|
Rocko Bonaparte posted:Is there a Unity asset for tile-based 3d level editing? I was hoping to define 3d tiles from models, define how they interconnect, and then be able to draw walls and paint floors with them like they were otherwise 2d tiles. I did this with the first game i tried to make and I used an external map editor called Tiled. Performance sucked because my tiles were too small and I had a ray-based fog of war system (also instantiating that many models was slow so load times sucked rear end), but it wasn't hard to get working.
|
#
?
Jun 25, 2023 07:57
|
|
|
|
| # ? Jun 3, 2024 12:51 |
|
|
https://crocotile3d.com/ maybe?
|
|
#
?
Jun 25, 2023 09:24
|
|
|
Crocotile is the only 3D modeling (/tiling) software that makes sense to my spatially-stunted brain, though trenchbroom is probably the most technically correct choice if it makes sense to you.
|
|
#
?
Jun 25, 2023 15:06
|
|
|
Rocko Bonaparte posted:Is there a Unity asset for tile-based 3d level editing? I was hoping to define 3d tiles from models, define how they interconnect, and then be able to draw walls and paint floors with them like they were otherwise 2d tiles. It works ok. You want 2d tile editing software for layout, or some other data that you can turn into tiles (eg map data). Then you need a poo poo load of tiles to connect different biomes/elevation tuples per edge. If you're clever, you can manage some of that with rotations (and you should because pooling will be important for performance). As far as making the tiles themselves, just load up your 3d modelling software (or have someone do that for you) and make the tile you need. Last time I did a project where we did this, the system spat out "missing" tile definitions (4 edges of biome/elevation), which we could then put into requests to the art team.
|
|
#
?
Jun 25, 2023 16:16
|
|
|
We just put up our Steam coming soon page for Goblin Camp ( https://store.steampowered.com/app/2431980/Goblin_Camp/ ). It's a fantasy city-builder inspired by Finnish mythology. To this day no-one has made exactly the right kind of city-builder/colony simulator so I figured I have no choice but to do so myself. Feels great to have the coming soon page out so I have a place to point people at when I'm talking about this. Conveniently I also haven't changed my avatar in 10 years and it's now topical again.
|
|
#
?
Jun 26, 2023 13:36
|
|
|
NICE. Wishlisted!
|
|
#
?
Jun 26, 2023 13:42
|
|
|
Aryoc posted:We just put up our Steam coming soon page for Goblin Camp ( https://store.steampowered.com/app/2431980/Goblin_Camp/ ). It's a fantasy city-builder inspired by Finnish mythology. To this day no-one has made exactly the right kind of city-builder/colony simulator so I figured I have no choice but to do so myself. Holy crap I remember when you started making that. You're still working on it!? Edit: Holy graphics batman
|
|
#
?
Jun 26, 2023 14:56
|
|
|
Aryoc posted:We just put up our Steam coming soon page for Goblin Camp ( https://store.steampowered.com/app/2431980/Goblin_Camp/ ). It's a fantasy city-builder inspired by Finnish mythology. To this day no-one has made exactly the right kind of city-builder/colony simulator so I figured I have no choice but to do so myself. Is this the same ascii Goblin Camp game that I still have from 2010?? Or is there a new goon game called Goblin Camp
|
|
#
?
Jun 26, 2023 14:59
|
|
|
Looks great! Wishlisted also
|
|
#
?
Jun 26, 2023 15:02
|
|
|
gently caress SNEEP posted:Is this the same ascii Goblin Camp game that I still have from 2010?? Or is there a new goon game called Goblin Camp This is the same game in the sense that Doom (2016) is the same game as Doom (1993). It's a game about a bunch of goblins that find themselves in the middle of a forest and need to build themselves a village, protect themselves from all kinds of danger and start some fires while they're doing it. So a lot is the same but this time it's going to have real 3D graphics (wow!), mechanics that have considerably more thought put into them, and all in all just a much higher quality game. We did our best to come up with an alternative name but Goblin Camp just feels right.
|
|
#
?
Jun 26, 2023 15:06
|
|
|
I am finally almost done with the core game mechanics of the idle game I've been working on: I can not overstate how much of a motivation boost it was to actually get eyes on the game and feedback, I've gotten more done in the past 2 weeks than I'd gotten done in the previous 6 months (granted, I did dedicate a full 4 day weekend to it). Aryoc posted:We just put up our Steam coming soon page for Goblin Camp ( https://store.steampowered.com/app/2431980/Goblin_Camp/ ). It's a fantasy city-builder inspired by Finnish mythology. To this day no-one has made exactly the right kind of city-builder/colony simulator so I figured I have no choice but to do so myself. Oh my, I remember following the ASCII version way back. Definitely going to be watching this.
|
|
#
?
Jun 26, 2023 21:55
|
|
|
Hammer Bro. posted:Crocotile is the only 3D modeling (/tiling) software that makes sense to my spatially-stunted brain, though trenchbroom is probably the most technically correct choice if it makes sense to you. Trenchbroom looks like a brush-based editor from what I get with all the Quake stuff going on. I did that a bit when I was younger so it wouldn't be alien to me. I suppose I don't necessarily need to stick to tiles, but I was figuring some kind of tile-based editor would be faster to use. Maybe not. Crocotile looked the best so far. I guess Unity has some built in stuff to do it? I saw some videos but I wasn't sure how much lifting was required to get to the point of doing it in the Unity editor.
|
|
#
?
Jun 27, 2023 08:59
|
|
|
Deki posted:I can not overstate how much of a motivation boost it was to actually get eyes on the game and feedback, I've gotten more done in the past 2 weeks than I'd gotten done in the previous 6 months (granted, I did dedicate a full 4 day weekend to it).
|
|
#
?
Jun 27, 2023 12:32
|
|
|
So I've finally managed to take the advice "make a small game" and I just want to make a clone of desert golfing basically because I think that game is the perfect example of "do one thing really well". The problem is I am falling at the first hurdle of what to actually build it in. I have some experience in Unity, not going to use Unity though because I just find it horrible to use. Godot seems good but I have to learn it, love2d seems good but I have to sort of learn it (a lot of pico8 dev and it's a similar structure and lua). My real life job is a ecmascript dialect so something like photon might be worth delving in to but man I really don't know. I just haven't actually made anything in so long my morale is totally shot so I really want to get somewhere with this project so I want the path of least resistance. Maybe I'll go with haxeflixel because back in the day however many years ago that was where I was most prolific. Basically what advice would you give someone who is a professional non-game developer, has dabbled in game dev for 20 years but never really finished anything (normal for most hobbyists I bet) and really wants to make this one thing? Part of it is making this statement of intention to try and give myself some kind of i've forgotten the word but yeah. The next stick is picking a weapon which will probably (and should probably) be Godot.
|
|
#
?
Jun 27, 2023 21:59
|
|
|
I'm a firm believer in never reinventing the wheel. You know Unity, make a Unity project, it's not like you're at a level where this has to be production ready and stable for release you're just making something fun. But it sounds like you've already made your mind up with Godot and I'm also a strong believer in just following your heart and it's not like you're doing this to put food in your stomach either.
|
|
#
?
Jun 27, 2023 22:03
|
|
|
al-azad posted:I'm a firm believer in never reinventing the wheel. You know Unity, make a Unity project, it's not like you're at a level where this has to be production ready and stable for release you're just making something fun. Usually I would agree with you but I think Unity has some part to play in my motivation dropping off in the last few years (along with career changes and parenthood). I find the thought of using it puts me off actually getting started with anything.
|
|
#
?
Jun 27, 2023 22:16
|
|
|
al-azad posted:I'm a firm believer in never reinventing the wheel. You know Unity, make a Unity project, it's not like you're at a level where this has to be production ready and stable for release you're just making something fun. I feel like Unity starts your project out with square wheels these days, kind of forces you to reinvent them. For what it's worth, quitting Unity and picking up Godot made my productivity shoot through the roof. Agree with the sentiment though, sitting around and pondering about it doesn't help much. All options are viable, even Unity, so long as you have the discipline to just sit down and get poo poo done.
|
|
#
?
Jun 28, 2023 15:35
|
|
|
Curious what specifically about Unity you find so bad. It's not a perfect engine, or company, by any means, but it's fine imho, and none of the alternatives are better enough to make them worth it for me. The only exception is Godot, which I find really intriguing, but I would hesitate to use it for anything I didn't plan to release for free or nearly free on itchio
|
|
#
?
Jun 28, 2023 17:39
|
|
|
When I started a game project in Godot last year as a novice programmer (after having spent some time working off of learncpp.com to try to fix an otherwise abandoned open source game), I was surprised at how easy it was to get into. I have to imagine that for an experienced coder, it'll be a cinch 
|
|
#
?
Jun 28, 2023 17:41
|
|
|
Polo-Rican posted:Curious what specifically about Unity you find so bad. It's not a perfect engine, or company, by any means, but it's fine imho, and none of the alternatives are better enough to make them worth it for me. The only exception is Godot, which I find really intriguing, but I would hesitate to use it for anything I didn't plan to release for free or nearly free on itchio I've been using unity full time since 4.x. The biggest problems I've seen are that it's extremely opinionated about workflows, but not good at guiding you to the good workflows. So a lot of people build towers to the sky because they don't want to accommodate the workflows supported by the engine or aren't aware of them. That and they ran out of useful features to work on, but instead of polishing UX they kept adding new features. And there's a handful of things that they _do_ go back to but are just garbage because they try to dumb things down below the base complexity or the people building them don't understand the domain or both (eg their networking solutions)
|
|
#
?
Jun 28, 2023 17:48
|
|
|
Polo-Rican posted:Curious what specifically about Unity you find so bad. It's not a perfect engine, or company, by any means, but it's fine imho, and none of the alternatives are better enough to make them worth it for me. The only exception is Godot, which I find really intriguing, but I would hesitate to use it for anything I didn't plan to release for free or nearly free on itchio Honestly I couldn’t tell you a specific thing. Stuff like the endless progress bars. The 2.8gb base project size. Probably more than anything it’s that I have so many unfinished projects kicking around in it that it’s a morale dampener. Though I don’t have the same feeling from pico8 so maybe not. I decided to start in godot earlier today so fingers crossed I’ll have something to share soon.
|
|
#
?
Jun 28, 2023 17:50
|
|
|
leper khan posted:I've been using unity full time since 4.x. The biggest problems I've seen are that it's extremely opinionated about workflows, but not good at guiding you to the good workflows. Yes, this is my #1 biggest problem with Unity... you can do just about anything in the engine, but good luck googling how you're actually supposed to do it... LeFishy posted:Honestly I couldn’t tell you a specific thing. Stuff like the endless progress bars. The 2.8gb base project size. this is another huge one! I had a handful of empty projects for testing purposes and they ended up filling up my hard drive >:-(
|
|
#
?
Jun 28, 2023 17:58
|
|
|
Unity really needs to do thread archiving. Googling a problem and finding a thread that was posted in 2012 is not going to be a good solution. It might work but it's definitely not going to be best practice in the modern engine.
|
|
#
?
Jun 28, 2023 18:05
|
|
|
Does Unity still require you to process all assets before you can load into the Editor? That's not a big deal for small projects, but was probably the biggest pain point for large games. In 2012, when I last worked in Unity full time, their suggested solution to that was to pay for the remote cache server upgrade. I think the longest I was waiting was a few hours before we switched to the cache server. I remember talking to devs at other studios on larger games that had 24+ hour waits to load the editor the first time when they couldn't get the cache server to work. It also used to not preserve processed assets between changing deployment targets. So if you swapped from iOS -> Android, you had to reprocess all assets, then when you switched back to iOS, you'd have to reprocess everything again. Chainclaw fucked around with this message at 18:08 on Jun 28, 2023 |
|
#
?
Jun 28, 2023 18:05
|
|
|
Chainclaw posted:Does Unity still require you to process all assets before you can load into the Editor? That's not a big deal for small projects, but was probably the biggest pain point for large games. In 2012, when I last worked in Unity full time, their suggested solution to that was to pay for the remote cache server upgrade. This problem has ironically gotten worse and more complicated with assembly definitions, which produce an additional damned-if-you-do damned-if-you-dont lever for compile time vs link time on any change.
|
|
#
?
Jun 28, 2023 18:14
|
|
|
I'm a complete newbie, but I've tried Godot on and off, and I like a lot of it, but some parts of the UI are weird and hard to use like the animationplayer for me. I've been toying with Gamemaker and it has actual released commercial games like Nuclear Throne and somewhat more straightforward guides since its not iterating as much as Godot is trying to.
|
|
#
?
Jun 28, 2023 18:56
|
|
|
I don't think animation is easy in any engine, unfortunately. It's fiddly by definition.
|
|
#
?
Jun 28, 2023 19:31
|
|
|
Polo-Rican posted:Curious what specifically about Unity you find so bad. It's not a perfect engine, or company, by any means, but it's fine imho, and none of the alternatives are better enough to make them worth it for me. The only exception is Godot, which I find really intriguing, but I would hesitate to use it for anything I didn't plan to release for free or nearly free on itchio I stopped using Unity in like 2018, so maybe they've fixed a few problems but a lot of it is like core to the engine. The biggest issue I kept having was with the physics engine. I make mostly platformers and stuff, and the default physics are just woefully incompatible with that type of movement. I'd constantly have game objects bouncing/sliding off of or over penetrating different colliders. I eventually found a tutorial that went through the process of making your own physics to clamp objects to the floor so they don't fly off of slopes or do weird poo poo. But then you use Godot and it just has all of that custom stuff built into the KinematicBody, by default it's just there already. I've also found the component system to be too open ended for me. There's so many different ways to split data and functionality up, I could never settle on one method of handling it all. So I spent most of my time in Unity refactoring my code into whatever my new standard was. Which is partly on me, but also a problem I just have not had with Godot. Having just a single script per object is just way easier to manage even if I have to add some occasional jank for edge cases. Having to establish everything as a prefab was also just way too much work for me. The vast majority of games these days need to dynamically spawn objects into a scene, so I don't know why Unity insisted that I define every single prefab that was going to be needed per scene. Godot just let's me load a Scene file wherever and whenever I want, no need to predefine it. The final straw for me though was the way they handled the networking updates.... I started working on a big multiplayer coop game around 2014 and I got really far, but it was kind of a learning exercise because I had just started using Unity. Then they announced that they were deprecating the networking code for new networking code called UNET. So I ditched my learning exercise project and started fresh with UNET and worked on that for about 2 years. Then they announced they were deprecating UNET, the thing they literally JUST added to the engine. Not only was it being deprecated, they were also decommissioning the Matchmaking and Relay servers that they recommended everyone use, so if it took me another 2 years to finish the game, then it would be completely unplayable after a year or two unless I overhauled the entire networking layer (which was not easy because a feature of UNET was coupling the networking code directly into gameplay code). BUT THAT'S NOT ALL! Unity did all of this without actually replacing the networking code that they just deprecated... Instead they just said that the networking code was "going to be pretty similar to the low level API", but it wouldn't be ready for another year. If I remember correctly, they didn't even make that deadline. I waited an entire year and there was still no networking replacement. About 6 months later I found a weird github page where they had an "alpha" of the newest networking layer, but I have no idea exactly when it came out because no one seemed to notice it. So yea, I never really liked Unity, but them making me restart a project 3 times and then sit on my hands for a year while they worked on a hypothetical replacement was too much. I haven't looked back once.
|
|
#
?
Jun 28, 2023 19:46
|
|
|
Polo-Rican posted:this is another huge one! I had a handful of empty projects for testing purposes and they ended up filling up my hard drive >:-( Meanwhile my ue5 folder is approaching 300gb for the engine build alone... 
|
|
#
?
Jun 28, 2023 20:03
|
|
|
Unity has one of the best compression methods I've seen and I've never had a final build over ~150mb unzipped which kind of makes me side-eye any really big Unity project. I had to have a seasoned technical artist explain to me how to remove unnecessary modules from Unreal 4 and it was still 4x larger than it should've been. The problem I see with Unity as a platform is that it's like playing in a sandbox but you forgot your toys so you have to bum hand-me-downs from other kids. I never had an issue with it until I wanted to make interactable grass that I could paint using their terrain tools. But the terrain system is so arcane my only recourse was either using polybrush (which I wasn't going to hand paint several square kilometers of terrain!) or scour the internet for the tech wizard who found a cheaty hack that let me place my custom grass using the terrain's splat map system which was only possible via a 2022 update so I had to upgrade anyway! Unity is fine until you run into a very specific issue then you need a dedicated software engineer or crowdsource a solution.
|
|
#
?
Jun 28, 2023 20:07
|
|
|
Tummyache posted:I stopped using Unity in like 2018, so maybe they've fixed a few problems but a lot of it is like core to the engine. this is a big one that's bizarre to me, because you absolutely do not want to use the physics engine for the majority of games but unity puts it forward like the thing you should be using by default. tons of tutorials and stuff you find through googling too even for the most basic super mario poo poo uses the physics engine and just dial in some random numbers until the jump is "mostly fine" etc and it's not at all sustainable to be expanded into a full game. to add insult to injury the build in kinematic controller (i think they called it character controller?) was extremely outdated, buggy and just... not good. the only realistic option was to pay for a third party kinematic character controller on the asset store, which is what i ended up doing for my 3d platformer project and it was absolutely night and day. everything just worked, collision detection was good, slopes worked properly, moving platforms, everything. it cost way too much money but at the time it was the only option (i looked it up and it seems the asset is no longer developed or maintained but at least it's free now...) it just feels like having a working character controller is one of the reasons you'd use an engine over writing all your own collision code by hand and it should be one of the most basic features offered but... Unity. with that said i also haven't used unity since 2018 so if there is a new up to date and actually working kinematic character controller built in now someone update me on that
|
|
#
?
Jun 28, 2023 23:25
|
|
|
al-azad posted:Unity has one of the best compression methods I've seen and I've never had a final build over ~150mb unzipped which kind of makes me side-eye any really big Unity project. I had to have a seasoned technical artist explain to me how to remove unnecessary modules from Unreal 4 and it was still 4x larger than it should've been. This really comes to team and game size. In 2012 at least, we were making Unity games with between 10 - 30 engineers and artists per project, and a dedicated tools and tech team supporting many Unity games. Because it was mobile, keeping size down was important and we had to build a ton of our own tooling to manage that. We also had to build our own tools to examine the relationship between content so we could figure out what was referenced at what point in games, so we could push things to the cloud to download on demand later, but include the first ~2-5 hours of gameplay in the app store download. Even with all that tech, we still had to resort to dumb mobile game tricks like downsampling the audio to hell and making a lot of it mono instead of stereo. I think nowadays, if you're putting more than 10 people on a game, I would choose Unreal or something else over Unity, even for mobile.
|
|
#
?
Jun 29, 2023 00:29
|
|
|
The main reason I stick with Unity at this point is that as crappy and weird as HDPR is, it still does a better job of rendering nice-looking large scenes than Godot does. I've basically locked my entire codebase away from Unity in anticipation of hitting a point where it's just not worth it anymore. I tried porting over to Godot as an exercise, and about 90% of my code can literally copy & paste between engines. The only stuff I'd have to rebuild are the classes bridging my poo poo to the engine's poo poo.
|
|
#
?
Jun 29, 2023 02:03
|
|
|
Chainclaw posted:It also used to not preserve processed assets between changing deployment targets. So if you swapped from iOS -> Android, you had to reprocess all assets, then when you switched back to iOS, you'd have to reprocess everything again. I use Unity full time and generally like it but this is still true. If I think I am going to be working on platform specific features I will usually pull a repo twice so I can have an Android folder and an iOS one. Switching between them has gotten faster over the years but I still like to avoid doing it if possible.
|
|
#
?
Jun 29, 2023 02:21
|
|
|
Unity's in a pretty good sweet spot of having a lower learning floor than a lot of other engines. C# also has a lower learning floor than other programming languages like C++, while still being powerful enough to get a lot done.
|
|
#
?
Jun 29, 2023 02:40
|
|
|
I'm deep into my semi-homespun network code for this 2D space game I'm building but need some critique and feedback on the implementation of this system. I'm really concerned about client simulations diverging from the server's authoritative state, so I've been working on some code to allow clients to validate their state against the server and rollback+replay if needed. However, I've overengineered aspects of it or missed some key tricks. My game has space ships that are constrained to a 2D plane and act with ~Newtonian physics. The server has a fixed tick rate for physics (configured between 25ms to 100ms per physics frame), and at each tick it processes queued player commands and runs the physics simulation for that frame. All of this happens over ENet / UDP with reliable packets (for now). The physics process takes player commands for (for example) acceleration and applies that to the player's ship based on various properties (thruster layout, mass, maximum allowed velocity, etc) Clients may only ever send commands like ACCEL_FORWARD or ACCEL_REVERSE and their current frame number to the server, which will: 1. immediately reply with an input acknowledgement packet, containing an echo of the client frame number and the server frame number at which the input is applied 2. send a snapshot of all players' kinematics for the frame after the simulation updates Meanwhile the client applies the commands locally and runs the simulation while waiting for responses from the server. The server is always authoritative, and fixing up the client's state when it's wrong is where things get a little funky. Right now, the client: 1. keeps track of it's previous kinematics for each frame, for some chunk of frames (as a function of latency to the server and the tick rate) 2. keeps track of the server authoritative kinematics for each past frame, for some chunk of frames (same as above) 3. keeps track of which client frames are associated to which server frame (per the input acknowledgement packet) 4. keeps track of all previous inputs over some time period. Each physics frame, the client looks at its server packet history and checks to see if the client/server still agree. If the simulation starts to diverge*, I want to roll back the client to the last tick where the client/server still agree and reapply inputs until it's caught back up using the last known server frame as a starting point. In the rendering frames, I would interpolate between the currently displayed bad position and the newly re-calculated good position. * the simulation can diverge for a number of reasons, but for the most part the floating point math in the simulation is basically okay since i'm using 64-bit floats everywhere. one of the big reasons for divergence is latency spikes which cause the server to not receive any packets during a few ticks, but continue calculating position/velocity for the player during those ticks sans acceleration input. the client has meanwhile possibly been applying acceleration in those frames, and the whole sim goes to poo poo  I've been reading a lot of the Glenn Fiedler networking posts, Gabriel Gambetta, Valve Source networking, etc. Wondering if I'm totally off the mark here or if this is a sensible approach. I certainly feel like the client is holding onto a lot of different buffers of old data, but not totally clear to me how to simplify.
|
|
#
?
Jul 1, 2023 05:30
|
|
|
Lazy pathfinding https://twitter.com/bonerman_inc/status/1675022695438880768
|
|
#
?
Jul 1, 2023 07:15
|
|
goku i won't do what u tell me
|
Congrats, you've made an enemy movement system that actually allows the player to run away from fights.
|
|
#
?
Jul 1, 2023 14:41
|
|
|
After 3 years of the in-game combat being keyboard/gamepad only, I've started putting together an optional mouse-driven UI: https://www.youtube.com/watch?v=jJ0Pashb-YQ Layout/UI here is placeholder, my artist collaborator is going to send me some assets to make this look better. If you have any ideas/suggestions though let me know!
|
|
#
?
Jul 1, 2023 22:04
|
|
|
|
| # ? Jun 3, 2024 12:51 |
|
|
Making another Blender importer for another underwhelming 1999 third person game, this time Ultima 9, based on the specs on the series' wiki (which are not quite correct at times): The terrain was reasonably easy but I can't seem to figure out how to correctly access mesh data yet, so no buildings. ---- KRILLIN IN THE NAME posted:Lazy pathfinding I mean, it sounds lazy but tracking which tiles and which enemies have line of sight to the player involves more work than you'd think and functionally makes it a sort of dijkstra variant too. The core interesting idea of it is a sound classic though, that if you're managing a ton of things with a common target then actually tarting from the target will allow you to share the search. Chev fucked around with this message at 23:02 on Jul 1, 2023 |
|
#
?
Jul 1, 2023 22:56
|
|