Jump to content

FirefightGI

Members
  • Content Count

    44
  • Joined

Community Reputation

1 Neutral

About FirefightGI

  • Rank
    Scavenger
  • Birthday December 14

Personal Information

  • Biography
    Lead software engineer at Paradox Solution Softworks - Game Development - Blockchain - AI.
  • Location
    In your game files.
  • Interests
    Modding, Game Development, Gaming
  1. I would like to note that I have been a daily user of Unity for about 5 years now using it for both non-commercial and commercial work and have followed the game since kickstarter. That said I'm a game dev, I like modding. One of the major blockers for me making an external mod toolkit with a nice GUI is the fact that we need the Unity editor to create asset bundles to add to the game. Now I have tried to reverse engineer Unity's asset bundle system but without much luck. So instead of modders requiring a whole game engine in order to import content consider this: https://assetstore.unity.com/package...-package-91777 TriLib: it's $25 package from the Unity Asset Store, I have used this A LOT, it can be set up in under an hour. What it does is allows the runtime import of 3D models including rigged characters and animations. A barebones implementation would be a command or XML argument to lookup a file path to send to TriLib. Better yet would be an in-game editor/automated system/XML params to enforce that things are set up correctly i.e. things oriented, parented right, scripts attached This would do 2 things 1) Modders can eventually ditch having to have the right version of Unity installed to create asset bundles. 2) Modders would not have to decompile your game files and have access to the source art assets/prefabs to look at how things are oriented. bonus 3) Replace existing art via xml? In my experience using TriLib in projects you may have to create a way for imported models to persist such as keep them in a pool tagged "don't destroy on load" and/or a mod file format that can store "prefab" like info so models don't have to go through the setup process each time the game is loaded. i.e. convert them to XML with the path for the 3D model stored. I have not posted in pimp dreams since like 2016 so idk if the devs will ever see this in the flood of posts, but if this is to be implemented please feel free to ask me questions about the system in a modding context since I have implemented it a few times in commercial projects. or just hire me for a month...
  2. This Unity voxel engine was around 5 years ago, even back then it was probably better than 7DTD's voxel system is now. Always makes me wonder.
  3. @ZehMatt As an asset store publisher I sort of agree, depends on the asset though. Personally I think A* pathfinding project is to complex for most uses, Unity's own A* pathfinding (that you find on GitHub) in my opinion is far superior but most people don't know about it. The only thing UFPS really has going for it is the procedural camera bobs/shakes and UMA is open sourced so is a bit messy (at least since I used it last). As it stands it would probably take way to much time to upgrade 7DTD to UFPS 2 considering they have already modified a lot of the code for their project but UMA to MCS might take me a day to two. The asset store you either need to publish very niche or very broad to cover use cases. My experience on the store is, don't publish anything with stock art and don't add any more scripted extensions than necessary. Users will recognize stock art and feel standard settings for things like UFPS. Currently I'm writing a tool to upgrade all textures in a scene/project to PBR and optimize 3D assets in which users wont notice and developers don't need to extend, assets like that are preferred. It is a shame that people think they can just buy things and make a game out of it, you can but probably shouldn't. btw @madmole if you still are upgrading PBR textures after December hit me up, my tool will make things 100 times faster (still writing the custom editor).
  4. 3)re-add the center indicator for sneak, we have whole screen indicators for everything (heatstroke, damage, etc.) but sometimes I forget im sneaking or am distracted by zombies and have to look at the bottom left, maybe just add a subtle sneak vignette. Been caught in situations not realizing I crouched and it slows you down. I might look for a hold to crouch setting.
  5. I felt more invested in and found the last two horde nights far more intense than any of the main FO4 plot climaxes. Heck the dungeon POI's are almost as fun. Great work I'm having the most fun with this out of all the alphas.
  6. Do you have enough space on your hard drive to store the temporary generated data?
  7. So you don't cull UI that is scrolled out of screen bounds or pool listed elements? Btw Does the saw do anything other than make logs, are log spikes gone?
  8. So we won't be able to max out our characters before A18? Challenge accepted.
  9. Am I suppose to have batman + underwear in my profile pic until A17 drops or something.....batman underwear is everywhere... Does it drop when A17 drops?
  10. I'm a simple man, I just want the spider meme.
  11. It's been what 8 months since Alpha 16.4? We can wait just a few more days for a video...…………………...or riot.
  12. But man would Unity's new ECS improve 7DTD pretty much could remove the overhead of the voxel engine if implemented right. But Faatal is right it would take months, to implement ECS (which would go along with the job system) you have to relearn how you look at writing code in Unity unless you're familiar with data-oriented workflows. You would then have to rewrite and restructure entire systems from scratch; and on top of that ECS isn't fully implemented so you would have to use a hybrid approach up until 2019 when ECS would be production ready. The render pipelines may be easier since it is finished and shouldn't involve re-writing any code, but I'm not an artist and don't write shader code so im not 100% on that. Would be sick to have 10,000 zombies through ECS though. Next game....next game...
  13. Interesting, I assume each voxel is counted as an individual object so that might be why, I'd think it would be better for large objects over many small ones. Not my direct area of expertise but maybe try batching or instancing textures to be shared among common objects in the scene on the fly so you don't stream to each individual object but stream to the batched group that way FPS is only impacted by changes per object instance but overall groups of objects sharing textures stream from the same source until otherwise changed.
  14. @Madmole, if you want more freedom with textures look into one of the texture streaming solutions for Unity, as a Unity dev myself I can say they help a ton as far as texture memory management goes. For Unity I know of two solutions: Granite: Web link: http://graphinesoftware.com/products/granite-for-unity Asset Store Link: https://assetstore.unity.com/packages/tools/granite-for-unity-indie-51026?aid=1011l3btC Amplify Texture 2: (PC only) http://amplify.pt/unity/amplify-texture-2/ Unity can achieve far better graphics even with a voxel world just look at this voxel engine for example: https://forum.unity.com/threads/terrainengine-voxel-terrain-smooth-cubic-2d-hexagonal-infinite-procedural-terrain.174595/ ..side note, UMA is gross switch to MCS better for texture memory and looks way better; at least for player characters, it would only take a day or two to switch out then you just convert the UMA clothes to MCS which is a simple process, shouldn't break animations either. #4k graphics 7DTD 2018?...don't let texture memory hold you back! -Tolin Simpson, Owner/programming lead, Leatherbound Games
×
×
  • Create New...