Vapourware Codexian Game Development Thread

[Catacombs]

Next enemy:

Seems that the Codex's mascot has taken up body-building.

Giga Codex Demon Chad

 

[AdolfSatan]

And I drew this portrait today.

Any portrait artists here with some feedback? I think it's not awful. But it's nowhere as good as those Korean dudes on Artstation. Maybe one day I'll be able to sit down and do one digital painting a day for a few months and then I'll really kick ass at drawing. For now I have to keep splitting my attention among all of the other gamedev tasks.

There's the codex drawing thread where you might get more views/opinions. imo it looks kinda faggy because of the anime eyes + low contrast, here's a quick paintjob to kinda portray what I mean by it

Spoiler

 

[Morpheus Kitami]

And I drew this portrait today.
Any portrait artists here with some feedback? I think it's not awful. But it's nowhere as good as those Korean dudes on Artstation. Maybe one day I'll be able to sit down and do one digital painting a day for a few months and then I'll really kick ass at drawing. For now I have to keep splitting my attention among all of the other gamedev tasks.

Not exactly the best at it myself, but I'd say the problem is the eyes, mouth and the "mustache". Its hard to tell what you're going to do on the finished drawing, but the black lines you left made the eyes look girly and a psuedo-mustache growing out of his nose. That's not all of it though. The eyes as-is look too small and the whole eye-mouth package look like a horror antagonist's face, not an arch-demon/devil figure. Give it a better contrast around the teeth and redo the eyes.

 

[Catacombs]

imo it looks kinda faggy because of the anime eyes + low contrast

Shit. It's way better than the stick figures most of us draw.

 

[Nathaniel3W]

imo it looks kinda faggy because of the anime eyes + low contrast

Shit. It's way better than the stick figures most of us draw.

Thanks. And maybe it is better than most people can draw. But artists who are content with "better than most" probably don't progress beyond that.

 

[Bad Sector]

Lately i ported Post Apocalyptic Petra to DirectDraw (looks pretty much the same as the DOS version) and Direct3D 7, meaning that it can now run on early 3D graphics cards. So here it is running on my retro PC with a Voodoo 1 . Also optimized the software renderer a bit (including rewriting the inner loop in assembly) which had a +40% increase in performance, making the game actually (mostly) playable on a 200MHz Pentium (still not as fast as something like Tomb Raider but considering the renderer is too generic and i just happen to use boxy levels whereas TR1 takes advantage of its boxy levels to speed up rendering, i doubt i'll ever reach its performance without rewriting the rendering code completely - which i'm not that interested in doing ATM, i only take small jabs at the code whenever i do not feel like doing anything else in other projects i'm working on :-P).

I also updated the SDL backend so the game now compiles and runs on Linux too, but i haven't written any audio code meaning that while it runs, it is completely silent (might be an improvement considering my sound effects though :-P).

I'll probably try to finish the SDL port to be fully usable and write a Windows-based configuration utility too since the current one is only for DOS and under modern Windows you can only configure the game by editing the settings.cfg text file.

I think by that point i've spent way more time tweaking the game's code than i actually spent on writing it in the first place :-P.

 

[Ninjerk]

That's.... dedication.

 

[Nathaniel3W]

AdolfSatan better?

 

[AdolfSatan]

Fuckin A, man! You got the menacing look down.

 

[Metronome]

So, i finished the game. Well, "finished". I was fixing bugs until the end, i managed to submit it just three minutes before the deadline :-P

I got stuck on the part where you had to find the lube. The atmosphere is really good but controlling the protagonist is painful.

 

[Quilty]

Working on a submarine simulation in a post-apocalyptic world where most things are procedurally generated: the world, its cities, the crew you can recruit, the hostile creatures and other submarines you can encounter, and the missions you can take on.

Currently the only implemented mission type is a derelict exploration type of mission, which allows you to send your chosen crew members on a turn-based, roguelike style exploration of an abandoned submarine or other vessel.
The entire concept is a bit like Battle Brothers under the water: you maintain your crew's morale, food, and other needs while also maintaining and upgrading your submarine and taking on missions and earning money and respect.

The video linked above shows a build from earlier this week (been working on the game since January 2020), and much of the UI is functional but obviously not well-organized or perfectly aligned. Just there to give you a basic impression of the concept. The crew portraits are also currently only implemented for male characters, which means female characters have male portraits. Very codexian, I believe.

I'd appreciate any feedback.

 

[Bad Sector]

So, i finished the game. Well, "finished". I was fixing bugs until the end, i managed to submit it just three minutes before the deadline :-P

I got stuck on the part where you had to find the lube. The atmosphere is really good but controlling the protagonist is painful.

Which version do you have? In the latest version (0.99h), you can use WASD+mouse to move around. Alternatively you can use the arrow keys and the controls are similar to classic Tomb Raider games. Or you can use the Windows version with a gamepad, though the gamepad controls were hacked in (it translates analog stick motion to keys), there isn't true analog motion.

If you can't find the lubrication

Spoiler

after you open the first trapdoor you drop into a small area from where you drop into the larger area with the electric panels you need to jump over. Before dropping into that larger area, check the smaller area first, there is another area with electric panels above the drop to the larger area - you need to jump over to a corner, climb and move sideways before entering the area where you can find the lube

 

[Morpheus Kitami]

Working on a submarine simulation in a post-apocalyptic world where most things are procedurally generated: the world, its cities, the crew you can recruit, the hostile creatures and other submarines you can encounter, and the missions you can take on.

This is actually really, really nice, but there's not much to go on yet. I would say that the constant use of active sonar is a bad idea, revealing your position to your enemies IRL, since that's what other sub games have drilled into my head, but I believe you already know that.

 

[Quilty]

Working on a submarine simulation in a post-apocalyptic world where most things are procedurally generated: the world, its cities, the crew you can recruit, the hostile creatures and other submarines you can encounter, and the missions you can take on.

This is actually really, really nice, but there's not much to go on yet. I would say that the constant use of active sonar is a bad idea, revealing your position to your enemies IRL, since that's what other sub games have drilled into my head, but I believe you already know that.

Thanks for the feedback!

The short video I made was just for the purposes of gauging people's initial reaction (I'm interested to know if the UI strikes anyone as garish or a pain to look at, etc.) but doesn't show things that are already implemented: procedurally generated map with a depth for the water area, recruitable characters, town locations, enemies, missions (well, one type so far), vendors in cities, traders moving around the map, as well as the option to go on a mission within a derelict vessel and explore it in a roguelike style.

I'll make sure to put together a better preview of all the implemented and planned features and show everything in action.

I know what you mean about the sonar, and that works as intended: you can either use active sonar, and reveal yourself to enemies within their "hearing" range, or you can use a passive sonar and pick up the other vessels' signatures. Of course, in that case you can't see the location of the environment and might ram straight into an island.

 

[Bad Sector]

Uploaded version 0.99i of the game with all the performance improvements and new features i mentioned previously. I didn't make a Linux version since i haven't implemented the sound code yet. I think i'll implement the sound code at some point and perhaps even finish the Android port and call it done, unless there is some huge bug (or change my mind, it happens :-P) i most likely wont release a new version after that.

I'll probably reuse the engine for a future DOS jam though.

Anyway, here is the changelog for the new version:

(Win) Added DirectDraw and Direct3D 7 backends (can now run natively in Voodoo 1!)
(Win) Added wincfg.exe Windows-based configuration utility
(Win) Fix crash at startup when the sound was disabled
(DOS) Added readme.bat and readme view mode for the configuration utility
(DOS) Added dirinfo file for Norton Commander
(SDL) Fixed SDL 1.2 compilation, the game builds. No binaries though since sound is not implemented yet.
(All) Optimized the software renderer and rewrote the inner loop in assembly
(All) Added frustum culling for mesh scene nodes
(All) Added a not-so-quick-yet-still-very-dirty PVS which is used for occlusion culling
(All) Fixed grid geometry generator to not generate geometry for occluded faces - in addition to improving performance, this produces a cleaner output when the camera clip through the world geometry
(All) Rendering optimizations give a ~+40% performance increase for the DOS version, making it playable on a 200MHz DOS PC (though still far from smooth)
(All) Added descript.ion file for file managers that support it (e.g. Total Commander)

 

[zwanzig_zwoelf]

Slowly writing my own pathfinding module. Not sure if it'll come along quickly enough to be used in the current project, but it's still going to be useful in the future.

After using Aron Granberg's A* Pathfinding solution for DGS and DGT I've noticed that I wasn't quite satisfied with it -- it's a good solution, but there are a few things that restrict what I can do, and extending it to support things that I wanted to have would've taken too long for my taste. Besides, my desire to avoid using someone else's code (unless writing my own solution would be a big waste of time) only got stronger recently.

While my current goal is to reach the same level of functionality used in DGT and lift whatever restrictions I had to deal with, my ultimate goal is to eventually develop it into a personal pathfinding solution that can work with environments like these.

 

[Tavernking]

Went with Godot and C#, simply for the speed of development in Godot and my familiarity with C# (Coming from a Java background, they look similar at face value).

After years of hanging around various dev groups I've seen too many devs focus on technical efficiency or deeply complex features ... and then never finish their game. It doesn't matter how impressive your prototype is, a finished game is better than an unfinished game.

 

[Immortal]

Went with Godot and C#, simply for the speed of development in Godot and my familiarity with C# (Coming from a Java background, they look similar at face value).

After years of hanging around various dev groups I've seen too many devs focus on technical efficiency or deeply complex features ... and then never finish their game. It doesn't matter how impressive your prototype is, a finished game is better than an unfinished game.

"Finished Game" is a bit of an ubiquitous term.
I empathize with your comment but beware the "I'll just slam a buncha shit in and cobble it together" flip side of that coin.

 

[Bad Sector]

Recently started working on a new engine since my previous one (Runtime Engine) has several issues that aren't worth fixing without rewriting large parts of it (though the engine is tiny so it isn't that much of an issue) and i wanted to try a different design approach anyway.

I have a couple of game ideas in mind for using it but for now i need to focus on the core aspects first (and hopefully avoid the mistakes i did with the previous engine and editor). One thing i'm focusing with this engine is to make it easy to create stuff with it at the cost of pretty much everything else (within reason of course) by taking advantage of how ridiculously fast modern PCs are so i can cut a few (or a ton of) shortcuts :-P.

Also i took some ideas from making Post Apocalyptic Petra which (aside a few limitations because of it being basically a game designed for DOS) has a codebase and editor that are easy to work with (actually a few bits of the code are taken directly from it, though with several changes).

 

[Nathaniel3W]

In keeping with JRPG tradition, the heroes of Himeko Sutori must kill a god in the final battle.

This took a ton of work over the last few weeks. I finished up the demon lord sprite and portrait, then gave them some different weapons, along with some character classes and unique attacks. Then I made mister Tyrant God there, who also has his own unique sprite and portrait and unique armor, weapon, and character class. And then I had to write a bunch of new classes to introduce him in the middle of a battle. You fight the guy who you think is the final boss, and when he's dead, this guy appears.

I think it's all coming together pretty well, except now maybe the final battle is getting too long. I'll wait for player feedback and then I'll tweak it.

 

[Yaar Podshipnik]

Lovely stuff! IMO however, the pixel art of the sprites, and the simpler color gradation there, kinda clashes with the clean vector with smooth gradient look of the portrait.

 

[Ninjerk]

I don't see the issue.

 

[AdolfSatan]

Looks like bitmap to me rather than vector. But that aside, I think it fits with the hi-res theme of the UI and serves to set it apart from the game grafix. Sprites are fantastic btw, amazing job.

 

[Nathaniel3W]

The portraits are all bitmaps, but the general shapes and shading were done partly with Photoshop's path tools. The UI is part vector (made in Flash) and part bitmap. Mixing the UI, the 3D world, and the 2D characters has been a huge challenge from the beginning. When people say that things don't quite look right together, I get what they're saying. But at some point I have to say, well, I could work on this and redo it forever to try to get it right, or I could say it's good enough and just try to finish the damn game. Unless I see something with an obvious solution now, then I'm more inclined to take the latter approach.

 

[Yaar Podshipnik]

Nathaniel3W your game certainly looks "good enough" - I would go as far as calling it very good . Definitely stick to finishing the game and polishing gameplay, it's way more rewarding that catering to the little graphics whore inside some of us