Search Unity

  1. Good news ✨ We have more Unite Now videos available for you to watch on-demand! Come check them out and ask our experts any questions!
    Dismiss Notice

Games LANDNAV - Version 0.01a available now

Discussion in 'Works In Progress' started by BIGTIMEMASTER, Dec 1, 2019.

  1. TonyLi

    TonyLi

    Joined:
    Apr 10, 2012
    Posts:
    10,730
    That's my top choice, too. The patch look is distinctive. Changing it to gold lettering really makes it stand out.
     
    BIGTIMEMASTER likes this.
  2. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    thanks @TonyLi .

    before i publish the coming soon page i'll put it up here for review. always good to get some outside feedback
     
  3. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    https://store.steampowered.com/app/1291570/LAND_NAV/?beta=1

    Actually I go ahead and put that up there. It's preview of steam page. Nevermind the screenshots and lack of trailer, that will all be replaced. If anybody has an opinion on the write-up I'd really appreciate it.

    Main thing I want to accomplish: Give enough idea about the game to entice, but not so much detail that a person loses interest because they think they know what it is (they won't really know without playing it).

    Write in a way that is clear and understandable but also use a light amount of jargon to suggest the technical depth of the game. (should be interesting, not intimidating)

    Deliver message in an honest tone, like i am your good friend trying to get you to play a game I love (because that's basically the truth. Rockstar can't pull that off, but I can, so I figure I ought to).

    The game mostly appeals to military and military hopefuls, but I feel like I don't need to do a lot to sell to them. It's the only landnav game, so they're going to play it so long as it doesn't look like ass and have bad reviews. I target my writing then to people who might be predisposed to having interest in something like this but don't know it yet. The "swing" audience I guess you could call them.
     
  4. TonyLi

    TonyLi

    Joined:
    Apr 10, 2012
    Posts:
    10,730
    Remember to post your Steam page to Feedback Friday to get more eyes on it for opinions.
     
    BIGTIMEMASTER likes this.
  5. Deckard_89

    Deckard_89

    Joined:
    Feb 4, 2016
    Posts:
    284
  6. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    I believe you will need to be signed into steam beforehand for the link to work correctly. Were you already signed in?

    if so and it's still not working, i ought to have this ready to go live by feedback friday and will post there

    @Deckard_89
     
  7. Deckard_89

    Deckard_89

    Joined:
    Feb 4, 2016
    Posts:
    284
    Ah sorry, I wasn't signed in.
     
  8. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    i am making some major changes already though. if you have some critique might as well wait until this weekend cause its all different already :)
     
    Deckard_89 likes this.
  9. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    996
    very cool landing page graphics, i like the embroidery look on the logo. How did you make that?
     
  10. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
  11. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    996
    Thank you for the tutorial, I don't know anything about art. My friend puts all together she says she uses something called 'krita' so am not sure if it is compatible with this software. In any case your landing page looks nice.
     
  12. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    Small update about some big decision:

    We are having a recurring issue with a crash related to some sort of memory hangup. Is it something we can solve? Sure, but it takes a lot of time to troubleshoot because it's not a simple problem. Because I got to pay my programmer for his time, and in this case even try to get help from other people, increasing the total time cost even more, I am considering bypassing this problem altogether because the cost is going to put me over budget.

    I have to do some testing, but no doubt the issue is just related to the amount of grass and trees we are using. I can just decrease that in total, but then the game doesn't look very nice, plus the vegetation is actually an important gameplay variable. So that's not a super attractive compromise.

    Another option which I think is the route I want to try is actually to change the theme a bit. If we are in the mountains rather the forest there is a few benefits:
    1. Don't need nearly as much vegetation
    2. terrain association will be easier for beginners

    but it would mean i got to change the theme a bit. That could possibly be better from a marketing standpoint. Rather than military training, we could be a crashed helicopter pilot who has to use land nav to escape pursuing taliban fighters. Same gameplay, but more exciting theme that could possibly interest a broader audience.

    I guess it isn't essential to change the theme, but I like it when everything fits together correctly.

    Well, i will be doing some experiments, let me know if you have any thoughts on the matter.
     
  13. BonneCW

    BonneCW

    Joined:
    Jan 22, 2017
    Posts:
    81
    Is it a guess that it is related to the grass or have you tested it?

    How do you know it is memory related at all?

    Just asking because as far as I know you're no programmer and then people tend to speculate around, even though in your case you at least have some experience.

    You'll probably get some ideas for the cause if you have some more informations to share.
     
  14. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    I have a thread about this in the general support area that has screenshot of the error.

    I am doing some testing to determine if it's related to the VGS system or somewhere else.
     
  15. TonyLi

    TonyLi

    Joined:
    Apr 10, 2012
    Posts:
    10,730
    It sounds like you already sense it's an opportunity in disguise. You can bump up the excitement factor to reach a broader audience, and maybe also bump up the visual factor since (at least for me) mountain vistas with lots of vertical are more interesting than flat terrain. Can you still keep the gameplay loop legit for hardcore military orienteering folks even though the story premise is altered?
     
  16. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    yeah it could be for the better.

    the gameplay loop will change, but the skills you are using is the same. there must be a trianing zone no matter what, so those people who are interested in the game solely for learning shouldn't see a difference.

    i think the game could be better if the stakes are raised like that. a survival situation where you are being pressed by outside forces, rather than asking player to exercise high degree of self motivation in a video game.

    and i think simplifying it down from "find tons of these metal stakes scattered all over the place" to "start in random location, figure out where you are and escape pursuing enemy to a safe zone" is probably going to be more interesting for most people.

    still, its kind of tough call to make. me doing some different art is no big deal, thats like a week of work. With this alternate theme I think i can cut down the amount of work for programmer to do as well.
     
  17. TonyLi

    TonyLi

    Joined:
    Apr 10, 2012
    Posts:
    10,730
    How about adding some reason for intermediate checkpoints before escaping the enemy? Kind of like having to get to CPs in regular orienteering.
     
  18. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    Definitely some opportunity but I think we better start with the simplest approach first. Once we get the details hashed out I'll make an update.
     
    TonyLi likes this.
  19. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    Here's a summary of changes we will be making:

    • To improve story mode experience and performance/stability, we are changing the map from North Caroline woods to Afghanistan mountains.
    • To ensure high quality long range visuals and improve performance, we will be doing some experiments with loading each terrain tile as separate scenes, making use of asynchronous loading. I'll use low poly meshes to stand in for distant terrains.
    • Story mode gameplay will be simplified - you are a pilot who has been shot down. Taliban pursues you. Evade, use your map and compass to triangulate your position, then figure a route to refuge while avoiding the enemy.
    • Each new game will drop you in a random location and refuge zones will be selected based on proximity from spawn point.
    Training area will remain unchanged, except for the setting.
     
    Deckard_89 likes this.
  20. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    I've got the new Afghanistan map started. Only terrain and textures. I'll accent the rocky ridges with some boulder models, and the valleys will get some greenery where there is irrigation.

    The new story mode really simplifies the amount of code we need, and I think will add a lot more replay value and emergent gameplay. But we do have to design some AI. We are trying to keep it really simple to start with. Will update once we get a bit further with that.
    Desktop Screenshot 2020.04.21 - 00.41.17.10.png Desktop Screenshot 2020.04.21 - 00.43.18.23.png
     
    TonyLi and Deckard_89 like this.
  21. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    996
    Really cool update.
     
  22. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    Desktop Screenshot 2020.04.21 - 23.39.56.84.png A long day but I achieved my goal and am happy.

    I wanted to make good use of tesselation shader on the terrain and also do a better job in world machine. Result is a believable terrain using shader only, no extra models to spawn.

    Only thing missing is the very far background and then in the lower valley we will have some villages with irrigation along a river.For the far background I am not sure how to accomplish that. I have some ideas, but it's low priority and I'll finish that later.

    I am using 5km terrain tiles. A total of 100 of them. But only the play area has higher resolution height maps (about 2 pixels per meter). The background terrains only need 512 heightmap and then even lower for the furthest ones. So that's kind of like an LOD system to cut down on the texture sizes.

    Haven't run profiler because programmer is working on the enemy AI, but entering playmode is instantaneous and all the performance measures I understand are in the green.
    Desktop Screenshot 2020.04.21 - 23.35.26.73.png
     
  23. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    reworked the map to more accurately represent afghanistan. using microsplat here for the texturing.
    Desktop Screenshot 2020.04.24 - 00.38.02.99.png
     
    Last edited: Apr 24, 2020
  24. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    So for a 25kmx25km playable area there is no way were baking navmesh for enemy AI agents. We instead use a technique derived from the old command and conquer games. I won't try to explain the technicals because I only have laymens understanding, but it works sort of like satellite GPS. We are shooting rays down in a grid around the agent to determine what is around it, then it can make it's decision about where to go.

    This approach is not limited by terrain size, and for our purpose it is "smart" enough and fast enough. Our Taliban patrols will move along waypoints in a search for the player. If player is spotted, they will pursue aggressively. Most of the time that will mean game over unless player is pretty far away.

    We aren't giving player a gun because this is a realistic game. Even if you had a weapon, it would be suicide to try and fight off a squad in unfamiliar terrain. So the game is only to escape and avoid being seen.

    My idea is that I don't want it to be like a stealth game with the usual trappings. "Watch the enemy pattern, run across to next cover when they face the other way." I want enemy AI to force player to have to create new route plans on the fly. It's open country and if enemy sees you you are as good as dead, so the thing to do is get out the map and find a new route, rather than just sit and wait for enemy to turn around.

     
  25. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    How to make contour lines for 3d game world map:

    This will be a short write up for future google searchers. I've seen a few people ask about how to do this and I have only found one answer myself - that answer which relies on using terracing and slope detection in world machine wasn't accurate/predictable enough.

    So my solution is laborious and will take you an entire day, but it is precise and relatively simple (a dummy who doesn't script can do it.)

    Overview:
    We want contours at precise, set intervals. For my map I need contours every 100 meters. The idea is to measure the increments with cubes and use an intersection boolean to cut up the terrain mesh into separate objects. We can then bake an object ID map in Marmoset Toolbag.

    Steps:

    1. Export terrain as mesh from World Machine. Must be same resolution as the one used in game. Single mesh is easiest, but you may need to export tiles.
    2. In my case, a single mesh terrain of 25km area with same resolution as in game was 16 million tris, so I have to decimate in zbrush.
    3. In Maya set your mesh to scale. I have a 25km terrain, so I set it to match a 25cm grid.
    4. Create cube which is as wide and long as terrain, with height at the appropriate scale for your intervals. For 100meter intervals, I use 0.1cm tall cubes.
    5. Align your cubes with the grid in orthographic side view. Turn on grid numbers and xray shading for easier work. Keep a backup copy of the terrain and cutter cubes. Put these in a group and hide the group.
    6. Grab a copy of terrain and cutter cube, duplicate. Boolean>intersection. Maya will fill in the center - delete this for clean baking layer.
    7. Do this for all contour intervals.
    8. When all booleans are complete, double check that the centers are not filled with a big n-gon. SHoot a planar projection for UV's, then export to Toolbag for baking.
    9. YOu use the same mesh for low and high in the baker. Bake either an object ID map or a UV island map. Object is better because if you have your objects named by elevation, that will save you trouble down the road.
    10. Move that baked map to photoshop. Select your colors and apply layer style > stroke and > color fill. Now you've got perfectly accurate contour lines.


    Desktop Screenshot 2020.04.29 - 18.20.56.83.png Annotation 2020-04-29 181737.jpg
     
    Last edited: Apr 29, 2020
  26. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    Hey everybody,

    Been a minute since I made an update.

    My programmer got a cool job from a big client so he hasn't been able to work full time like before. Still is making good progress anyway. Our enemy AI is able to move along waypoints, perform simple searching functions like stopping every so often and looking around, and detect player.My goal for the AI is for them to feel competent and fearsome. They are not to be exploited - more like a dynamic obstacle that is there to throw a wrench is your plans.

    For that they don't need to be complex. A few simple rules do fine. 1, they have realistic peripheral vision and can see you from kilometers away. And 2, they stop and look around just like real soldiers would if they are searching. So if you can see them, chances are prety soon they will see you too.

    To help avoid such a dangerous enemy we changed up the controller slightly - you can now go prone to reduce your profile, plus you can sprint. If you are about to crest a mountain, it's a good idea to go prone so that your silhouette isn't highly visible from a long distance.

    Even though the enemy AI isn't fully functional, I am enjoying playtesting with them on the map. It definitely changes the dynamic. The main crux now isn't in finding precise points under time limit, but more of figuring out where you are while under pressure to keep moving because an enemy is searching for you. It kind of creates a situation where you want to panic but the calmer you stay the better you'll do. A more dynamic challenge it seems like.


    As for art stuff left to do, I will have to make a few models for the enemy AI, but they don't have to be super detailed. Once the enemy is within 100m they are going to shoot you so a simple model will do.

    You will have a couple options to escape to safety each playthrough. I think it will be a choice between a predetermined HLZ where friendly forces will pick you up, or some friendly villages. For the villages I have made some buildings, but I am going to experiment with the amplify imposter plugin. Again my idea there is to have them only be seen from a distance - once you get within a few hundred meters it would give you a dialogue box to ask "enter the village and ask for help?"

    it would be nice to make a realistic afghan village to walk through, but then you need villagers. and goats and donkeys. and children. and they all got be alive and doing stuff... well, you know. too much work for something not important. The artist would love to lovingly create a quaint afghan village but the boss is saying it's not in the budget right now :)


    I did make a model for the spawn point. Here's your crashed blackhawk. It puts out a big smoke plume for awhile that could help savvy players stay oriented when trying to get their initial bearings.
    upload_2020-5-6_10-49-12.png
    upload_2020-5-6_10-49-53.png
     
    Deckard_89 likes this.
  27. Anox

    Anox

    Joined:
    May 28, 2013
    Posts:
    1
    Wow, looks amazing! :)
     
    BIGTIMEMASTER likes this.
  28. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    996
    Looking great! I will have to check out this microsplat, tool looks very similar to unreal engine's "brushify.io" but not sure if it works on hdrp. Have you consider using something called 'emeraldAI?' and if so do you have any advice?
     
  29. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    @unit_dev123 , no I don't know anything about emeraldAI
     
  30. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    3,517
    I'll keep updating this blog here as we make progress, but if you want more detailed breakdown of whats going on I will be making those sort of updates at https://www.patreon.com/R2N.

    I didn't intend to do any sort of crowdfunding for this game but a couple people asked if there was any way they could support the project so I made that as an option. It's not going to be a big popular game but there isn't really anything else like it so if somebody is excited about the idea I'm happy to accept some help from them. Things are going a little slower with development now so having some outside help is really nice for the long distance motivationals.

    I'll also tie playtesting of more finished builds to patronage as well. Probably the next build or two will still be available for anyone who wants to try it but once the game is getting final fixings I think I want to reserve that for people who are a little more invested in the project - and also reward them by essentially giving early access to the game before it gets published.

    In development news, things are going good - we plugged AI into the full sized world. A few issues as expected but nothing major. I do really like the way it changes the game. It feels like a really intense survival game sort of now. Just knowing that any moment a squad of baddies could appear over any horizon and not having any means to deal with them really adds a ton of tension. You can't just sit with nose in the map for ten minutes at a time now - you really got to keep head on a swivel and learn to plan quick. It is probably too hard at the moment, I may need to add some cover so you can get a little sense of safety now and again, but I'll wait and see how playtesters feel first.
     
  31. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    996
    Most definately! It makes the game. Really looking forward to future updates sir!
     
unityunity