesting, and that we might want to entertain them especially if it means quicker computations. After all, units in reality sometimes choose their own route, not necessarily the absolutely quickest or even the one given by their commanding officer. Then again people may expect their soldiers to march through the forest only to be surprised (and slaughtered) when they instead flank around the left into a fortified enemy base, and too much of this can definitely be detrimental. I will continue to play with this, my mind is full of ideas. Most of them I'm sure you and others have hashed through, but perhaps I'll offer a fresh perspective It's a learning curve though, this is my first time working in Java so I'm a little unfamiliar with some of the rules. Luckily, all scripting languages seem to be similar at least in the thick of things. It's just all those -> and .push commands I will have to figure out hehe As a question for setting up the alternative pathfinding system: you said "you can try it by using the "-sim2" command-line option." ...I know this may sound stupid, but ehh, how do I do that? I'm new also to building in the workspaces so I'm a bit clumsy with it. In case it matters, I'm on Windows XP. I would love to dedicate some time toying with pathfinding though. Off the topic slightly, I had a dream about pathfinding last night Anyways thanks for the reply, I will go spend some time dissecting the file you mentioned
  15. I've been playing around with the pathfinding engine and I found it to be extremely laggy. Even on short distance paths (IE around my buildings) there's substantial lag. I really am interested in the scripting behind pathfinding and I've been reading an article for the last two hours or more about the A* Pathfinding algorithm. It looks very interesting and potentially effective, *if* it has the proper settings (eg the heuristic or 'guestimate' on where it should check). I was wondering if the triangulation calcs that I see in the pathfinding engine are truly necessary or if they're just adding a ton of overhead to the simulation when it comes to actually moving the units around. From my understanding the A* algorithm applies a value to every point on the grid based on distance to the goal (h) and distance to the starting point (g). The actual algorithm is f(n) = g(n) + h(n), where f is the total 'cost' of the point on the grid. Using the heuristic 'guess', it snakes to the goal and on contact searches for the least costly route back. It seems pretty straightforwards and there's some real-time programs based off it which lead me to think we can possibly fix the 0 A.D. pathfinding lag and issues. Here's some examples: http://pixelwelders.com/experiments/Pathfinder2/ http://www.vision.ee.ethz.ch/~cvcourse/astar/AStar.html# http://theory.stanford.edu/~amitp/GameProgramming/index.html After studying A* I think it would be possible not only to have units avoid other units, but also to have them prefer roads over grasslands, or one elevation over another - realistic features for a realistic game! To be honest I'm still trying to figure out the logic behind the 0 A.D. engine. I'd love at least tinker with it but maybe someone can give me some pointers on how it's working? Is it using A* math or something else I could research to get myself up to par? It looks like fun!
  • Create New...