Vapourware Codexian Game Development Thread

[bionicman]

Marketing is very tedious in today's Internet. Most people spend their time in walled gardens with very strict rules. For example, you cannot just go to Reddit, find the relevant subreddits and post your work there. No matter what you are marketing or how you do it. You have to be "part" of a subreddit's "community" (=suck everyone's dick) for weeks before they allow you to make even a single post about your work there. Even then you are fighting an opaque algorithm box and it just randomly might shadowban your posts.

This is why the influencer economy is so valuable. Going through these people is one of the few ways to actually make people aware of things existing.

This site used to be better, but here are some idea's that might be valuable:
https://draft.dev/learn/marketing-checklist

Are there any infuencers for 90s style 2.5D isometric RPGs?

Spoiler

https://www.youtube.com/@GAMEGODFLUENT

 

[bionicman]

For real though, look up other released games of the same type on youtube and twitch and you will find ppl willing to play such games.

 

[The_Sloth_Sleeps]

This took far longer than it should

there is 0xffffffff point to craft an isometric 2d engine anno 24
admit it and move on

There are zero functional 2.5D isometric engines in 2024. Admit it and move on.

Admittedly I'm entering this conversation late, but Unity is a functional engine that can create isometric 2.5D games. Fairly easily to boot.

 

[Tweed]

How do I into game dev? I'm trying to learn Godot, but I am simple and dumb and skipped CS.

 

[The_Sloth_Sleeps]

There are zero functional 2.5D isometric engines in 2024. Admit it and move on.

Ex-Fucking-Actly!!! (except openxcom)
you know why? cause there is no fucking advantage to it! only pain and wasted time
if you are writing a renderer, a lighting system, all the ai, all the tooling for it, for something that should see the light of day: you - are - a - moron
the pseudo 3d just adds a ton of non-functional worthless work on top to make it look proper
you either go rudimentary 2d and focus on innovative, interesting features that dont exist on the market or you grab an engine that has at least all the visuals functioning so that you can focus on kewl features
retarded faggot

God no! Fake 2.5D using 3D looks awful.

 

[bionicman]

How do I into game dev? I'm trying to learn Godot, but I am simple and dumb and skipped CS.

you don't really need CS knowledge to code in Godot or unity. I started with just trying to make a game I'd like to play. The more time you invest into playing with a game engine the more you learn about it and how it works, so it's just about the time invested really, like any other skill. Also, once you learn one of these engines it will be easy to switch to another, e.g. I used to use unity a lot before I got my job and never touched unreal, but now i am an unreal programmer (although I still use unity for my side project). Godot's scripting language is a bit like python so I think it's a great starting point, although I found its architecture with scenes and nodes a bit confusing... scenes are kind of like prefabs in unity and are not literally scenes but can be characters and similar, it's just a template container which stores nodes.

I recommend you try this tutorial for godot scripting language: https://gdquest.github.io/learn-gdscript/

 

[Viata]

How do I into game dev? I'm trying to learn Godot, but I am simple and dumb and skipped CS.

Studying CS is not required for making a game. Learning Godot is good enough, see some tutorial making a simple game so you get the ideas of how making a game works and then try to make toy games or literally start the game you want to make. But first do follow some tutorial to get some concepts down. However, avoid the problem of thinking you need to watch more tutorials before making your game. It can get to a point you are never making anything because you think you don't know enough. Everything you don't know, will be learned as you try doing it. A tutorial is good to get the basics down, everything else you search as needed.

 

[baboogy]

Going to post my WIP here, and try to give weekly updates, in the hopes of keeping myself honest in working on it. Aiming for an hour a day at the minimum after work, family and other obligations considered, but sometimes shit happens and I miss a few days. I have no game dev experience to speak of so this is as much a learning project for me as it is one of passion.

In that context, I would love if anyone had some good reading recommendations that touch on the more conceptual layout of a game engine, preferably one that has as little to do with programming as possible. Meaning more along the lines of, in the abstract sense, how is a typical game engine laid out, what are their most important components, how do they interact with each other, etc. And any design pitfalls I may run into going into this blind that will leave me kicking myself later and wishing I didn't have to do major rewrites.

Anyways, I've always loved blobbers, been a genre I come back to time and time again, and after recently binging through a few in a row decided I want to try my hand at making a Wiz1 clone since it seems like a fairly approachable project. Using C + raylib for the GUI + OpenGL + lua, and I decided to start by getting some GUI/OpenGL practice in by working on the toolset I'm going to use. My end goal is to have as much as possible outside of the engine, I want to be able to sit down and bang out maps, quests, puzzles, items, creatures, classes, with my own tools to save time.

Here is my WIP mapmaker/renderer that allows me to view the map in real time and walk around as I work on it. Drawing/viewing the map and picking individual textures for walls/ceils/floors/objects (or using the default texture set for all) is currently supported. Currently working on a way to attach lua scripts in editor to various objects, tiles, etc. that are saved/loaded with the map and I can edit to say, have a tile teleport the party somewhere else, link stairs/exit points to another map file, require a certain item to open a door, or require some game flag or world state to be set before progression is possible.

Here's a short webm of it in action.
https://files.catbox.moe/qlah3t.webm

 

[std::namespace]

And any design pitfalls I may run into going into this blind that will leave me kicking myself later and wishing I didn't have to do major rewrites.

it all depends how complex teh whole game will get
if you do c++/java style oop, dont use inheritance to characterize things, use composition, google entity component system
design the application of effects to agents before you write any code, especially if they interact...
go to oreilly media and search for game engine book...?

 

[Viata]

In that context, I would love if anyone had some good reading recommendations that touch on the more conceptual layout of a game engine, preferably one that has as little to do with programming as possible.

Tbh, I don't think there is a better book than this: https://www.amazon.com.br/Engine-Architecture-Third-Jason-Gregory/dp/1138035459

 

[pumpkinteractive]

Marketing is overrated, I don't think it's the problem here, nor the price, it's the art style. I like rpgs (not all tbh), but I find this too retro for me. pumpkinteractive maybe you should try your next game to have modern interface and gameplay but imitating the retro look? E.g. I really liked the skald demo https://store.steampowered.com/app/1069160/SKALD_Against_the_Black_Priory/ this game looks retro, but feels modern somehow. But I'm a zoomer so idk... I get intimidated by such low res and huge pixels :D

also, btw steam algorithm catches up after first 10 reviews: https://howtomarketagame.com/2022/01/25/why-your-first-10-reviews-are-the-most-important/

Hey I just saw your review (I suppose it's you because you refer to yourself as a zoomer), thanks for your support, though I don't know if I should be happy because it reads like you bought the game out of pity for me lol
Thank you in any case, you are definitely very kind !
The price might be too high because I saw a drastic increase in wishlists (over 2000%) on launch day but just little converts from previous WL's, plus someone in the chat was saying it's to expensive...
I'm not here to lament or anything, I'm also grateful for the negative response because I can improve on what's clearly broken (I don't know how I completely missed the wrong background in the city).
I think I should also highlight more that it's not a RPG in the common sense, it's more a adventure game with combat.
I think there are some wrong expectations when the description reads it's inspired by classic RPGs which refers to the art style, not the mechanics.

 

[baboogy]

And any design pitfalls I may run into going into this blind that will leave me kicking myself later and wishing I didn't have to do major rewrites.

it all depends how complex teh whole game will get
if you do c++/java style oop, dont use inheritance to characterize things, use composition, google entity component system
design the application of effects to agents before you write any code, especially if they interact...
go to oreilly media and search for game engine book...?

I don't plan on making anything terribly complex for my first real project, Wizardry 1 seems perfect for that. Read a little bit about this entity component system you mentioned, that does seem like a very flexible approach, thank you. I'll do some more studying about it in depth.

In that context, I would love if anyone had some good reading recommendations that touch on the more conceptual layout of a game engine, preferably one that has as little to do with programming as possible.

Tbh, I don't think there is a better book than this: https://www.amazon.com.br/Engine-Architecture-Third-Jason-Gregory/dp/1138035459

Thanks! Looking at some indexes for it I found online, this looks very comprehensive and exactly what I was looking for. I'll give it a read before I start putting any real work into the engine part.

 

[Curious_Tongue]

How do I into game dev?

Find beginner tutorials on your preferred engine. Follow along even if you have absolutely no idea what they're actually doing.

Keep doing them.

 

[bionicman]

Marketing is overrated, I don't think it's the problem here, nor the price, it's the art style. I like rpgs (not all tbh), but I find this too retro for me. pumpkinteractive maybe you should try your next game to have modern interface and gameplay but imitating the retro look? E.g. I really liked the skald demo https://store.steampowered.com/app/1069160/SKALD_Against_the_Black_Priory/ this game looks retro, but feels modern somehow. But I'm a zoomer so idk... I get intimidated by such low res and huge pixels :D

also, btw steam algorithm catches up after first 10 reviews: https://howtomarketagame.com/2022/01/25/why-your-first-10-reviews-are-the-most-important/

Hey I just saw your review (I suppose it's you because you refer to yourself as a zoomer), thanks for your support, though I don't know if I should be happy because it reads like you bought the game out of pity for me lol
Thank you in any case, you are definitely very kind !
The price might be too high because I saw a drastic increase in wishlists (over 2000%) on launch day but just little converts from previous WL's, plus someone in the chat was saying it's to expensive...
I'm not here to lament or anything, I'm also grateful for the negative response because I can improve on what's clearly broken (I don't know how I completely missed the wrong background in the city).
I think I should also highlight more that it's not a RPG in the common sense, it's more a adventure game with combat.
I think there are some wrong expectations when the description reads it's inspired by classic RPGs which refers to the art style, not the mechanics.

I will rewrite it so it doesn't sound like I bought it out of pity. I played the game a bit, and it's actually kind of fun (although I initially did kind of buy it out of pity LOL). Make sure to get to those 10 reviews as quickly as possible btw, this is how algorithm starts to help you out. Tell ppl to write a review.

I don't think the game is overpriced, you can put it on a sale later and ppl who find it overpriced now will buy it then.

 

[Bester]

What's this about 10 reviews? What's the "algorithm" you keep mentioning?

 

[bionicman]

What's this about 10 reviews? What's the "algorithm" you keep mentioning?

I have no experience with publishing on steam, but in the indie dev circles people have noticed that their game has increased visibility (I guess through discovery queue and the "New and trending" section, and maybe suggestions) after getting the first 10 reviews. I'm guessing steam has some sort of algorithm to hide most garbage games this way.

 

[bionicman]

Bester there's this Chris Zukowski guy who researches/investigates how steam works and how to increase your game's chances in the store (kind of like SEO but for steam games), he's got paid courses (i haven't bought those yet), but some informative free videos too

 

[Bester]

What's this about 10 reviews? What's the "algorithm" you keep mentioning?

I have no experience with publishing on steam, but in the indie dev circles people have noticed that their game has increased visibility (I guess through discovery queue and the "New and trending" section, and maybe suggestions) after getting the first 10 reviews. I'm guessing steam has some sort of algorithm to hide most garbage games this way.

Do they have to be reviews from bought copies or do free copies count?

 

[bionicman]

Allegedly only bought copies count https://www.reddit.com/r/gamedev/comments/zpnwfh/did_you_know_when_you_get_10_reviews_on_steam/

 

[Twiglard]

For people recommending the use of an entity component framework— imagine you're implementing Fallout 2. How'd you encode the differences between static scenery, scripted scenery (e.g. a ZAX terminal or vendor shelves), plain unlocked doors, and locked and/or scripted doors? That involves object proto used to instantiate it, per-object internal state, sprite, and update member function.

If you prefer composition over inheritance I'd like to hear it too.

Let's say you have such properties:

All objects have some intrinsic properties such as rotation, position and bounding_box. All scenery can be interactive, and currently active which are boolean values. Some scenery may be currently playing an animation where its current frame is an unsigned integer. Doors may, when active, be currently opening or closing— an additional bit of state is then necessary to express it.

 

[Bester]

Some scenery may be currently playing an animation where its current frame is an unsigned integer. Doors may, when active, be currently opening or closing— an additional bit of state is then necessary to express it.

An overkill, needed only for multiplayer, when a person can jump into a scene and be sent information that the door is currently being opened at a certain percent progress.
In a single player game, the door has two states, and the in-between is animated, but the state of the animation is not serialized.

 

[The_Sloth_Sleeps]

Well I consider ECS far too much for what I need. I dont see the need to over complicate. In fact I prefer to have a system which has concrete in the foundation instead of the wankish "anything plugs into anything" approach (no thanks).

Now, I have triggers, actions, objects you can interact with, script based objects, etc.

But basically I make two distinctions. It is either a scene object or a game object.
All these objects have common properties such as rot, pos, bounds, etc.

A scene object can always be potentially interacted with and exists in a scene.
Could be a fireball, a light, a trigger, a door, a house, etc.
This has a set of 20-30ish properties which says what it can/cannot do.
These properties have an associated script which is executed if that condition for that property is triggered.
(These objects have a template form stored in Json. It can read in out at any time and thus we can create many variations outside of code itself).

CanInteract -- interaction triggers the interact script
InteractScript -- run script when interacting, can also be null

CanPick -- can this object be registered to the inventory
PickupScript -- what happens when its picked up

others include;
Locked
LockLevel
IsDoor
IsOwned
OwnerId
IsWeapon
IsWearable
IsEdible
..etc

I just have these as properties I guess you can have a map with null properties. (But C# supports null types so why bother).

So if you have a door I simply write a door template in Json, it will have a lock level etc.
Any "special" doors like booby trap doors for example would just use the door template and have a script that differs from the door template script, which triggers damage to whomever is executing the action (caller being an automatic parameter passed to the script engine).

For me this allows for great flexibility yet the core functionality is kind of locked in.

 

[zwanzig_zwoelf]

How do I into game dev? I'm trying to learn Godot, but I am simple and dumb and skipped CS.

Godot is weird, either it gels with you or it doesn't. I found it less intuitive to get into compared to Unity or even UE3.

 

[std::namespace]

This has a set of 20-30ish properties which says what it can/cannot do.
These properties have an associated script which is executed if that condition for that property is triggered.

thats an ecs with composition, its a fancy name for old school c style data driven code

gaylord = { // entity
age = 13 // component
canfuck = true // component
}

system = select_entities_with_following_components(age, canfuck)
system.do_shit()?

How'd you encode the differences between static scenery, scripted scenery (e.g. a ZAX terminal or vendor shelves), plain unlocked doors, and locked and/or scripted doors?

i dont think thats a great usecase example for an ecs
where is the hard conflict in the inheritance hierarchy?

inheritance:
weapon -> sword
weapon -> wand
weapon -> swordwand // thats a class with copypasted code from sword and wand

ecs composition:
sword [melee]
wand [casting]
spellsword [melee, casting]

inheritance:
tile -> static
tile -> interactive
tile -> scripted
tile -> interactivescripted

ecs composition:
tile [static]
tile [interactive]
tile [scripted]
tile [interactive, scripted]

https://ericlippert.com/2015/04/27/wizards-and-warriors-part-one/
https://github.com/bakpakin/tiny-ecs

 

[The_Sloth_Sleeps]

This has a set of 20-30ish properties which says what it can/cannot do.
These properties have an associated script which is executed if that condition for that property is triggered.

thats an ecs with composition, its a fancy name for old school c style data driven code

gaylord = { // entity
age = 13 // component
canfuck = true // component
}

system = select_entities_with_following_components(age, canfuck)
system.do_shit()?

How'd you encode the differences between static scenery, scripted scenery (e.g. a ZAX terminal or vendor shelves), plain unlocked doors, and locked and/or scripted doors?

i dont think thats a great usecase example for an ecs
where is the hard conflict in the inheritance hierarchy?

inheritance:
weapon -> sword
weapon -> wand
weapon -> swordwand // thats a class with copypasted code from sword and wand

ecs composition:
sword [melee]
wand [casting]
spellsword [melee, casting]

inheritance:
tile -> static
tile -> interactive
tile -> scripted
tile -> interactivescripted

ecs composition:
tile [static]
tile [interactive]
tile [scripted]
tile [interactive, scripted]

https://ericlippert.com/2015/04/27/wizards-and-warriors-part-one/
https://github.com/bakpakin/tiny-ecs

I am talking about why I dont adopt an ECS framework like https://github.com/Chillu1/CSharpECSComparison Basically I dont want these kinds of things in my life, introducing their bugs and weirdness.

The spiel for these ECS are something like "Real ECS framework for c#. Enables to write encapsulated, decoupled, maintainable, highly efficient, data oriented, cache friendly, code without pain. Although the framework is platform agnostic (compatible with c# 7 and above and .net standard 2.0 and above), it comes with several Unity extensions." or some similar kind of horseshit.