⊖ Thread has reached reply limit.
READ THE RULES
>#8/agdg/ via irc.rizon.net
>https://matrix.to/#/+agdg:matrix.org via matrix programs
>Dev resources: http://8agdg.wikidot.com/resources
>Previous breads: https://archive.vn/BQHzj - https://archive.ph/vi3yT
rate my track
groovy, but i think the snares are too loud and the rest of the song is too quiet, had to turn my volume up to max to appreciate the main melody
Nice like ice.
i rendered it on my machine where my volume settings are all fucked up, need to figure out what a good baseline volume is for everything because it's all over the place across my projects
Don't actually hear it with your ears, just check it in something like audacity or whatever audio program that shows you the decibel levels, you want it to hit in the middle of the yellow zone and the loudest thing should barely touch the red zone
>Regardless of whether you consider them to be self important jack offs, the threat has teeth behind it.
No this doesn't seem to be the case, this section appears to talk about either trying to impersonate "red cross using" organizations or trying to appropriate the symbol but not just displaying it. On top of that I find it hard to believe that at least in the US with its vast protections for speech and art simply depicting a symbol would be disallowed.
>tfw there's not enough good monster girl models available to make a coherent game with
I'm terrible at making my own models, and I want to just make demo, no matter how disjointed. I guess if push comes to shove I can justify it with some bullshit.
It's pretty good. I think >>136638 is right and you need to work on the volume of the main track.
do like the snes and change palette
Did the feds get to the Matrix room? Element says that it's no longer supported, but doesn't give any correction.
the matrix room has never worked properly, i think someone just threw it up there but forgot about it the next day. probably wouldn't hurt to have one going.
here's another one i've started, took about an hour including getting distracted and fucking around. volume should be better on this one.
oooo Ominous, although I think it is a bit repetitive, it's mostly due to the tempo I'm used to fasted paced tunes transitioning into different melodies faster and i'm sure the full song wouldn't feel like that at all. I'm sure it fits on a gloomy or dark level that opens up into something challenging. It's pretty good.
Volume seems to be good for most instruments as well, so well done.
What program do you use?
Can you maek gaem if you can't do math?
Yes but you still need to read a book or two.
Say you're storing 2d positions. Floats have ~7 significant digits. Doubles have ~15. That includes the part before the decimal. So a value like 1280.123 can be represented in float. That's like each pixel being a 100x100 grid. Double can represent 1280.12345678901. So each pixel would be a 10,000,000,000^2 grid. The lack of precision with floats can be a problem. Use doubles to represent 2d points.
most advanced math I've seen so far is physics vector math, which is jewish and evil and a false abstraction of reality but not hard
I suck at texturing and I have no clue how to make it better ;_;
Do you guys know of any damn search term I should use in a search engine to look for FPS games ammo item models? I've tried doing that and I only get shit result, I want to make models that doesn't look like fucking wooden boxes, but I'm too brainlet to think off a better shape.
Woah! I fucking LIKE IT. Pls make full version of this.
It seems its because of recent Matrix/Element upgrade where rooms are now managed by spaces, similar how Dicksword handles its "Servers", I just checked the link and it doesn't work for me either. However there is #8chan+:matrix.org instead which I manage now after the Anon guy left, I had to create a new space because he forget to change the ownership of the space to me. I pretty much almost administrate all of this 8chan matrix room.
I barely know shit about math and with Godot 2D you still can go pretty far with this as stuff like position and rotation is straightforward, its only in 3D when shit starts become bullshit tier.
I feel tremendous despair seeing anons making progress
this link should work now https://matrix.to/#/#8agdg:matrix.org the thread and the wiki needs to be updated to reflect this change.
Make your own progress sperg
Just use a green cross instead of a red one? Or put a square around the cross?
You could try looking for games on https://www.textures-resource.com and https://www.models-resource.com for various shooter games and see if they have the pickups, then stydu them from any angle. Might take some digging though, for instance I checked Half Life and Halo and they didn't have much.
>in 2022 A.D.
You don't recolor the cross, you extend the ends at 90 degree angles. I hope getting associated with "literally evil" was worth it for the fags in the Red Cross organizations.
I just looked it now on models resource and there is barely any interesting ammo models that I can get "inspired" from it. God damn it.
>"Those responsible may be unaware that use of the red cross emblem is restricted under the Geneva Conventions for the Protection of War Victims of 12 August 1949, and that unauthorised use of this sign in the United Kingdom is an offence under the Geneva Conventions Act 1957."
>in the United Kingdom
Of course it would be. Being British is a crime in the UK.
>hmmm today I will make game
<10 hours of procrastinating later
>wow what a hard day at work! better hit the hay!
How Do I end this cycle? I really want to make a game but not have duke nukem forever tier development hell. Every time I go to do something or attempt to make a game I just get sidetracked and fap to loli or do something else
>using the red cross is a crime
So that's why pharmacies use the green/blue cross? I thought it was just some gay meme but its true you can go to jailerino for using a stupid symbol. any other stupid shit you can get V& for in the united cuckdom?
>any other stupid shit you can get V& for in the united cuckdom?
Making offensive twitter posts. Also in Cuckmany you can get jailed/fined too for making offensive twatter post.
>fined/jailed for making offensive twitter posts
And the people haven't gone full 1488 yet why? you'd think by now getting jailed for your speech would cause some civil war but people just seem to roll over and accept it?
I guess those people don't care or are so conditioned to accept this state, I bet most of them don't know how America is the only country with a true freedom of speech. Cuckmany has in theory something similar called "Meinungsfreiheit" but due to 6 gorrillion hate speech laws it may not exist even. What a fake and gay country.
>How Do I end this cycle?
Make a to-do list for each week.
Oh dog there still more crate version left to do. So far I have done this now:
SmallCellBox (Cell), LargeCellBox (CellBox)
SmallShellBox (Shell), LargeShellBox (ShellBox)
SmallBulletBox (Clip), LargeBulletBox (ClipBox)
SmallRocketBox (RocketAmmo), LargeRocketBox (RocketBox)
StimpackBox (Stimpack), MedikitBox (Medkit)
What's missing is these now:
HealthBonus, ArmorBonus, GreenArmor, BlueArmor,
SoulSphere (?), MegaSphere (?)
And some kind of a chest for when the player clears the map with 100% kills or 100% items, I'm not doing a bonus for the secret one because some maps have some obtuse bullshit secret. The difference I want to make is that items drops from monster gets put inside the inventory (lol) instead of being picked up directly hence why I made models of those items.
I already have a basic script system for that working, though I didn't write yet a scoring system which also takes for more accounts such as +missile +missilemore +boss flags and the size which affects the monster score too and thus its loot drops. I guess this mod make the player even more OP but eh I don't care, almost all of my playthrough with the Lambda mod I did felt relative easy anyway and I don't care as I just want to kill shit.
Though I have no idea now what I should do with those 3 colored boxes now, maybe I guess I scrap those models instead, hmm.
I'm here to shit up the thread. I'm assuming most posters here are probably from the Discord and already saw this stuff, but I'll drop it here anyway.
Main game (ANUBITEK) is no longer a tactics game, moved to a Mega Man Legends and Sin And Punishment inspired squad shooter that will have rail shooter mechanics. I couldn't figure out how to make the tactics mechanic fun or engaging in any way that wasn't frustrating or would require over complicated GUI. Started making the training room/demo room in TrenchBroom, I think I'll have a demo out by the end of the month if I can consistently work on it.
I also started a rocket jumping puzzle platformer called BoomeRocket, I plan to have most of the maps made for it by today and to have it ready for sale by tomorrow. That game is almost done.
I have a Twitter, YouTube, Newgrounds, and Patreon up for these but they haven't really went anywhere yet because the progress isn't significant enough. Some engagement though, which is a start.
I had a game dev job for a bit that fell apart, been almost 3 months without pay so I'm back to regular wagecucking again. I wont be paid in time for rent so it might be ogre, but I'll still dev at the local McDonalds or whatever.
Who did you steal this from?
>rail shooter mechanic
So it contain no exploration and the map progression is linear only?
I've been around for a while, haven't posted in the threads because they got really shit for a long time. Don't want the spirit of agdg to die though, so I'm back. This is my game, I could post screens of anything from it.
The game will be split into to segments, free roam and rail shooter. Planning to do something with chapters similar to Left 4 Dead, but every other chapter would be switching between the two. While in a rail shooter segment, you'd be on one of these.
To be honest, why would they? Banning speech has been done since time immemorial and most people don't have need for it anyway.
"Being a bigot is not an opinion" I am sure you've heard that one.
sage for /pol/
Stop being a discordnigger.
inb4 muh crabs /agdg/ is le ded
Hardly anyone uses the Discord anyway. We'll see how long I stick around this time, I'd prefer to stick around but I can't read programmer wars and /pol/ posts all the time in a game dev thread.
Have some concept art, so I'm not just shitting up the thread myself.
Hmm I'm not exactly fond of rail shooter, I like Doom style of combat/exploration the best. Still I wish you the best of wishes with your game.
>I couldn't figure out how to make the tactics mechanic fun or engaging in any way that wasn't frustrating or would require over complicated GUI.
That sucks, care to elaborate a bit more what was hard to make?
Maybe you should stop bickering around and post progress instead?
>muh discord is bad
Then fucking contribute to the Matrix room already instead of letting it to rot.
>"Being a bigot is not an opinion" I am sure you've heard that one.
I think I heard something similar to it, though I can't recall its phrase. In Cuckmany they would say shit like "X (some internet site) is not a rechtsfreier raum (legal vacuum?)" and other such crap despite I'm sure that internet sites like faceberg or twatter is hosted in America and not in Yuropoor, so which means Cuckmany laws wouldn't apply to them.
Though I'm pretty sure the bureaucrats over at Jewrop Union would say otherwise.
Welcome back. I admire your dedication to constantly redesigning your furry game, and not giving up on it through engine changes.
I had a whole write up, then my internet died and I lost the post so I'll just greentext.
>squad management was a mess because engaging in combat would send all units to a nearby square, which would sometimes send units off an edge or into a hazard/trap/field of view of NPC
>turn order means waiting for NPCs to do actions then passing input back to player, I hate this about tactics/RPGs in general
>too many abilities to balance against each other because some modified others
>trying to plan levels around mechs being able to be on the board was time consuming because I wasn't using TrenchBroom to make maps yet
>I wanted platforming to be a primary focus with stealth/tactics gameplay being secondary but couldn't figure out how to incorporate tactics without game flow being awkward. If one unit was spotted, all nearby units would go into tactics mode and if they aren't in position like they should be then it would just frustrate players.
>I don't like turn based combat or tactics games lol
These all might have had easy solutions to them, but these were problems I simply didn't want to script around.
As for turn based you could maybe make it like in UFO Aftermath series where actions actually happen in real time but all actions still require player input so the soldiers won't absolutely do anything unless ordered to do so or do it like in X-Com apocalypse where they can have manual actions applied to them or they act on their own.
LMMS, i build it off the master branch occasionally. took two days to get all the deps in order but I think it's been worth it. they've been undergoing a giant "unfuck everything" phase for about two years now so the last stable release is years old and lacking lots of features/fixes.
these are just clips, i have the basic structure for most of them but don't want to post them entirely since i don't want to get bullied.
most of the songs i'm doing are based on late 90s/early 00s eurodance to fit with the Y2K aesthetic of my game. i actually started working on them almost six years ago and my game kind of grew out of these autistic fantasies i had in my head that i built around the music i was making, until one day the idea just flashed in my head and i thought "hey i can do this". but enough blogposting, if you like my shitty music that's cool.
<shitting on crabs
I have an Idea!
What If we remade CS but open source? some guy already tried making CS in godot and almost got it right, so im thinking its possible! I was thinking that it could be a mix of all 3 games with the best of each, the mods of CSS, the comp of 1.6 and the weapon variety of CSGO. Any codefags want to make a prototype? I could make some models and animations within a week, i gotta learn how to animate in blender and the rigging. Only question is how much muh graphics do we go for? 1.6 levels of graphics or source? it needs to run on toasters AND look pretty so we could go for something like condition zero the middle ground.
Thoughts? I just want to make something fun and open source for linux users and toasters similar to how open arena made quake 3 accessible to all I think something needs to be done with CS or similar games because fuck valve, fuck DRM and fuck kikery
What if you just pirate 1.6 and play it right now instead of being an useless ideafag?
Networking mostly. If you have to find a coder and have to learn animating, you're already in the hole for maybe a couple of years of development. Would it be worth it? CS 1.6 is already dying/mostly dead.
>pirate a dead game filled with brazilians and russians
For what purpose?
>would it be worth it?
I think bringing a game into the free and open source realm is worth it, sure you can pirate the 1.6 but that only works on wangblows, I want a linux native game that doesn't run like ass and doesn't need valves DRM to play
I have a system that allows for characters objects to load modules containing mechanics, from inventory, to movement, this character nodes are instantiated by the game core, and the game core controls the rules of the current game, however, the game core does not contain any content, everything is loaded separately.
The current problem is how I will be load the player character, do I make a character selection screen, like in UT2k4/Timesplitters, but keep hidden from the player on the final release, that was the original idea, but I don't want to implement a necessary parameter for the player character in the game core, and I want to re-utilize system in rhw future for custom characters.
I think will parse an argument to the game core, telling it to load the model and animation set for the player, and for the final campaign core, I will add argument in player spawn to load a specific model.
I would make an input module for your character, then make one version that gets input from keyboard/mouse/gamepad, and another that gets input from your AI.
>pirate a dead game
>For what purpose?
Anon wants to play CS, he should play CS. OSing existing stuff takes years of hard work and hardships, it's not just a "oh do this for me would be really cool" kind of thing.
>linux native game that doesn't run like ass
Aren't there builds of GoldScr that work natively in Linux? Obviously not release binaries, but there's ought to be linux releases in steam for 1.6 or at least Half Life.
That is exactly how it works, I have the player controller, NPC controller, module controller and player module controller.
The character node is a separate object, in memory, that can be controlled by both NPC and Player.
How many times are we going to go through this cycle?
You try to learn to code.
You ask people to build your game for you.
Do you know the definition of insanity?
More sounds + music added, physics bugs squashed, targets now activate goals, and more maps have been introduced. Working on map 4 now, I plan to have at least 30 of these.
Last few things to work on are a main menu/options, enemies that slowly move toward and try to shoot at the player, laser walls that block player only, and kill barriers for lava/etc. Then it is just a port and ready to go on play store or wherever. I might sell a pc version too, I don't see why not. I'm worried about a high refund rate though, that's why I'm looking to pack this bitch with levels.
how do I stay motivate
Start a project then just work on it. Add stuff you want to see in a game, if you're unsure how to do something the go out of your way to either ask other aggys for advice or join some discord servers. I know people here hate discord, but there are enough communities on there to help you through things or give you advice on ideas. I've found that joining retro gaming servers like arena shooter servers has given me some of the best advice I could have received.
Post progress, don't shy away from mistakes. Also ignore people who get mad at everything and anything for no good reason, they aren't people as far as you're concerned. Just Like Make Game, if you can't do that then there is the very real possibility that game dev might not be for you. Or a myriad of other conditions, like maybe you're just not inspired or you're a lazy SOB. Get yourself on a dev schedule and try to maintain it, post even if no one looks at it. Eventually those posts may add up to something.
Welp that's how it looks like in game now, I got all those Items implemented and working now, some of those items drop a bit more than one item depending what it is and its based on chance value too, so they don't always drop the exact same amount but there is at least 1 item being always dropped.
Later I need to look into translating the score drop system I made in Godot to ZScript so that it uses a similar system, loot drops are already determined by score but so far its done by using certain sets of loot drops. Also I need to come up with a better formula for determining score too since most of it is based on absolute value instead of something that scales.
I also added height/radius check of the monsters so that Lost Soul monster don't drop ridiclous amount of items (I would rather prefer it should drop none but I have no idea how to make a condition check that only works on lost soul and any monster pack mods that adds a lost soul replacer), which fucking sucks.
Some people liked the models that I have made so far but in my opinion some of the items such as the small clip crate and the rocket ammo crates could still be improved, though I can't think of any better design yet atm, meh.
Someones gotta shit up /agdg/ and that someone is ME.
t. eden pro
Bleh, I will never make maps in Blender again, that program fucking sucks dick for that.
What kind of maps are you trying to make?
If you're looking for smooth exterior terrains and you're using Godot you could trys this:
If you wanna make interior maps I would use Qodot which is what >>136785 is using.
Basically I need map editor which either has good support for terrain editing or at the very least lets me import model formats such as dae/obj so that I can use other tools which can do terrain editing. Trenchbroom doesn't let me import any form of models which is a big bummer since I like the editing workflow of Trenchbroom.
>what type of maps
Basically I need a mix of terrain and structures which may or may not have interior design.
>godot addon for terrain
that looks interesting, I'll take a look into it.
Have you considered using the height terrain addon for all the smooth terrain parts and then just placing structures or Trenchbroom levels on top of it? You don't need to do it all in one editor.
If everything fails there could still be level transitions between interior and exterior levels if needed.
Hmm I could try, but it sounds like then when I would use Trenchbroom for structure/interior design I would have to constantly fiddle with the sizes a lot because either the structure is too big or too small when I would place it in Godot. Because I have no means to check in Trenchbroom if the size of the brushes looks right, or at least I can't think of any way how I should do it.
So if I understand it right I would have to do it like this to make my maps?
>Make terrain in Godot terrain addon editor
>Keep adjusting the terrain until it looks good enough
>Make model objects in Trenchbroom
>Place them around with Qodot
>Because I have no means to check in Trenchbroom if the size of the brushes looks right, or at least I can't think of any way how I should do it.
There's a setting in all Qodot Map Nodes which is called "Inverse Scale Factor" which is what's used to convert Quake units into Godot meters.
I use 32 because that's best for my texture size and what I'm going for which just means that 32 Quake Units translates to one meter in Godot. TB comes with dev textures so just place a couple of them with a 64 unit height (assuming you're using 32 inverse scale) around the map so you know roughly how tall a person is. Figure out a general size that is right for you.
>Make model objects in Trenchbroom
>Place them around with Qodot
That doesn't sound right but then again I don't know what you're planning to do. Just make the levels or objects you need in Trenchbroom and then load them as Qodot Maps, then translate these to wherever you need them on the terrain.
I don't know why you'd use TB to make simple objects though since that sounds inefficient.
Also I checked again and you can actually export TB maps as OBJ files.
Basically I want to do something similar like the UT2004 maps in my Game which for onslaught maps its a mix of terrain and brushes/static meshes. Which if I am not mistaken the UT2004 map editor includes both so you can pretty much make all form of maps with that map editor which is what I'm looking for. I hope the provided screenshots are clear enough what I'm trying to say.
Yeah, just make the terrain with the Godot heightmap and then place the models you make with Blender and Trenchbroom around in Godot.
So how would I go about making pre-rendered backgrounds for something like resident evil? is it really just as simple as taking a picture of an existing level and then layering it on top of the cameras viewpoint?
Work on anything.
Doesn't even have to be game.
Lift, make healthy food, clean your room.
That will get you the basics, but you may also want lights in the scene that match the pre-rendered background so your "live" models are lit properly.
Realistic shadows are trickier, since you need some kind of proxy geometry for the shadows to be projected onto.
Looks promising. Kudos to you for actually making something.
Working on unfucking code, instead of having a static camera attached to the player in each scene it creates a camera and sets the params once the player is loaded in. Also allows me to set specific spawnpoints for each scene. I should've done this weeks ago but I have a habit of procrastinating like a motherfucker. Including writing this post.
>I should've done this weeks ago but I have a habit of procrastinating like a motherfucker. Including writing this post.
I have the habit I just end up playing games instead of getting anything done, meh.
Demo out here:
oh good it fucked up the link, cool.
god that took way longer than it should've, there was probably an easier way to debug that than going through every single level from the start each time but it's done now. you can now seamlessly go through each stage and show up in the other one where you're supposed to be. maybe i'll finally add enemies tomorrow.
Always good to get rid of hardcoding for more flexibility
actually i probably should've set them up as node points in the editor so i can drag them around and not manually enter coordinates each time the scene layout changes. i keep thinking of better ways to do things that invalidates all my previous work.
>people are making progress and are motivated to do more
OY VEY It's anuddah holocost!
TrenchBroom .map to Qodot .tscn conversion is a little strange, the chain/grate textures and image texture had opaque backgrounds when converted but you can handle individual materials by a couple ways. Also the chain meshes had two be two brushes that have their backsides clipped, putting chain textures on both sides (front and back) produced some weird UVs as a result.
haha I forgot the images
>The class "XYZ" couldn't be fully loaded (script error or cyclic dependency)
I love when Godot spills this error, it slowly tips away from my willpower, fuck Juan.
Yeah the errors cascading through other scripts are annoying, since any custom gdscript class that has a compile error will cause an error anywhere that class is referenced.
Try restarting Godot if you think you've fixed them all and it's still happening.
Fortunately Godot is really quick to restart.
Oh great, I'm getting a error with this plugin on Godot v3.4.4, rip terrains.
I tried to instantiate the same class inside its own function, that was causing the error, but it was showing on another script.
Works for me on a fresh project, installed it through the asset library.
Does anyone know how I can create team/player selection in godot? there are no tutorials on it but I assume it goes something like
>create button with Join game as text
>once in game the player faces another menu for character selection buttons and spawns in as a spectator
>if you pressed button A to become X class or just for a player model you spawn in as that class/player model
Am I right or am I wrong? Could I do the same thing for weapon selection and other things in game? so if the multiplayer is asymmetrical or has class based game play or the game has a commander mode you can just use a few buttons and spawning in as that specific role? Also how would it work for a team based game? Do i do the same thing with the buttons and do something with the player node to differentiate it from the enemy team? I Can think of a few fun game modes that would be fun to put into my game but I have little knowledge on how to think about them in a practical sense, for a game mode would it be
>press button in main menu to chose a game mode
>spawn into level with mode already affecting the world
>spectators given different choices based on the game mode of choice on which way to spawn in
I'm following a few tutorials in godot and have some basic knowledge on how things work and I assume people are waiting for the networking features of godot 4 before making anything serious such as a multiplayer FPS game. I would use godot 4 myself but there are no good resources on GDscript 2 out there or maybe I am retarded and missed them on the godot website?
I just want to make something relatively basic and fun then once I get to grips with the engine make something more complex. I'm wondering how I will be able to do the different game modes such as gun game or bag tag or virus whilst still being able to play on the same map but with different variables. can someone give me a quick rundown on how to think about game modes and their nodes?
I think for selecting classes you could make one scene per class, script those player class accordingly, use a common object to inherit from to avoid some code duplication and variables too.
For my 2D game I did used a janky method where I wrote a function that a couple of nodes gets removed and loads other nodes from other scenes in order to implement tank chassis swapping.
Heres how it looks like:
#TODO: Move this to a seperate script
Swapping = true
# var Vehicle = Item["Vehicle"]
var SScene = load(Item["Scene"]).instance()
var Nodes = null
Turret_WeaponSystem = SScene.Turret_WeaponSystem
#The Partial nodes swapping
Nodes = SScene.get_node("Chassis")
ChassisSprite = Nodes
Nodes = SScene.get_node("Turret")
Turret = Nodes
TurretSprite = Nodes.get_node("Sprite")
Nodes = SScene.get_node("FireSkele")
Turret_Firepos = Nodes
FireSkeleRoot = Turret_Firepos.get_node("root")
Nodes = null
Swapping = false
Swapping = false
As I said previously its not a good example of a code and there is a lot of room for improvement.
Baking lighting, after these are done (4 more rooms to do) I'll be going back to scripting. Demo later this week if everything stays on course.
You will NEVER finish your game
Your game will never get published
Your game is a shit
You WILL stay unmotivated and demoralized
That's weird, I enabled the plugin again and now it works, though is there a way to have shading + grid rendering? Its right now hard for me to tell how the terrain is shaped because everything looks white with barely any shading applied to it.
You will finish your game
Your game will get published
Your game is AWESOME!
You will stay motivated and full of ENERGY!
Your waifu loves you and counts on you!
Shoo shoo Schlomo
Why are you replying to yourself?
Geh fuck this, trying to edit the terrain in Trenchbroom is such a pain in the ass. I have to create a single brush and copy it manually 6 gorllions times till I get the size I want, then once I double click on the brushes to edit the vertices for some reason Trenchbroom doesn't always register my keys and it doesn't move the vertices upwards instead it always moves in left/right - front/back directions. I constantly hit the shift and control keys and it rarely moves the vertices upwards which is highly annoying for me.
What the fuck is this shit. At least with Blender it doesn't fuck my keys that bad. Meh still I won't use Blender for making maps, it sucks too much for it.
My game is already in the state that it could badly need maps and for unknown reasons there is no decent general purpose map editor for linux available. I tried also DeledCE program which seems to have have only windows binaries so I have to run it through wine and with Lutris 6.4 wine the camera lags badly, making the program unusable for me.
What's the problem? Making terrain maps is a terrible idea with TrenchBroom. Can't you just make a subdivided plane in Blender and use proportional editing and smoothing for slopes?
It shouldn't be that hard.
Even if I would do that, the other aspect of Blender is highly unintuitive for making maps with, such as trying to model structures with interiors, I tried making one and I gave up how Blender constantly selected faces when I wanted to move the camera.
>how Blender constantly selected faces when I wanted to move the camera.
I never have this issue as you cannot select faces when moving the camera, using middle mouse and shift+middle mouse you literally cannot select faces when moving the camera around, maybe its something wrong on your end but I am able to make interiors just fine
I assume embeds don't work, but this is a Plutonia Experiment wad edit I had been working on a while ago. 7 maps were finished, added some new music tracks. This is played with Brutal Doom Sperglord Edition.
I want to make EEnE:O a reality! What would be the best way of going about doing that? Do I have to re-watch every episode to find out what items/equipment they used, places they've been, or is there a wiki for that? What type of combat should I use? I don't want to do what ever other MMORPG does and use shitty gookclicking with number pressing for abilities
First, be blamed for a school shooting by federal police.
Consider playing through it with hideous destructor Final Doomer.
Has anyone made any mods for duke nukem forever yet? didn't it come with full source code or am I mistaken?
What I mean is when I do interior design and I want to move the camera the faces gets selected instead of moving the camera which makes it for me highly annoying, also the lack of FPS styled control and Trenchbroom styled grid snapping for modelling makes Blender less than ideal for me to make maps with.
>lack of FPS styled control
They changed the flycam shortcut to ^` (shift + grave accent)
>What If we remade CS but open source?
You know Warsow exists? There are at least 2 other opensource fps games out there.
Then what are you doing in this board?
What's the best way to learn music theory?
I reckon is not dicking around with my guitar.
Also dicking around on your guitar.
Neat. They even give you a piano, how about that.
Figured out how to project viewport textures onto a panorama sky texture. Going to be using this to make Source-style skyboxes.
Music theory is gay, you don't really need it to compose music. Just like you don'[t need to be a linguist to speak english.
I kinda want to make music that doesn't use western piano scale with 11 notes to double frequency, but like 14 or 9 instead. Gonna need to make my own software for that though.
You could just sample generated tones and put them in a tracker.
Just like a true Scotsman.
I think my game is too big. checking the design document, the amount of environmental assets that I will need to finish, I'm not sure if its feasible for 1MA within the time space that I'm looking for, its not even the complexity of the models, I'm want something similar to the models of Unreal Gold and Clive Baker Undying, I just want more in the same scene.
There is the option moving to another genre game, I already have ideas of 3 other possible games that require less models, a procedural beat 'em up, a shmup similar to Tyrian, and Mount & Blade with mechs, each one having a different levels of complexity and their own set of problems, but they generally require less models than what I'm currently working and allows me to recycle textures and meshes, two of them however require good music, can't imagine playing a beat 'em up or shmup without some good d&b or jungle, it feels off, and I don't know how to compose, and I don't want to expend money hiring composers.
I have no problem writing code, is just asset creation, maybe I'm just afraid of the magnitude of the work load, maybe it is feasible, I don't know.
dont worry about creating assets, coding is the hard part, making assets is really easy unless its a 2D game with lots of sprite work but even then you can still use 3D models then make sprites of them to fill in the gaps.
>mount and blade with mechs
Unironically do it faggot, sounds like a really cool idea
Just learn how to into music, its not that hard once you learn the basics then you just string everything together!
Depending on the assets, you can get a lot of stuff under CC0 online. You would want the hero models like the mechs to be custom and consistent, but stuff like a generic car model or a tree could be something you get offline and modify rather than starting everything from scratch.
Where do you people get your 3D assets from?
I'm planning on making everything by hand, that is why I'm worried about the work load.
You can also kitbash with CC0 or public domain models, Smithsonian has a library of public domain models.
That's a lot of bones.
Baked lighting screens, though this map has to be scrapped and rebuilt again due to some geometry issues. It will look the same, just some small fixes to be made.
Where do you get your textures from?
On a more dastardly note, do you know of a way to get 3d models off of those sites that ask you for money for free? I don't mean to use them for commercial purposes but it makes my blood boil that you cannot find a good model for specific instances unless you fork over an unreasonable amount of money.
OpenGameArt.org and various CC0 sources off Startpage.
>Hardly anyone uses the Discord anyway.
That's surprising to hear. I know all the yes-devs and guys who contributed to the wiki silently backed away. I assumed they all went to Discord and Twitter so they could be in a yes-dev utopia but it just seems like that energy fizzled out completely.
>I can't read programmer wars and /pol/ posts all the time in a game dev thread.
I've been checking in on and off and the programmer wars died years ago. I don't know about /pol/ posts but the self-sabotage shit is worse than it's ever been because all the people who are actually serious left causing it to condense in on itself.
I used Element/Matrix all the time before and I wasn't even aware there was a /agdg/ dicksword.
I just tested out the Git version of Trenchbroom and it seems despite the patch notes mentioning Quake 3 patches it won't let me create one which if I am not mistaken seems to be used for terrain, that sucks because I really need the damn terrain already.
Yeah. It's been a thing even since 8chan's days. It makes no sense to join though because even if the yes-devs who went there were the most talented posters on here once you open up the floor to massive platforms like Discord or Twitter they're pretty below average in the scheme of things. At that point why sabotage yourself and make an inferior game by associating when you could just get in touch with people who can give you skills and resources to make something actually worth playing?
I must have missed it then.
I only decided to use it because then I get faster help related to ZShit, but since I had a debate with those chucklefucks and they got mad at me they banned me from their cabal channel so there is for me no reason anymore to use Dicksword.
Meh I used GNU+Social/Pleroma for over a year and then stopped using it, I absolutely loathe this fucking twatter clone format as its just terrible for having conversation with and with almost every post I have made there virtually none gave a single fuck about my shit, fuck that.
>a game worth playing
I don't care if others will like my tank game or not, I made this game for myself and I only made a gitgud repo for it in case anybody else might find something worthy to salvage my shit code.
Got off my lazy ass and added compound collisions to my game objects so I can simulate concave colliders with rigidbodies
During the stress test I lost about 300 frames but honestly that's to be expected since I haven't really optimized the game for that level of intensity and don't really plan to atm
But I'm impressed with how well the physics engine is keeping up with how I've handled the interpolation with my prop_physics script.
The way that script works, when any object with interpolated physics (such as the player) collides with a prop, that prop will have interpolated movement until 3 seconds after it's no longer being touched by an interpolated object. This also implies that if an interpolated prop_physics collides with another prop_physics, the other prop's movement is also interpolated until things finally cool down.
Just thought that was kind of neat.
>Also is this the game that shall not be named or is it your own doughnut steel?
I haven't been very active in these threads for a while but which game are you referring to?
This is not Peripeteia but I can see how the lighting might resemble it. I'm still experimenting with the atmosphere of this project but one really big thing this project and Peripeteia have in common is that aesthetically both projects are inspired by the Half-Life 2 beta and take place in Dystopian post-modern hellholes.
Are there any good old 3D games that Could be remade with todays technology? What game should i remake for you fags? Im thinkg about games made 2000 and before, anything made over 2000 doesn't look too bad unless its some console shit for the PS2 with its shitty hardware.
I'd like to see a remake of C&C Renegade someday that isn't just a janky MP project.
Some fuggen /cow/ tier shit right here.
make one with a proper skirmish system so that I can play alone and kill stuff with all the C&C Renegade guns.
I remember some ideaguy in here was talking about how he wanted to make stalker but with anime waifus instead. Everyone laughed him out of the thread and he got real asshurt. I wonder if he became one of the peripetia devs because it's a similar concept just sci-fi.
Why havent we made this a reality yet? Stalken with heckin cute grills sounds like a blast!
I use this thread as an anonymous, or semi anonymous public record of the game that I'm making, however, I cant post screenshots, because all that I have is programming related progress.
No one gives a shit about an implementation of a command system,it's just a tool, there is not pzazz to it, but its an important tool non the less.
Working on skyboxes now. Might make a video on how to do it.
4agdg has been winning, do your part to make better games.
No! You're going to steal my genius idea!
>All the yesdevs leave
>Crabs are left behind
>Suddenly the thread sucks and the yesdev area thrives
Who woulda thunk?
>Sounds like a good idea because it seems like all the insufferable faggots here are all people who hate anime or constantly going into shitposting mode.
No one hated anime they just spammed bait in every thread until someone bit and you got gaslit by it. All the guys who made threads for games like Stalker left years before that happened because they got hit first since Mark didn't moderate them at all. Eventually it spread to every thread outside of /agdg/ and Mark started doing mass deleting of posts for people he banned because suddenly it was affecting him. Mark only banning people who upset him, ignoring the rampant baiting and attention whoring killed the board. 8chan was unusable the two years it was up because every thread was full of people referencing deleted posts and bait.
>Just be careful because there will be some attention whore that will try to ruin your life on discord. Happened with the 4agdg one.
That type of shit happens here too. Look at how many people have badmouth peripetia, tetra and other devs just because they were somewhat successful. It's a cycle that keeps repeating.
> the two years
The last two years*
Pic related, it's you.
Hot dogs anon, any progress from one of the half dozen yesdevs posting in this very thread?
I can disregard entire posts because of nitpicks too
Also in my day "yes-dev" meant someone who actually finished a game which I don't see any of here.
>Look at how many people have badmouth peripetia, tetra and other devs just because they were somewhat successful
yeah dude I'm sure that's why
>multiple devs start posting concrete progress
>no stupid technical arguments, just questions answers and cool shit
>thread gets nearly a hundred muh crabs posts in the span of three days
>starting immediately after a faked school shooting
You're doing an awful lot of complaining and not a whole lot of deving.
>I assumed they all went to Discord and Twitter so they could be in a yes-dev utopia but it just seems like that energy fizzled out completely.
More like drama and general faggotry. It just turned into the same indie-clique bullshit that always happens. A number of people just slid away, either being driven off by the largest dramawhores to do their own thing or on their own choice.
Judging from your post history I assume you like FPS games, right?
So I would suggest made a copy of Terminator: Future Shock if Doom and Wolfenstein3D are too basic for you.
The game uses rough 3D models of robotic enemies so maybe that should be easier to make.
How goes tankgame dev, is he still kicking? I'm interested in /k/ projects like that. Seems promising for now.
P.S. Is it locked on digimon or legallydistinctfrom!digimon monsters being the enemies, or is his other idea (pull a Spellcross) still considered? I still want to run orcs over.
Pretty sure he's c0423d but I might be wrong.
Here is the skybox in action, plus a water shader. If I can't figure out how to fix this weird seam issue, this might be an abandoned feature unfortunately.
Looks like a UV problem. Can't you make a half-sphere in Blender and use that as the skybox model instead of trying to do it in Trenchbroom?
TB isn't well suited for making organic and seemless geometry since that isn't its purpose.
>How goes tankgame dev, is he still kicking? I'm interested in /k/ projects like that. Seems promising for now.
I did nothing for the past few days and I just played GZDoom because I got bummed out with the lack of good map editing tools for Godot, now I'm trying to do it again with Trenchbroom and settled for Doom but in 3D styled maps instead, its not ideal but meh what can I do about it.
Right now I'm stuck in the issue that I cannot test out my new map to see if the map scale is good enough because fucking Qodot is crashing on me with a *** stack smashing detected ***: terminated error and Godot lags like a motherfucker when I try to import the map as obj.
If only Volvo weren't bunch of assholes and released a standalone loonix/proton compatible version of Hammer 2 then maybe with this tool I would have better luck.
>or is his other idea (pull a Spellcross) still considered? I still want to run orcs over.
Looks like you to wait for that one longer then, since I want to finish this current idea first before attempting next idea. Once my base game is finalized I'll try to modularize it as much as possible to make modding as easy as possible, I already separate Base and Core content into its own folder so when I did everything right a mod should be able to unload most of the Base content and provide its own content so modders can add create their own TC with my game.
And as for Spellcross since I enjoyed this game much despite I haven't played in ages because I hit in a major roadblock I still consider to make a Spellcross themed addon for my game since I still like the premise where "Other Side" a force with magical abilities all of suddenly appears and causing havoc world wide.
I too want to drive to a destroyed town, where the sirens is crying and cleansing the town off all the orc scums.
Though the Spellcross part will be harder since I would have to learn how to make Character modelling and animation, so I guess I still would have to attempt it somehow even if the result is far from ideal, better than nothing I guess.
The issue isn't the UV, it is that the viewport texture isn't wrapping properly when rendering to the sky. The skybox itself is fine, though there are some UV issues as can be seen in the video.
Oh, I misunderstood then. Tbh I have no idea what you're trying to do here. If I were you I would probably just make the skybox a model that's placed inside the map. I wouldn't bother with using a second viewport at all. Seems like some kind of weird solution to me.
The idea is so the skybox moves with the player visually, so you never get closer to any edge of it. It is a solution that Unreal Gold and old Source games had.
Animated skyboxes are staying in, found a 360 camera class.
Nice! Good job finding that.
Does anyone know how to do different aiming methods in godot such as those seen in red orchestra with the dead zone aiming or like in MGS with the gun wobbling about the place in first person view? I've already seen some tutorials on how to make a basic FPS but having the gun be static in place is very boring!
You mean they don't do that normally?
Spitballing here, untested, but what you could do is make an FOV model that has a 3DPosition node offset on the end of the gun and track that screen position to maybe get relative firing vector of the gun on a bone attachment point (or however you want), then you could do some procedural animation stuff with an animation tree to offset where the gun would be. Then you'd just use the input from the player, probably with some hotkey and mouse movement, to move the gun around. I'd be interested in seeing something like in Metroid Prime/Goldeneye/Time Splitters where the gun moves around based on where you aim, maybe even view roll like in Quake/Turok 2.
Do mean, weapon sway or the camera will only move after the weapon passed a certain buffer within the viewport ?
camera will only move after the weapon passed a certain buffer within the viewport, something like the weapon moving along the screen but not focused on the center of the screen and is able to shoot at any point within the dead zone or in MGS case the weapon moving whilst you shoot around the screen also not focusing on the center unless you ADS first
Just calculate if the mouse position is within the buffer area, and use the difference from the center of the screen to the current mouse position as a data for the projectile angle.
I've written my own system which goes kinda like this:
>bullets are spawned at a "bullet spawner" spatial node at the barrel of the weapon
>firing function randomly selects a rotation for X and Y coordinates within a negative and positive accuracy range
That's the basic version. Here's what I also added on top:
>have a minimum and maximum accuracy rotation float
>accuracy change variable that increases the current accuracy rotation range until it hits the maximum
>current accuracy returns to minimum if you don't fire
>change-factor multiplier that is also applied to accuracy change each shot which you increase or decrease based on player's movement stance (1.0 when standing, 0.5 when crouching, 1.5 when moving for example)
I'm very happy with how it feels although I still haven't gotten around adding a proper recoil system yet.
Speaking of which, does anyone know any good recoil systems? I'm not looking for code but rather a logic since I wanna implement it myself.
Nice concept art! Did you decide on coloring yet?
>late 90s/early 00s eurodance
Fuck yes, please post more.
There was an EEnE game for the PS2/PC and it was glorious, play that first for inspiration.
who drew this? what game is it for?
Color scheme will be somewhat close to hiw Fallout Tactics Deathclaws are colored.
I haven"t asked them about crediting yet, so I'll ask before I answer yet. The art is for my game, ANUBITEK. Here is a photobash they did. Someone else made the model(s).
Reminds me of Nightmare and Shockwave a bit
I hate how godot will only register collisions if some object is moving, I need to check for a ramp collisions even if the velocity of the kinematic body is zero.
How feasible is it to make a game in Godot using C? I have a game engine written in C and SFML. It's 3.5k lines of code. I want to port it to Godot for its features and to be able to use 3d. I tried Raylib but it didn't work for my needs. Should I learn GDNative or GDExtension?
In my engine, game logic updates at 1khz and the draw rate is configurable. If the game lags, it drops frames instead of slowing down. This is essential because replays need to be deterministic and the motion of entities must be depicted as clearly as possible. Is it possible to have this kind of setup in Godot? Will it limit me to 60hz?
as far as main game loop goes, the default is this:
>_physics_process() is the fixed loop, default 60hz, you can set it to any high number i think
<hard to find actual information about how it deals with dropped frames if the hardware can't keep up, but it's implied that if the game lags, it skips frames instead of slowing down and instead increases deltatime to compensate (which isn't determinstic it's fucking variable)
>_process() is for rendering / matching the render framerate, should use the delta time given
you can create your own game loop, you can get by with using gdnative but as always you still have rework your written code a little bit
Remember that battle against a giant enemy crab actually happened!
>Fuck yes, please post more.
since you asked so nicely
it's supposed to be boss music clearly, so I don't want to get carried away and go into something completely different, it should loop cleanly and not get annoying to listen to a dozen or so times if you keep dying to it.
also here's a little fuckaround track I did to get the "feeling" of eurodance. i mostly used webm related as a musical reference and because it's comedy gold, i love that "in your face yeah some crazy shit is going down fucking deal with it" stone-cold serious mentality that all these 90s music videos have
Good start, your tracks remind me of 90's tracker MODs that play in demos. With some polish (and audio level normalization) they can become solid vidya OSTs.
Cheesy, but nostalgic. It's an abstract kind of feeling watching this...
It is based on the mech from Solatorobo, has magnetized arms for swinging around or shooting. Idea is that any ability you can use as a player will be amplified in the mech, to make rail shooter segments more fun.
Idea is that the characters in my game (called Boltrs, Boltr singular) are industrious and militant, with safety being a secondary concern to all else. They made the mechs as their answer to all-terrain navigation. Some also wear HND-CNN devices (play on the phrase "hand cannon) which use proprietary chips containing schematics to manipulate matter and produce various outputs like a nail gun, a rocket launcher, shock traps, etc. They also figured out how to rewind time in a small area and cause the results to become congruent to reality, basically figuring out resurrection by restoring matter in a localized space to a previous state of existence (hence ANUBITEK).
Why not just add a very small velocity to both objects so they're both moving at all times?
Are there any genres (You) would like to see more of on /agdg/? for me it HAS to be DMC clones, i wish a nigga would, goddamn
Point and click, or general adventure games that are not "talking simulators à la TellTales". It seems they have fallen out of fashion big time.
Crazy action tank games or good shooter games that are inspired by Doom 1/Half Life 1 but with good selection firearms, good weapons balance and monsters that scale well. So basically easy to moddan shootan games. I like shooting at things.
haha shotgun goes boom.
RTS games that actually incorporate basebuilding and Diablo-likes. I think both of these have enough potential to either be given a nice unique spin or innovation without requiring a big team or even a team at all. For example I think you could put a lot of love into the map generation of an ARPG and how the game treats drop rates and difficulty for late-game players even as a one man team.
I want to see a game that gets finished.
Games that people actually will spend money on. I'd actually be making games if marked wasn't oversaturated and hostile.
Could work, but just want to move on, I shoot 4 rays and get use the angle to check for ramps, I'm tired doing hack work trying to circumvent Godot's shitty physics system, Bullet Physics is not a good alternative because of the jitter, objects move up and down, its a mess.
The worst part is that the contributors knows that the physics is broken, and they refuse to fix because, reasons, I think a stable physics engine this is more important than the vulkan render, a less buggy game is better than an ugly game.
I will be honest I don't any hope for Godot 4.x because they are still using Godot Physics, maybe GDExtension will allow the creation of bindings without a headache, I really want to creating a Newton Dynamics binding for Godot.
>Physics: Bullet: KinematicBody compound shape collision fix (GH-56801).
Maybe there is hope
Here you go: http://8agdg.wikidot.com/general:finishedprojects
List would be longer but a lot of guys who finished their game didn't post it here when they were done with it.
>play alone and kill stuff with c&c guns
I mean. Isn't that just the single player campaing?
If the guy that did Among Stars and Robots is here, how many copies did you sell total? Did you do any kind of advertising beyond just throwing it on Steam?
I need more ideas for buildings and billboards. You anons got any?
Any game that could be in the bonegaem64?
Looking good anon. I like the zombies. What does the chink on Ken's Kannibal Kitchen read?
>I need more ideas
>Libbie presidential campaign
>"Canned Spotted Dick" and "Christmas Tinner" ads
>Hotdog cart tuning and hot rodding
>Cheese store run by Nazarin
>Bones n' Noble: Pathway to Reading
>Intervertebral Land Noir
>Osmosis Jones for grave bacteria
>Going Postal Services
Your express delivery to the afterlife of your choosing!
>Obvious gas station
>Billboard of Weird Al dressed as a lawyer
Got a problem?
I'll sue ya!
Got a problem with a rich asshole?
I'll sue for ya!
>Bonegaem64 - Dr. Calcium Doctorate 2: Fracture Rapture
>Jellyfish Hill 3: No Bones Allowed
>Graveyard 64: Rave in the Grave
>Bonezone of Enders
>Skulls: For Skulls Sake!
>Battle Joints: Let Us Clack Together
>Exo Squad: Bugs in the System
>I need more ideas for buildings and billboards. You anons got any?
Gamestop, but it would be MonsterStop "Power to the rapers®"
Well my FPS game's progressing nicely but now I've gotta deal with the biggest hurdle and that's developing an enemy AI.
I've been procrastinating for a long time but I think I figured out a good way to do it. I'll try to go down the evolutionary path of FPS game AIs, starting with basic Doom enemies and hopefully finally reaching FEAR at the end of it.
This'll take a long time but at least I have clear steps in between that'll be functional.
Wish me luck, guys.
What kind of FPS game you are going to make? You better have a good Shotgun/SMG weapon in your game else I won't play your FPS game release.
I'm not gonna say too much because from experience I know that talking about your ideas isn't productive or helpful. I'll better use that time to dev in silence.
All I can say is that I'm going for something mid-2000's similar to Half-Life.
As for shotguns? In general the quality of an FPS can be measured by its shotgun. So that means the shotgun better be damn good. I'll be dedicating a lot of time to make it perfect. But that's for another time.
I don't want to use a shotgun all the time though in close range, that shit is getting boring since Doom 2 with its SSG spamming.
Fucking hell I hate Blender 3.10 so much, what the fuck did those damned morons do? I want to know because every time I open that piece of shit my whole fucking computer lags and everything becomes slow poke this is not funny, do those asshole excepts me to buy 32GB of RAM or what? Fucking piece of shit.
You best be using bullets and not hitscan boy.
Use whichever version was best for you at the time. If the current versions keep messing shit up, stick to what works.
I haven't even programmed hitscan attacks yet. So far all weapons use projectiles.
The thing with Blender 3.10 is that some of the plugins that I need such as W3D importer only works with this version, not with 2.79 so the plugin devs shafted me out of this one. Otherwise I stick with 2.79 most of the time.
They're focusing on features rather than performance.
>do those asshole excepts me to buy 32GB of RAM or what
Welcome to modern software, where the programmers let the hardware people fix their shitty code for them. To give the hardware people credit, they actually do really good job, it's just that modern code is soooooo shitty.
Upgrading my toaster is currently unobtanium for me since I don't heff enough money for dis ;_;
What gamemaking engine is best if I want to make an RTS?
Get Godot, Unity, and UE4
Get squad pathfinding working in all three of them.
By this point you will know which engine you like the best. Throw away the work in the engines you dislike and keep going with the one you're the most productive in.
Have you tried blender 2.8x? The LTS versions may be more suitable for you.
Install 32 bit version stupid. Max 4g ram requirement.
What a twat, so the 32bit Blender 3.10 version uses less then 500MB of RAM instead of around 1-2GB of RAM? I have bunch of other shit running such as Gimp and Godot at the same time so I have less RAM available for other programs.
Nope I haven't tried it.
Well it states 8GB RAM as the minimum requirement on the homepage, not sure what you have going on there.
>8GB minimum of ram
>when older Blender version could and did use much less RAM
It's always a can of worms what the program can use and what the devs pick as a safe minimum to guarantee that it works. That said of all programs I do expect a 3d modeling one to be among the most intensive in terms of resource consumption. Do you use swap?
Yes I have 8GB of swap and even that is not enough.
Imagine making a game and only provide a deb file and cuck windows users. That would be funny.
You need to pay modern games to even touch your game nowadays.
I figured as much, the other programs are likely too active for the system to consider swapping them out. You can try changing your swappiness value but beware that this does not actually change how aggressively the system swaps but rather what it swaps.
Wall running, map preview, music track. Needs work, buggy, but working.
Are there any tutorials out there on how to make a combat system like half life? I still don't know how they made it so you can have multiple items in each numbered slot, i think I can get around to coding the health and armor pickups, the inventory is the only thing giving me any hassle, no idea on how to pick up ammo or weapons automatically and have them be assigned a specific slot within the inventory or how to make weapons have an alternate fire mode, are there any suggestions on how to improve upon the systems of half life anyways? for things like keys or essential items like those found in sven coop. im looking to combine the best of both HL1 and HL2s systems to give the best of both worlds. once I get the basics down I hope to make a few levels then release a demo
I don't understand how you don't understand.
what i would do is have an array or dictionary of values such as gun possession, ammo count, key ids, etc and just increment values or modify booleans based on collision with objects scripted to modify said values. then just set control binds to toggle gui elements and cycle through the list of elements that are enabled/disabled depending on ammo count or item availability.
item 1 in dictionary is set from false to true on pickup, increment item 1's ammo count, enable equipping if switching via gui/keybinds, or enable auto-equip which should just automatically call a pickup function in a gui script. for health/armor, add some values to a player scriot and just use a signal to update either or when the player takes damage or gains either, use a formula based on a damage/heal signal to modify player health based on armor percentage. this is all very simple, so get to it anon.
The Loonix kernel is retarded and swap is irrelevant when a process requests another huge chunk of memory, the kernel will first sacrifice the entire disk cache for it and that will slow down the whole system. For Blender, try limiting undo levels to a low value. It should be in the User Preferences somewhere.
What game is that?
That looks like fun, will there be a demo later?
If he's on loonix he might as well just use the 32-bit version of blender, but he'll have to get it from his distro's repos because the last 32-bit version offered by the official website is 2.80:
Soon. Road map now is getting shooting working with animations again, AI to follow the player, patrol behaviors, dispatch behavior (think MGS2 dispatcher), more ammo types, then the demo is ready. I was pushing for a week ago, but I feel just being able to platform isn't enough for what I'm trying to present the game as.
Am I the only one who noticed most of the guys posting progress are cross posters from 4chan's /agdg/ discord?
Starting to realize how daunting it's going to be to make a GDD, since the whole idea behind it is that it should be your every word and thought if you're not there to describe it to someone working on your game. Everything from how movement should be done to how functions should be called and what should be attached to what needs to be described, It's like I have to do everything twice, once for writing it down and another for actually implementing it.
if they keep their faggotry where it belongs then i really don't care. if i don't notice it it's not a problem.
Meh some minor progress with my village map, modelling the brushes to specific shapes can be such a hassle. I didn't textured anything yet because I need to sort my textures stuff first so that its not over 3-5k in textures because Qodot seems to have odd method to load the textures (I'm just going to haphazardly guess it is because it loads ALL of the textures instead of textures of that is being used by the map)
Wait, I didn't know you had on-foot sections.
I didn't implemented on-foot feature yet, I hope Cinema4D is really easy to make animation with so that I can do that soon. Though I have no idea how I would handle the other crew members because I would need gud AI for that and I'm not good at math either, fugg DD-----:
Give me an idea for an FPS game, or at least what FPS game I should be ripping off. I need a project to do whilst I learn the basics but I also want to create something fun
Exotic Bullet Types only
And not some smoke and mirrors extra damage with stock fx bullshit,reality better start crying whenever John YesDevMan pulls the trigger.
> I have a game engine written in C and SFML
Isn't SFML written in C++? Surely you mean SDL
Rip off C&C Renegade with freeroam maps.
multiplayer class based shooter where all the characters are arachnids
The formatting on the site messes up every time some one mentions Cee Plus Plus. You need to type C + + (with spaces) or Cpp.
Is there any GOOD description about how to make a simple RPG turn based combat? I've been trying to implement something on my own, but right now depending on how I tweak my numbers, it's either piss easy, unbeatable, or pure RNG.
The problem isn't your balancing, the problem is that simple turn based combat is largely shit. If you have a limited number of options then the game becomes a solved problem. At that point you are no longer making any interesting decisions since you can always make the best decision. If you can always make the best decision, then the game can only be one of three things: piss easy, unbeatable, pure RNG.
If you want good turn based combat, you have to add more interesting decisions for the player to make until the best decision is no longer clear, or there is no singular "best" decision. But at that point your RPG combat is no longer "simple".
If you have examples of good, simple RPG combat then stop listening to me and spend your time analyzing what they're doing that you aren't. Copy it wholesale then play around with it to see at what point it breaks down and what elements are required for it to work.
Make it less simple and move it closer to Mario RPGs?
>If you have a limited number of options then the game becomes a solved problem
Some Flash games had that problem, tried to copy a well-known game but with less options, became a boring puzzle.
>entire world blacklists Russia
>AGDG dev adds Russian localization to his game
Yes, I know it's pretty simple right now, I just wanted to have something playable before I start complicating things... but it looks like it doesn't work that way.
I've been fucking around with Unity for a couple of weeks, and this is the result. I'm a fucking noob, but I'll learn more about it as I keep working on it.
Okay. Now what?
Almost all the progress posters are from 4chan lol.
Laugh at them
>have an idea
>spend a week making assets
>do something unrelated irl and skip a day of work after completion
>lose motivation to bring myself to actually program the thing
how do I do it
just copy dragon quest
Radar rotates with the player's direction, animations are all back in, and wallrunning is now determined by a while loop rotating a raycast 8 times (per physics frame). Going to use this to get normals/collision positions.
Post advice and/or alternatives, you wont convince them of anything by communicating with memes.
Etrian Odyssey is probably a good inspiration. There's RNG, but unlike something like SMT it isn't stupidly important.
>Give me an idea for an FPS game,
combine Galactic Conquest from Star Wars Battlefront 2 with your FPS game.
Ignore GPT-3 posts. How's your game progress going, anons?
There's no shame in making an easy game that people can actually play and beat. Most games are pretty easy.
What GPT-3 post? And as for my progress, I haven't done shit. I'm totally bummed out from Trenchbroom limited tools and that it constantly messes around with my selection all the time while trying to edit brushes is annoying me more.
At one point I nearly lost tons of brushes because I didn't noticed I had other brushes selected while I was trying to delete a brush so I had to press the undo several times over to get them back. Meh I don't know, it doesn't motivate me anymore to continue mapping seeing for weeks how my map is flat as a Wolfenstein 3D map.
I made some player models and some guns, making a simple FPS game and hope to make it multiplayer
I've seen too many games, where I though the game was too easy, then I went to ((( steam ))), and saw people whining about how impossibly hard that game is... So maybe what I think is piss easy is not that easy actually. But on the other hand I don't give a flying fuck about what normalfags would think about my game, if you don't want even a little about of challenge go watch a movie or read a VN or whatever.
Don't give up anon, if I can do it you can do it! Remember your subtraction CSG operation and to vertex edit, also layer your brushes and lock layers you don't use.
Slowly, but I'm keeping to the schedule. Recently finished a refactoring pass and now I'm incorporating some feedback we got from play-testing.
GPT-3 is an AI that generates human-like text. You give it a prompt, it writes something in response. If someone has an axe to grind or they're bored and want to farm (you)'s then they could use this or a similar AI to waste everyone's time by creating realistic but ultimately empty posts.
It's a more specific form of "don't feed the trolls". While normal trolls will specifically try to get a rise out of people by making deliberately false, contradictory, or inflammatory posts, GPT-3 posting does none of that. It just wastes time.
Are there any resources on active ragdolls for godot? I want to create a fightan game like chivalry but without all the spin to win bullshit
>What GPT-3 post?
He lost his right to accuse anyone he disagrees with of being cuckchanners so now he's going to call them AI spambots instead.
went back to work on my "game" after two months without any progress.
i got mad at cpp's stl containers and wrote some custom containers based on unreal engine 3's, but found myself lost deeper into the language's complexities, thus deciding to port everything to C.
some other "big" changes are:
- implemented custom containers for dynamic arrays, (robin hood) hashmaps, dynamic strings, and double linked lists
- added material types, where its definitions are read from a json file, and they may have an albedo, normal and AO maps, with metallic and roughness parameters for PBR
- added a texture type (stores the image handles in the device + number of references)
- materials and textures are now stored in the hash maps - using their names as keys, and are referenced by their hash instead of pointers, allowing the maps to be resized/rebalanced
- switched from tinygltf to cgltf for parsing gltf models
- merged world stage into model stage
- removed source bsp lightmap support, as they weren't working well with the PBR pipeline, may revisit this later on and implement own PBR correct lightmap compiler
- dropped bsp culling as it was broken, might look into culling with compute shaders
- dropped support for reading textures from plain images, leaving KTX as the only supported format
for reference, there are 3 render passes in place: deferred, "after deferred", and forward
and these are the "stages" executed in them (dunno what else to call them):
- model stage (deferred) - writes position, albedo, normal, material parameters (metallic and roughness) and depth into each respective images
- ssao stage (after deferred) - name says all, still no fucking blur as i think it would require a new pass
- composing stage (forward) - builds the final image from the results from the model and ssao stages
- skybox stage (forward) - re-uses the depth from the deferred pass and draws the sky in empty pixels
- ui stage (forward) - draws text on top of the final image
moving on i'll be looking into shadows, and networking afterwards. might also fix the leaking objects that vulkan's validation layer keeps complaining about.
by the way, is there a more "standard" map file format? i'm using source's bsp format since i know how to do some stuff with hammer, but i'm not using some of the format's features, and gltf doesn't seem to support or even suitable for bsp. otherwise i might bite the bullet and try to make one based on source's.
getting started is one of the harder steps, even more so when you hit obstacles like >>140516 has. but you must keep trying, as the motivation will only come from your work
Oh well I know what GPT-3 is its some closed text generator thing its just I don't know which of those posts was generated by GPT-3 so hence why I got confused by his mention of that.
Besides I am tired of the constant turf wars between various Image board and other boards, as long the anons that come along don't act like absolute niggers I don't mind them, like if they act in good faith and don't post some abhorrent stupid posts and such.
>He lost his right to accuse anyone he disagrees with of being cuckchanners so now he's going to call them AI spambots instead.
But he only made 1 post? I don't understand.
I use CSG operations sometimes to make a bit more complex shapes, the fountain I did was made out of CSG. For layers I think I am using 5 or 6 layers already to "sort" the brushes into layers for a bit easier management. Hmm I guess maybe I should group the cliffs too I should try that and see if its a bit better editing that way so that it doesn't select other brushes that I do not want to edit and for performance reason.
I like the tools of Doom Builder more as its easier to create cliffs in that program, I can just left click on the lines and it automatically creates a vertex which I can edit it immediately without having to press first "vertices tool" which is the case for Trenchbroom.
Meh I tried looking yesterday again for another tool that I can use instead and well I have found this: https://christt105.github.io/Elit3D/ I didn't tested it out yet so I am going to do it later when I feel like it, hmm it seems it can create from tiles a 3D map out of that, which might be useful in my case so that my maps looks a bit more interesting to look at since I want to have heavy focus on exploration in order to obtain items.
Which is one part I liked Doom so much but I disliked on that game that it takes sometimes what it feels hours long just to get a specific item or secret as items are sometimes placed in places that are hard to find, the secret sections on various Doom wads can get quite obtuse too, some of those secrets are literally hidden behind walls that look exactly like the other walls so you won't be able to find these unless you happen to bump into that specific wall.
>but you must keep trying, as the motivation will only come from your work
It doesn't help that I feel low energy every single day, so I need to basically force myself to get anything done which is hard as I don't know how to describe it, my brain is like constantly put inside a narrow blinder or something. The psychologist I go to every 8 week is not much of a help either.
>i got mad at cpp's stl containers and wrote some custom containers based on unreal engine 3's, but found myself lost deeper into the language's complexities, thus deciding to port everything to C.
Sounds so familiar, except I still haven't gave up of using custom containers. I'm curious what you have right now, how about sharing your code for me to shamelessly steal while maybe still crediting you? Valve and EA and Carmack all wrote their own containers instead of using stl shit, EASTL is even available for use for general public, but source code is so complicated looking that I'm not sure if that's a good idea. Anyone here looked at teeworlds source code? They seem to use their own custom containers also, arrays and ranges.
Finally my module system is working properly, all that I need is to type the module ID, and the modules is loaded directly into the character node, the player controller detects the module and search for a controller for that module, its great, work like a well oiled machine.
Now, I just need to make the game, yeah, the game, maybe I should improve the camera system, making sure it can works on 2D environments as well, or maybe I should work on system to allows me to create a viewport to another scene, so I will be able to create that gimmick from Zelda a Link between worlds, or maybe I should test the sound system, or improve the AI scripting system.
Good Lord, I don't feel comfortable making the assets, every project that I worked ends up here, in this exact point, I have a good system working out of the box, ready for the assets, just drag and drop, that fucking simply, but I give up because I lack the ability to create good assets, its been what, 3 projects by now, 5, or 6 years on the same fucking cycle, I lost count.
I have no idea how to escape this, its like I'm the protagonist of E.Y.E., living the same nightmare for God knows how long, because I refuse to give up or move on, I'm stuck.
>Oh well I know what GPT-3 is its some closed text generator thing its just I don't know which of those posts was generated by GPT-3 so hence why I got confused by his mention of that.
It was the "do my homework for me bros" faggot, all his posts are so similar I assumed they were generated by an algorithm. It's deleted now, either because someone finally did their fucking "job" or because the poster deleted it himself for more D&C ammo.
>how about sharing your code for me to shamelessly steal
go ahead: https://gitgud.io/glue_sniffer_420/game_test
you can find the containers under "src/u"
you may also go back in history to find the old CPP containers - though i suggest agaisn't using the hashmap, as its over-engineered.
the C map is based on (read: stolen from) https://github.com/tidwall/hashmap.c , everything else's based on their UE3 counterparts.
unfortunally the C containers can be more prone to errors, when compared to the older CPP implementations
>Valve and EA and Carmack all wrote their own containers instead of using stl shit, EASTL is even available for use for general public, but source code is so complicated looking
that's one reason why i dropped CPP all-together:
for example, take a look at this hashmap implementation: https://github.com/martinus/robin-hood-hashing/blob/master/src/include/robin_hood.h
sure it may have plenty of features, but they come with the cost of complexity, when you probably just want some of them
idk why godot makes me set collision masks in binary but it works i guess
No visible progress right now, but I'm working on navigation and combat abilities.
Regarding abilities, I'm down to 4 abilities and 4 modifiers that can be applied to bound abilities. A small amount of interesting and useful combinations is better than a larger amount of more niche combinations. Going to try to program these in and see what I get.
>just being able to platform isn't enough for what I'm trying to present the game as
It could be enough for a demo, and you'll get feedback on movement so you can refine it if needed before building other mechanics on it... Either way, I'm looking forward to the demo!
Does the GDD need to be very elaborate? You can focus on the general ideas and mechanics in the GDD, then work out the details while actually programming the game.
Do what makes the most sense for you and your game, there's no single way to go about it.
What program is that?
How nice of him!
Generally speaking any developer will always think his game is too easy, he literally designed the game and knows every small detail about it!.. Make the game, even if you (or others) think it's easy, and publish it. Don't get caught in analysis paralysis.
Cool font but not monospace, why do this to yourself? Here:
I use Source Code Pro, but hypothetically where would I go to get a pro-imperial, autocratic, anti-democracy monospace font?
Many have tried before you to change my font and none have succeeded.
>What program is that?
Unironically jewgle: fonts.google.com
There's also a list of free/libre font websites here: https://opensourcedesign.net/resources/#fonts
Well at least I tried ¯\_(ツ)_/¯ I guess it doesn't really matter as long as you're comfortable
Unpack/rip apart another game to vampirise it's assets and keep coding until you manage to convince some shmuck to make assets for you.
> I give up because I lack the ability to create good assets
So you're the same as every other computer programmer?
I have some basic combat system, a few types of attacks and skills and items. Some attacks require some skills, in other case having a skill just increases the likelihood of a successful attack, equipping certain items can also have some active or passive effects... My question is, how to do this without your code ending up as a fucking spaghetti with random if (has_xy_skill) and if (has_z_equipped) checks all over the place?
Hard to say without having an in-depth look at your system, but just as some general advice you want to go with the simplest option available.
It sounds like you have a list of actions (attacks) that are modified by a list of things that alter those actions (modifiers). I would keep the list of modifiers in a central location and give each one a tag. Then I would maintain a separate list of actions. Whenever either list changes, I would reset the actions to their default state and then iterate through both lists, modifying the actions when they have a tag that matches a modifier.
Dynamic navmesh plugin:
Navmesh can be rebaked during runtime, but this plugin doesn't play well with Qodot generated collision. What I had to do was make each generated mesh generate a new Trimesh Static Body per mesh, set those collision groups to something other than what the generated floor collisions were, then it would build a navmesh properly. I'll look into implementing this with Qodot maybe in the future, next is getting NPCs to navigate on this with behavior tree nodes.
The behavior tree nodes will be what characters pathfind to in 3D. 2D pathfinding on radar is likely to stay in but at a reduced capacity, unsure how yet.
Organize what you need into data. Have it so that if data is added, it's not necessary to change the code used to access it. Load that data from a file. Make it so that you can reload the data without restarting the game.
Maybe use a Array of strings and do a recursive loop to check if a attack needs a certain skill or uses skills for modifiers? That's what I would do.
please tell me there's a better way to break this up
#scope creep, break this up
if not is_on_floor():
$LedgegrabHook.monitoring = true
$WalljumpCheck.monitoring = true
if playerVariables.faceDirection == "downleft":
knockbackSpeedX = 750
knockbackSpeedY = 1500
elif playerVariables.faceDirection == "downright":
knockbackSpeedX = -750
knockbackSpeedY = 1500
elif playerVariables.faceDirection == "left":
knockbackSpeedX = 750
knockbackSpeedY = 1500
elif playerVariables.faceDirection == "right":
knockbackSpeedX = -750
knockbackSpeedY = 1500
elif playerVariables.faceDirection == "up":
if playerVariables.lastFaceDirection == "left":
knockbackSpeedX = 750
knockbackSpeedY = 1500
elif playerVariables.lastFaceDirection == "right":
knockbackSpeedX = -750
knockbackSpeedY = 1500
elif playerVariables.faceDirection == "upleft":
knockbackSpeedX = 750
knockbackSpeedY = 1500
elif playerVariables.faceDirection == "upright":
knockbackSpeedX = -750
knockbackSpeedY = 1500
$LedgegrabHook.monitoring = false
$WalljumpCheck.monitoring = false
if playerVariables.faceDirection == "downleft":
knockbackSpeedX = 750
knockbackSpeedY = -1500
elif playerVariables.faceDirection == "downright":
knockbackSpeedX = -750
knockbackSpeedY = -1500
elif playerVariables.faceDirection == "left":
knockbackSpeedX = 750
knockbackSpeedY = -1500
elif playerVariables.faceDirection == "right":
knockbackSpeedX = -750
knockbackSpeedY = -1500
elif playerVariables.faceDirection == "up":
if playerVariables.lastFaceDirection == "left":
knockbackSpeedX = 750
knockbackSpeedY = -1500
elif playerVariables.lastFaceDirection == "right":
knockbackSpeedX = -750
knockbackSpeedY = -1500
elif playerVariables.faceDirection == "upleft":
knockbackSpeedX = 750
knockbackSpeedY = -1500
elif playerVariables.faceDirection == "upright":
knockbackSpeedX = 750
knockbackSpeedY = -1500
You could try using a mask or a similar approach, so that each combination of modifiers is a unique sequence of 0s and 1s, and then check to see what numerical value it is.
Now that I'm thinking about it, this may not be fundamentally different from your initial plan, though it may be shorter and simpler.
Thanks for the ideas, they should work if my modifiers only modify some parameters, but if I want to have other possibilities, I'm just refactoring my spaghetti into a different shape.
For example, let's suppose I have a skill that gives the player a small chance of reflecting the enemy's attack (so it'll hurt the enemy, not you). Ideally this should be somewhere in the skill's code, and not in the attack code. I see no fundamental difference between checking if (has_xy_skill) in the attack code versus if (has_xy_param) where has_xy_param is set to true IIF you have a specific skill.
That's what you wat to do in the final release, but if you have 1000 skills you don't want to have to look up which bit you have to set to a 1 to make something a prerequisite. Have strings for developer input, convert them to numerical IDs after loading, and further optimize those into bitflags if the code needs optimisation.
Like >>141721 said, change things to data, and as I said above mae it as simple as possible on yourself while developing, then compile it to faster stuff when releasing.
If you want to make it as easy as possible by not dealing with file shit yourself, use JSON or something similar. Each skill has a field for just prerequisite names, and you only write code that works agnostically for any data.
Your combat system uses algorithms on player and enemy data to spit out results and change parameters of both. Let's for simplicity say both player characters and enemies are of a Combatant class. HP is a nigh-required sort of parameter. But Dodge Chance, or for instance the Reflect Chance you mentioned are parameters too. You probablyalso want to have multiple parameters that together decide an overall one, such as multiple types of defense being used together to decide an ultimate damage reduction per attack instead of just a single one.
Meanwhile all your skills need is a name field, a prerequisites field that just lists names, and a parameter field that list parameters in the Combatant class, and some value. It'd be ideal if you can get it to interpret simple math here, so you can do multipliers and additions without needing separate parameters for them, but technically you could do without that.
One big benefit of this is that it directly extends into stuff like equipment. That too just modifies Combatant parameters, and you can even have equipment that provides skills and/or have equipment be prerequisites for skills (e.g. needing a shield to use shield bash).
A lot of this is based on knowledge of how Etrian Odyssey does things.
You need to create a complete design document for this combat system if you want to reach a solution, either by yourself or with help. I can give an additional solution for what you describe, but I'd like to avoid
>here's a solution
>I forgot this other part your solution doesn't cover
repeated ad infinitum. You need to describe, in totality, the problem you're trying to solve.
With that said, damage reflection can also be done just by changing parameters. You have a list of functions (or function pointers, rather) attached to an enemy that run before the entity takes damage. Skills/actions and so on can add functions to this list. All of these functions take as parameters a reference to the attached entity, the targeting entity, and the damage (including value, type, etc.). You add the "reflect" function to that list, which performs all the relevant checks, can modify the damage taken, and can call the "take damage" function on the enemy that targeted you.
Any good resources for deving for older systems? I want to make PSX/win95 games, its not enough that they have the low poly style they should also be able to be played on the lowest possible hardware! What language do i need to learn and what engine can I use?
>You need to create a complete design document for this combat system
The problem is that I wanted to make it moddable in the long run (not like I expect anyone will make a single mod), so ideally I'd need something that's fairly extensible.
I guess I'll try having a bunch of function pointers, at least that's pretty general, and hopefully not too complicated.
C (maybe C++, but forget any modern features) and some compiler that can still target that platform. For windows you have openwatcom, the open sourced version of the old watcom compiler.
FreeDOS YT channel has a tutorial about C programming in FreeDOS https://www.youtube.com/watch?v=MpKoh-AabTM&list=PLzuACU-W7Omo3VEnMKuM0IPupdOHFDzL3&index=1 + >>>/tech/5755 contains some notes about FreeDOS.
If you are interested, for GB/C there are The Pandocs (https://gbdev.io/pandocs/), GBDev wiki (https://gbdev.gg8.se/wiki/articles/Main_Page) and RGBDS (https://github.com/gbdev/rgbds) & GBDK-2020 (https://github.com/gbdk-2020/gbdk-2020).
Should be (relatively) easy, but you'll most likely have to make separate builds for win9x and winXP/7. Also you will have little to no 3D capabilities and may have to fall back to 3D software emulation.
- Language: C (C89 or C99, check your compiler) or C (C98)
- Compiler: DJGPP (for MS-DOS but works on Windows), OpenWatcom, maybe older versions of MinGW or MinGW-w64 (32-bit)
- Build System: POSIX make (compatible with GNU make on linux, bmake on BSD, and nmake on win9x)
- Video/Input library: SDL 1.2, Allegro 4.x, maybe SDL2 & Allegro 5 if you compile them yourself although neither officially support pre-winXP systems
- Engine: make your own, alternatively you can try to track down the original Doom or Blitz3D source code before it got modernized
Forget it, that's orders of magnitude more difficult than programming for Windows. At least finish one game for Windows first before tackling PSX programming.
>The problem is that I wanted to make it moddable
Let's say that any attack ever has a chance of:
1. Inflicting damage on an enemy
2. Inflicting damage on the user
So any time the player picks up a powerup you calculate the 2 damage rates above and store them, then whenever the user attacks you just use both of these pre-calculated values (either can be zero). Same approach for defense or whatever... At least that's how I would do it.
You should avoid unnecessarily calculating values especially for a frequent action like attacking or defending.
>or C (C98)
That's C + + (C + + 98)
who the fuck thought this formatting was a good idea?
>have many ideas for different games
How do i Pick one? im ideafagging out the ass but I want to make something halfway decent for people to play
>print out your ideas
>put them on a dart board
>throw a dart
Now write a game design document that even a shareholder could read and understand.
The way you should go about learning C is to watch the tutorial series and read a book at the same time. Read either C Primer Plus or C Programming: A Modern Approach.
Dr. Dobb's Journal may also contain some interesting stuff: https://archive.org/details/DrDobbsDVD5
Also, read the Graphics Programming Black Book by Michael Abrash: https://www.drdobbs.com/parallel/graphics-programming-black-book/184404919
The reason why I recommend FreeDOS is that:
- FreeDOS is open-sores
- It has a package manager (a package is just a .zip file). You can distribute packages with CDs (.iso files) or over Internet.
- FreeDOS is the nicest version of DOS. And it comes with software development tools.
If you haven't made a game before, pick the easiest idea to implement (preferably a 2D game) and just start making your game. I think the best way to approach game making (or programming in general) is to make small "tech demos" while you are making your full game/program. The easiest way to make a game is to use Lua+Love 2D or Godot.
Gamejam a bunch of the ideas, see which ones actually work. If you can't gamejam it you also can't make it.
Speaking of which, why dont we have any gamejams here? we have demo days and other shit but ive never seen a game jam, would be nice tbh
Do something about it or kill yourself.
>The problem is that I wanted to make it moddable in the long run so ideally I'd need something that's fairly extensible.
Now I understand.
Your problem isn't the combat. Your problem is that you've tied a millstone to your neck by trying to optimize for modding support when you don't have enough gamedev experience to effectively do so. I realize you care about modding and I respect that, but right now the best thing you can do for your game is to forget about modding and do the minimum, simplest thing required to give your game the best combat. Once you have a build with a combat prototype that does everything it needs to do, then you can start thinking about how modding may fit into this. And once you have a game, you can worry about modding tools.
The best modding support you can provide is releasing the game and listening to what the modders want to be able to do, then providing patches that open up those possibilities. But you have to release a game first.
Be the change you want to see in the world.
What's the best way to organize a stat/skill system to be as efficient as possible? Currently, I've just got the stats and skills listed as variables within a class, and I'm curious if it would be more efficient to divide the skills and stats into structs or unions, or would it work well just as it is?
>for a site that's largely solodevs
Is this the new language wars?
wasn't the original doom source code for windows actually lost? only the linux port of doom actually exists
I made one and I fail to see the point of a fucking game design document, I already have vague idea what my fucking game is going to be so and I just use my current GDD as a list of check marks of shit to do.
No idea, but it doesn't matter. Any source code from the 20th century is a good starting point for a true retro game.
IIRC. they just had to use the Linux version because it didn't contain the patented sound engine.
Efficiency doesn't mean anything without some kind of measurement. How are you planning to use these stats/skills, and how often do you plan to access them? If you're just going to call skillClass.getSTR() for most use cases, then I don't see anything wrong with just having a big list of data.
>I don't understand what planning is
Working by "instinct" will usually end in failure and burnout.
>How are you planning to use these stats/skills
I should elaborate some on my planned system. I've got the stats arranged into three stat types, based on Body, Mind, and Soul. So combat and physical stats and skills would be affected by Body, intellectual stats and skills would be affected by Mind, and magical stats and skills are affected by Soul. There would be certain limits based on those stat types, so if you had a low Body stat, for example, you'd be limited in physical stats and thus couldn't become very strong in combat.
>how often do you plan to access them?
Unless something comes up where I have to use them some other way, I think the main thing I'll do is just getting values when needed and setting new values when leveling up.
Then your approach sounds fine. I wouldn't worry about optimizing for performance if it just works and takes up a trivial amount of RAM.
Interdasting infographs, thanks.
It was used in a hack to get around technical limitations. For all intents and purposes, this isn't AA.
>Is it white hair or is it a hat?
lol people actually thought that was hai-
>his hair is painted blue and-
some more examples
Is someone going to use these guides or is this just "I love space!"-tier involvement?
>use a dict/enum for all 8 directions, then make a class that holds all related vector2ds & knockback (put that in a vec2) or use a poolvec2array or whatever it's called in godot if you know for sure that's all you need
>use variables that dictate state rather than variables related to state, i.e player_variables.state == states['jump'] or whatever instead of player_variables.playerjumping
minimum example with the dict/enum/array, should look something like this
>var clwk = cannon_ledge_wall_knockback[face_direction]
>set_animatoin(clwk) or set_animation(clwk.sprite_pos)
obviously there is a special use case for when the character is facing up from the looks of your script
Late as hell, but I finally got around to trying this game out. Cool concept, but I'm strongly not a fan of the 8-directional sprites not accurately portraying where the launcher is actually being aimed.
It'd be cool if for clarity's sake, the rocket launcher would aim at the mouse similarly to the clip posted.
Aside from that, it's a pretty neat and cute concept going on here and I hope to see some cool puzzles and polish with this project
Decided to face it and create a 3D model, I used the character sheet as reference, its not a first time model, I've following tutorials for a while now.
I know low poly models live or die based on their texture work, but I'm unsure if this a good model or a bad model, it has around 2.5k triangles, I'm still a beginner I'm not sure if I can push to 1k polygons or even less.
I know modern hardware can easily render more than a 1 million triangles on a screen, but I don't want to ignore the fact people with toasters will play my game, and I want a style that allows me to finish 75% or more of a character on a bad day.
Looks good. I would suggest not modeling each finger and let textures take care of that.
See pic related.
Or model the thumb and index finger, and let the rest become one single part.
What a blast from the past, used to moderate the forums and did some beta testing at the earliest builds where Overweight spam was the most effective strategy.
Depends on what you're trying to achieve though, if you have scenes where you can see it close-up it can be weird, like the cutscenes in Vice City.
>got raycasting working, something that absolutely refused to work in stencyl but works just fine in godot
DEATH TO STENCYL
Show progress anon
Path finding in action, and DetourNavigation+baked Navmesh that uses collision siblings generated from floor meshes, set as children of StaticBody nodes.
I've spent the last week reversing some shitty old nip VN engine instead of working on my own project. FML.
if I just steal all of D&D mechanics and put it in my game can I get away with it
Which ones? As long as you change the names around you can almost certainly get away with it, since you can't copyright game mechanics yet. If you're stealing 5e mechanics, then you might as well build your own system from scratch since you're going to have to do that anyway. If you're stealing 4e mechanics, then you can probably make a fun party-based dungeon crawler. If you're stealing 3e mechanics, you will be working on a Dwarf Fortress-tier fantasy simulation for the next 20 years. If you steal 2e mechanics, you're remaking Baldur's Gate.
Sounds like a recipe for thirty years of success.
>you can't copyright game mechanics
You can patent them.
>You can patent them.
How does that work? I know Wizards did some copyright magic with 3e so that people could release third party supplements and make money off it, but I think that was still the branding of the mechanics and not the mechanics themselves. You can't copyright "roll 1d20 and compare the result to a target number to determine an outcome".
I don't know how it works, but video game companies have done it. Square patented the ATB system, and WB patented the Nemesis system used in their Mordor game.
I'm talking out of my ass here, but, based of what I've read in patent documents, you can't trademark basic or simple things like rolling a dice, but you can trademark set of equations and algorithms.
For example, Wizard of the Coast algorithm to play a card game.
As long you don't use the same algorithm, you are free to create your own ATB system or Nemesis system.
I would rather kill myself than fully read the Jews of the Coast legal documentation, but after skimming that for a few minutes it doesn't look like many mechanics of M:TG are actually copyrighted. Sure, they can sue you if you have a mechanic for "tapping your mana card" in your 40 card deck trading card game, but that's about it. They can't copyright trading card games.
NOTE: I am not a legal jew, or will I ever be. I am probably viewing that document through the lens of a human being with a soul rather than some kind of law-aligned automaton. Lawfags go die.
looked into it and it seems they have a list of shit you can't use
so as long as you don't take anything from the list you'll be fine
yeah they have the word "tap" copyrighted so games just use another word
they also have the word "d20" copyrighted, but not the design of the object itself
Anyone know the name of that Moonman game that was posted a while back? You could play as Moonman, Terry Davis, Ben Garrison, and I think Adolf Hitler, drive around in a car and run over niggers, etc. I think it was mostly or entirely driving gameplay.
Interesting. I wondered what all was copyrighted in D&D. It's surprising how few monsters are on that list.
>they also have the word "d20" copyrighted, but not the design of the object itself
Only when specifically referring to "the d20 system", which given how many games were shat out using it... people know what it means.
The vast majority of D&D monsters are actual fantasy creatures from myth and legend, even if a lot of them are obscure. Creatures like the Tanar'ri and the Baatezu were created not only so christian moms would get off their back and they could scrub all the 'demon and devil' references, but so that they could be copyrighted. I think it's really funny they have the Clockwork Nirvana of Mechanus on there, but not the modrons.
What's weird is that owlbears aren't on the list, considering they started with Gygax finding a Happy Meal tier chinkshit toy at a flea market.
I know the Gith were copyrighted due to a conflict with a George RR Martin book, and the illithid is a ripoff of Larry Niven's Thrint. Yuan-ti are a copyright-friendly Naga and the displacer beast is a copyright friendly Coeurl. Beholders and gauth are actually somewhat original.
"Owlbears" are probably too simplistic to copyright or something. I mean, you can't copyright a minotaur or a centuar. I dunno.
Black lives splatter, it was a shit game, >we could do better
black lives splatter
How do I properly update the Godot's animation track paths, I want to do via code, but, it will slow down the game, having to individually check hundreds if not thousands of paths, just to remove a few characters, it doesn't sound like a good idea, and I don't want to modify manually because it will break the preview, and its a waste of time.
Importing models and armatures to godot, it will always be a pain in the ass, it creates random trash, maybe in 4.x things will be better.
Could you do the update on a separate thread?
As far as if they've fixed 4.0 importing, you can install the current alpha for 4.0 and see for yourself. If it's not fixed there, then you'd have to search through the listed issues or ask one of the people involved in development.
Its not a good solution, but I solved the problem by creating a new scene via "Open Anyway", if you move the nodes in the editor viewport, it automatically update the animation tracks, but if I do the same thing via code, it doesn't work.
Is godot 4 worth using now or should you wait for more updates?
It's still in alpha.
anyone know if godot 4 will let you do hexagonal gridmaps?
I've been thinking on it and that sounds like a lot of fun
afaik it's not possible in 3 yet
>start making 3d game
<the gridmap doesn't have any collision for some reason
Maybe wait for Beta instead when things are bit more stable, you can use Godot 3 instead and make small projects in the meanwhile or something.
There should be a way to auto generate the collision for the gridmap, is there no new tool options being shown when you select its meshes? I don't know how grid map works as I haven't used it yet, worst case you could auto generate the collision via code by calling one of those function instead if there is tons of meshes so you can use either groups or a loop call to do it.
There needs to be an game engine library. Something that handles graphics and sounds and shit for you, except you use it as a library instead of having to use some GUI engine.
I'm too autistic to use an engine but too low int to make my own.
So basically Raylib?
Can you import animated models and blend between animations? Does it have shadow and light effects?
Or SDL / Allegro
Yes but only IQM and GLTF formats. I think the latter's compatible with FBX, just exported differently, but I'm not sure.
Yes, assuming you mean it has lighting for models. Not sure what you mean with light effects honestly.
I'm not saying Raylib is amazing. It just does the basics of everything. But that's all you asked, "handle graphics and sounds and shit". You didn't specify anything else. You didn't even specify 3D. Your post was stupid as fuck.A better option would probably to use something like BGFX for rendering, some audio library, some input library etc, and cobble shit together yourself. Stop complaining about how low your int is and put more points into it instead.
>only IQM and GLTF formats
As long as blender can export it it's fine.
>Not sure what you mean
Cast shadows because that's the part that adds a bunch of architectural complexity to your renderer and isn't just some shader calculation trick.
How hard is it exactly to make your own 3D engine anyways? is it super duper hard or does it just take a long time to develop? Remember when every company used to make their own engines instead of using pre-made ones? why did that stop in the past decade or am I just retarded?
Complexity ramped up while most games no longer need to squeeze every drop of performance out of the HW, so they no longer need specially developed engines. It's easier to get a pre-made engine than doubling the development team (and you need more capable people than your usual diversity hire if you want to end up with an engine that's not a dumpster fire) and adding a year to the development time.
It's not that hard, nor does it take that much time to develop.
The main obstacles are:
>If you don't have clear requirements, you'll end up having to re-implement things many times, inflating development costs.
>Making an engine also means creating editing tools, compounding to development costs.
>Your engine doesn't have years of being used worth of testing. Bugs are not just inevitable, but will occur on hardware configuration you do not have access to.
>The skillset isn't widely available, there's a lot of problems to solve that most developers, especially game developers, don't have the skillset for.
>Using an existing engine allows you to lift assets from the asset store, including code, significantly reducing development time of games.
>Most small/indie games don't actually require a specialized engine; existing engines meet all the requirements, so developing your own engine is essentially just a more expensive solution with little to no gain.
You want to make an engine for your hobby project? Great, go ahead.
You want to make an engine for a commercial project? What features do you need that existing engines (& its sold asset store assets) lack? And for clarification; "I want an engine that doesn't suck" isn't a feature, because whatever you make with your lack of experience will suck all the same, if not more.
The main obstacle is that making a custom engine simply does not make any financial business sense. It not only adds costs, but also adds negative value compared to using an existing engine that people might already have experience with.
If you wanted to turn a 2D game into a 3D game it would be almost as easy as changing vec2s into vec3s. It's not a big leap at all.
The problem is that 3D games have a lot more expectations for them. 3D games use proper physics (as in shape collisions that can rotate) more commonly than 2D games. 3D characters require a skeletal animation system. You can get by in a 2D game without raycasting but in 3D it's more common. 2D games usually don't use cast shadows at all but for 3D it's expected. Sorting transparent objects so that they render correctly is much harder in 3D. Making the rendering look "competitive" is especially hard, just supporting several light sources at the same time can be complicated. It's not so bad if you use stylized or "retro" (e.g. Dusk) graphics.
Whats the deal with arcade machines? Has anyone here developed an arcade game with the extra peripherals and everything? is it possible to do in to days market or are arcade machines a thing of the past? How hard would it be to make something like a lightgun game with proper working lightguns on some shitty LCD screen?
I don't really know anything about that market, but I can't imagine it being viable to enter, especially now with corona forcing the arcades to close for several years.
Maybe you'll find some success with hardcore enthusiasts. I think the question is less about being able to step foot into this industry, and more about there still being an industry left.
Devolver actually announced a Gungeon lightgun arcade for 2020, but it has been perpetually delayed since. Worse, I think that machine was only announced as a marketing gimmick (which shouldn't be a surprise considering all the crap they show and announce on their shows), and I don't think they were ever considering it as a realistic business model.
>replying to /cattle/
Since it's at least a mildly interesting question I'll bite the bait anyway.
The only new arcade games being mass-produced that are actually games are Golden Tee and Big Buck Hunter, both produced by Raw Thrills. In the past they've done a couple throwbacks to better days (like H2Overdrive which is a clone of Hydro Thunder on newer hardware) but the rest of the industry is a vast wasteland of carnie tier shit like you see at Dave & Buster's type places, sometimes fun but barely above kiddie ride level. There have been a couple indie dingleberries who produced limited production cabs for the barcade market, but I suspect that with the jews slowing down the free money printers you're running out of time to get funding for something like that. Weirdly pinball is doing great with multiple new tables per year.
it's pretty frustrating
if I create the mesh library for the gridmap using spatial nodes, then the gridmap won't have collision, but if I create it with meshinstance nodes, the pieces of the gridmap won't align properly, defeating the point of the gridmap
if the gridmap node actually worked it would be amazing but I can't get it to
You have to manually create a collision mesh for the scene, that is how I use in my game.
Tl;DW canuck dev reverse engineers his own DOS game which he made when he was 17, but stops halfway through after receiving the full source code from his co-programmer... Then proceeds to enhance the game a little on DOS, then fully remake it in Godot.
Subtitles included in video.
What game is that?
you don't want to know
>didnt make his game open source
woosh! what a thrill, why do faggots like this still exist? if he made it open sores then he wouldnt need to reverse engineer it
Source included, in case someone wants to play with what's shown in the video.
I'd tried about 5 different ways of setting up the mesh library and each seem wrong
seems that importing a glb file and then using the automatic trimesh collision thing works the best, in case anyone else has issues
what muusic editor are you using?
rather, which fonts though?
Default brushes imported using Qodot generate as StaticBody nodes with CollisionShape children, which I instead use as children in Area nodes. I am using scripted Areas to detect characters for determining under which index to parse PathNodes (1 if room 1, 2 if room 2, etc.) for simpler navigation while moving to a destination of any distance. PathNodes will be working with bitmasks, to modify executed behaviors similar to Halo 2 handled behavior tree and stimulus reaction. I want a character to be able to dodge out of the way of another character's projectile, to strafe, and to handle jumping, all while pathfinding.
Making a purely 2D game, am I remembering correctly that I should use godot 2 rather than 3/4?
stick with godot 3
there's no real reason to stick with godot 2
Doesn't matter, nobody makes games in Godot and you won't be the exception.
They heart looks really good in and of itself, but it's a bit distracting.
She has a big heart.
>trying to draw a Line2D from the player to the first collision point at a raycast which also comes from the player
>both have the same origin point
>the raycast is clearly hitting the right coordinates, but the line being drawn from it is slightly off, even after using to_local()
what could be causing this?
Portal rendering, before and after. I can't believe Godot engine developers didn't want to add culling features other than frustum culling.
I've had good results in Godot 3.4.
>nobody makes games in Godot
Not with that attitude. Go dev, anon.
figured it out, default positioning in godot affects a lot of things. never move your shit in instantiated scenes off-center in the editor if you can
It's looking good. Have you checked the profiler and measured how much exactly the performance has improved on your machine?
ideafags, give me ideas for a enemies in a 2D metroidvania that aren't just rehashes of what has already been done
A bug that weakens you and can only be removed at a checkpoint.
A humanoid creature that steals your weapons or abilities and uses them against you.
That's a good one.
i can see that being really fucking annoying, imagine putting a bunch of them in a hallway before a boss and you either go in gimped or keep going back to save/cleanse yourself. i can hear the speedrunner rage from here.
A speedrunner would just kill the bugs before they attach to you.
but if they're floaty and move around a lot they'd have to take their time to stop and kill them all
Not yet, I'll post specs and profiler screens later today. I am not sure if it gave much of a performance increase yet, I was up until 4am trying to fix a transparency issue with a sprite above a mesh with a transparent mesh. Tangent, but I'll be asking around to maybe get that fixed as I'm thinking the issue is with the render modes and setting albedo alpha erroneously.
pic unrelated, my next objective after fixing that issue is to get additional pawn functionality working, including health and ammo.
My idea for health is to do hybrid between Left 4 Dead's temporary health and Mega Man Legends' shield, where damage sustained will turn a green health block red until enough time without sustaining additional damage passes, then it will restore to green. If more damage is sustained but not enough to meet a certain threshold, green or red converts to white health blocks. From right to left, any more cently changed white blocks must restore to green before the next block changes from red to green. Taking enough damage will convert a white or red block to dark green. If dark green, healing is required.
I'm considering doing a sort of hype mechanic with ammo, landing shots will restore ammo as each shot costs however much amount. Otherwise it recovers over time. I will allow ammo to overflow, so you can go from 8 ammo to 16 if you land enough attacks. To give a positive return as oppossed to regenerating only the amount previously spent, I will likely return double ammo quantity on a kill and maybe 1.5x (rounded down) on a stun/shield break.
pic is old concept planning
with a transparency shader*
>responding to sagefag
>transparency issue with a sprite above a mesh with a transparent mesh
- sort your triangles for draw order
- disable Z-buffer writes for transparent polygons this will make transparent triangles combine in strange ways
- if it's GUI then disable Z-buffer comparisons
- port it to Dreamcast
If your engine doesn't let you do any of this, then good luck.
Shader code here, if you want to take a crack at it. I'll be looking into your advice, I'm just not at dev station atm. Engine is Godot 3.4.4.
I typically sage ancillary dev planning and comments.
Would enginefagging for peak performance be a good idea? I want my games to run on complete toasters and be able to customize anything about the engine for maximum efficiency. Where would I start for making my own engine? Do I learn C then go from there or should I learn another language?
Enginedev can be a good idea in certain situations, but you specifically shouldn't worry about it.
To be more specific that the anon above:
There's two scenarios where you should enginedev: you're making something that isn't viable in any other engine, or you're an advanced level programmer who is sick of certain things in other engines.
If you have to ask "do I learn C" there are zero reasons to enginedev.
To pile on, I would say that enginedev can be a fun challenge and you would learn a lot of useful stuff, but you should 100% not enginedev if your primary goal is to make a game you have in mind right now, because chances are it will never get made. It will probably take you years to become a good enough C programmer to come up with something useful. Just use one of the engines available for your game and if you must enginedev do it on whatever time you can spare.
The solution was to change the Render Order of the Sprite3D node's Material.
Shader in action:
Enginedev is only for people who like enginedev. Learn any language, try to just like make game with it, and then decide based on whether you like it or not.
reate my cave i know that a lot of the sprites are disjointed, still trying to find out how the fuck sprite artists make it all line up
Is this a platformer or some top-down shit?
Also fill the wall with some patterns too.
platformer, still trying to get the hang of making tilesheets.
Is nice cave anon. Don't know why you choose to do pixel art. Pixels are very 2010. You want to make big money? Uncompressed 4k textures and audio. Aim for a gigabyte a level.
If you are looking for tilesets to study I reccomend Owlboy. Possibly the nicest looking pixelshit released in the last 5 years. Don't feel beholden to the standard it set. Owlboy took 9 years of development.
I fixed my skies, again. Using Room nodes only allows for the current camera to be used, so my camera->panoramic sky projection work around stopped working because the second camera that was placed between a bunch of meshes wasn't being used and would give a black screen. Now my sky meshes all have a shader material that moves them relative to the camera position.
I'm going back to pathfinding stuff now. Demo soon.
Sweet, looking forward to that.
I've found some of the biggest help so far is sleeping. Sometimes my brain gets stuck thinking I need to do things a certain way, and stops questioning certain aspects of the code/game. Often when I start working again on something the next day I'll see something I've missed or realize there's a better way to do what I'm trying to accomplish.
Pretty gud, it reminds me of the Brinstar caves in Metroid Zero Mission.
I don't enjoy sleeping, every time I wake up I feel fucking dizzy and easily pissed off.
that's what i was going for, was using the Brinstar tileset as a reference
You wouldn't happen to know how to mouse lock in Clickteam Fusion 2.5, would you?
ANUBITEK v0.3 demo is up. Demo is buggy, but I did what I could.
when is next demo day? august 8?
I've been looking forward to this! Sadly it crashes on linux/wine :( The EXE launches but shows errors in the console on startup, then when starting a game it crashes...
Any chance you could make a linux build?
Try proton-GE? Gets into game with 7-20.
A lot of the errors from the first screenshot shouldn't matter too much, but I'll see about a Linux build. I might need to have someone else build it.
Just tried that, same problem.
Thanks for considering. I can build it for you if you don't have a linux machine btw.
How do you make a good or interesting damage calculation, I understand this is a genre specific question, but I'm sure there must be some type of similarities between all of them.
min + random() * max
I assume you're talking about a JRPG or a game with similar menu-based combat. Damage in such a situation is mostly a timer, which tells your player "you have X number of turns to kill this enemy" and if he doesn't, then he loses a party member or his party wipes. Most rank and file battles aren't that challenging in JRPGs, but bosses often have more specific attack patterns that demand better timing or planning. Consider a boss that deals 1/3rd of each party member's health to your entire party each turn, and a boss that charges up for 2 turns, then kills your party on the third. Both of these enemies give your player 3 turns until death, but the time spent fighting the boss feels different in each case.
Attacks that scale based on player strength, like always doing 30% of a character's health, are generally unliked by most players because there's no way for them to overcome the damage and because it means the enemies only exist in the context of the player. I think a good way to do damage calculating for run of the mill enemies would be to ballpark what you imagine the player's health to be, decide roughly how much damage each enemy should deal with a strong or weak attack, and use a bit of minor randomness to make it seem more interesting.
Can't you just use Godot's Linux export template?
finally fixed walljumping after starting from scratch and taking it one step at a time. apparently
if playerVariables.faceDirection == "left" or "upleft" or "downleft"
is not the valid way of doing this and breaks it. instead you have to add playerVariables.faceDirection before each string. considering all the other niceities of the language i'm surprised this isn't one of them. or maybe i am just retarded. part of me is afraid to ever bring other people onto this project because then they'll see the code and call me a retarded niggerfaggot for all my conventions.
is bandwagonning with cuckchan a good idea? there are a few games here that are explicitly "made by zzz" and that would just attract the unwashed masses here.
There's probably a playerVariables.faceDirection in ["left","upleft","downleft"] or .contains(facedirection) method that you could use that might be tidier than typing out all the conditions.
I'm having a design crisis. I've been planning to make a game similar to Terraria for a long time, the idea is that it's focused on building and automation rather than combat. I think combat is necessary to some extent because exploration needs to be interesting, but when I see something like vid related, I can't help but to feel like there's no point to adding combat at all if I can't go all out with it. I've had all kinds of ideas about combat mechanics that I think would be neat like multi-stage weapon projectiles, but the more engaging the combat is, the more it takes away focus from the building aspect. Terraria itself suffers from it, almost nobody bothers to build anything because the combat is what they want to play for.
I always though of my idea as a "competitor" to Terraria in a way, but now I feel like that associating it with Terraria at all is an obstacle. But what does it mean for exploration if there's little to no combat? What should the monsters and environment be like? Should there be some kind of tactical or building-based approach to exploration rather than a combat-based aiming/dodging challenge with monsters?
I'm concerned about how often I've had major changes in the way I think about the whole game idea like this. If I start working on the game too soon, will it be worse off because I hadn't come up with some important part of the game design yet? Will I come up with another idea a year from now that completely changes the way I see things?
Think of what your core pillars are, write them down, the focus on that. If combat is not a core pillar then you're going to have to get over this idea
"I can't help but to feel like there's no point to adding combat at all if I can't go all out with it"
and force yourself to keep the combat at just the level you need for the game, and no more.
>Will I come up with another idea a year from now that completely changes the way I see things?
Yes. And the year after that, and the year after that. It never ends.
>keep the combat at just the level you need for the game
I don't even know what that level is. If I'm not doing "proper" combat, should I even have combat in the first place? It sounds like I would be adding it there just because that's what games do, not because it's a good fit or I have a good reason for it.
What if instead the terrain is hazardous and you must maneuver around it and build infrastructure to get around safely? That compliments the building oriented design, probably makes the map way more interesting, and your progress is permanent since your bridges and platforms and whatnot will remain there. On one hand it makes the map way better and more suitable for the game, but on the other it's incompatible a bunch of other things I was planning. It changes everything about how I imagined the game.
The best way to find the answers to your questions is to start making the game and exploring your options. Try it without combat, try adding a little combat, does it work better? Then add a little more. Still good? Keep going until you've explored the space well enough you can answer the question of how much combat to add to the game.
It's very hard to just try something like this because it changes everything. The combat is good if the combat is good, but is good combat a good fit for the game? I would have to make 2 entirely different games in order to properly "try" it.
As an example, what role does weapon and armor crafting play in the game? What kind of loot do you find? When you search for resources what are you hoping to make with it? Are bossfights key points of progression? Whether there's combat or not changes so much that it's not a "just try" kind of situation.
>I would have to make 2 entirely different games in order to properly "try" it.
Yes. You won't be able to confidently design a good game until you've completed some shit games.
If you just need someone to make the decision for you, then here you go. Don't include any combat at all. Reducing scope and creating a more focused experience is usually better, especially for a small/one-man teams. Sometimes you just need to make a decision without knowing if it'll be good and push forward, because at least your making something.
>someone to make the decision for you
That's not what I was trying to do, but I can see how it's easy to think that everything needs to have a simple solution.
>You won't be able to confidently design a good game until you've completed some shit games
I don't think that's related. Making a shit game gives you game making experience, but it has nothing to do with design. Design is about making things that work together and compliment each other and make a cohesive whole, it compromises of everything that's in your game together. Making a combat focused game no matter how good it is doesn't really teach you how to design a building game, you just learned to make better combat gameplay (assuming you actually thought about it and didn't just haphazardly make a shit game to get something out of the door).
>at least your making something
I don't want to "just make something". I don't care about money or making a "career" out of games, I'm on this because it's something I'm passionate about. There's very few games in the world that I would consider exceptional and almost all games I play have lots of what I see as missed potential, even very obvious and easy stuff, mainly because not enough thought went into the design and/or people were thoughtlessly doing what everyone else is doing even if it's a bad fit for the game. It's uncomfortably common to see things in games that, if just flat out deleted from the game, would make the game better with no additional changes.
>I don't want to "just make something". I don't care about money or making a "career" out of games, I'm on this because it's something I'm passionate about.
Then why does it matter that you would have to make two games to answer your design questions? Why does it matter that you would have to make 100 games to find the solution, if the point is working out of passion?
I never said you have to keep what you make, I'm saying that if you're not careful you could flounder for years trying to figure out the answer to this design problem and come away knowing less than if you had spent a couple months working on the game. The point of working on prototypes is that it's 100x better at telling you what works and what doesn't than sitting in an armchair pondering the nature of how the mechanics should fit together. You already know this to be true. You talk about how you see games with obvious and easy solutions to their problems.
>and almost all games I play have lots of what I see as missed potential, even very obvious and easy stuff, mainly because not enough thought went into the design and/or people were thoughtlessly doing what everyone else is doing even if it's a bad fit for the game.
You've made it clear that your ability to find these obvious and easy solutions doesn't work on games that don't exist (specifically, your own game), so why not direct your problem solving powers where they do work, against a shitty game that actually exists? Make a shitty game and fix it. Stop treading water and being all "thoughtful" as you achieve nothing. You're not going to do it right the first time, so stop trying to. Pick what you think is the best option out of the available ones, or even a random direction if you genuinely cannot judge the quality of your possibilities, and just do it. Once you have your shitty game created thoughtlessly, you can accurately judge which direction to go and implement all of the obvious and easy solutions that now reveal themselves to you.
We can't all be Nicola Tesla's who can picture the finished project fully formed in our minds. The rest of us have to actually build it.
I have a HUD element that represents the charge of a player's weapon that is layered like this:
>background (dark blue)
>charge indicator (light blue)
Both elements have the same shape and dimensions and occupy the same space. What I would like it to do is that, as the weapon charges up, the charge indicator gradually covers up the background graphic until it is fully charged, then when the weapon is fired it releases back and charges up again. I think I have to use some form of masking like so but I really have no idea how to go about this.
This is Godot btw, and it seems the progress bar node almost does what I want, however this only supports a rectangle element and not a user-defined polygon.
Easy to do with a shader. If pixel's y co-ordinate < charge % * element height, use colour from texture one, otherwise texture two.
You can probably do it with a progressbar, but a shader will give you more flexibility. It should be a relatively simple one to learn with if you don't have any practice with shaders yet.