Development Tank

how and what

[bensiiiii]

Alpha Conte by Ben

A dark, atmospheric 2D game set in a futuristic research facility

the game has now been released, free of charge on itch.io!

a summary of the process, evaluation and experiences:

i created an atmospheric 2D game with a few puzzles and some interactive story elements. the game became much shorter than i anticipated, a lot of the reasons for this boils down to me being one person and deciding to implement things from scratch. 

developing the engine with an editor (which was also made from scratch with its own GUI system), doing all the art, scripting, testing, story states, sound fx, graphics fx, AI, tiled maps, game design, file-handling, particles etc... ended up taking so much time and effort that when i was faced with my initial deadline, there was no way to sanely improve or expand on the game itself.

from a game design perspective, the game has a bit too much linear progression, which i think is a side-effect of initially wanting a focus on story, but realizing that it was not possible in the time i had estimated. i had to force the content i had to work with into something that resembled a start to finish game with somewhat of a logical progression.

libgdx was the initial choice, and i was set on having the game be playable across platforms, but that idea quickly died as libgdx seemed to become exponentially more viscous to work with and i ended up focusing on desktop. 

i noted lacking areas such as documentation, consistency in naming and examples, ability to customize certain parts of the rendering pipeline.

there were some unexpected behaviors and systems i had to work around. all in all, it’s a decent framework IF you absolutely are going to ship to browser and phone.

it was a very character building experience for me to show that i can finish a complex project and execute long-term planning all without any other devices than myself and a simple framework. no engines, no artists, designers, existing assets, no nothing. 

i realized some valuable things about myself and the process, which are already greatly improving the way i work with my next project.

- B

[bensiiiii]

update:

- tightened up lighting

- added items

- more gui

- more graphics

- filled out the demo level more

- fixed some bugs with doors and terminal usage

- fixed most aspect ratio / resizing issues

todo:

- add config and options menu

- make sure transitioning levels works and persists states properly

- design and add last levels to the finished game

- add more items, entities and tiles to prepare for last polish

[bensiiiii]

i’m going wild with updates: showing off pathfinding with one of the NPCs as i said i would

- made particles slightly better

- made ai part of engine somewhat better and more flexible

i just wanna go so sky high

- B

[bensiiiii]

quick update:

- implemented particles (gif showing off spark particles with light-emission)

- implemented pathfinding using custom A* (will show example in next post)

- cleaned up some code and changed some game design decisions

progress is very steady, i’m so far staying on, if not ahead of time schedule for the june alpha demo

- B

[bensiiiii]

quick update on the project to keep everyone inside the loop. (picture is the main menu, which is animated and has music and sound fx)

have been working non-stop since last time, the game is now close to having a finished prototype. i’m ironing out engine issues and adding some necessary features like particles, proper resizing, AI and more

the game has a working console terminal which can be programmed to do basically anything inside the game. right now the debug level has a door that opens via the console.

i am aiming at releasing an actual alpha demo some time after mid/late mid of may, to prepare for alpha in june and beta in july. that is the schedule, but i can’t forget hofstadter’s law

meanwhile i am considering starting slowly expanding my networking to youtube and twitch. i am working all the time anyway, so i might as well start streaming consistently, this would simultaneously produce material for a youtube channel. it could be crazy, and i like it

- B

[bensiiiii]

serious progress since last post. 

i have meated in the editor a great deal, the current test level was created in it, without any issues. entities collide and can be scripted to interact arbitrarily with their environment.

collision was initially just intended to be straightforward O(n) by list traversal, but for fun i decided to go the memory road and implement tile tables such that entities register themselves through O(1) lookup. worst case i’m looking at O(n) traversal (as a side-effect) here as well, but that is theoretical, of course. practically speaking this is always much faster, based on geometric constraints, however it is less elegant to implement and maintain, it also hogs extra memory, but it’s apparently 2018, so.

there is a very simple main menu, and i’ve worked more on filling in the gamesave class.

the rendering pipeline has been tidied enormously, and it’s also possible to display text in various forms and ways now.

a little cloud list:

- finish gamesave and directory-collaboration fully for the main menu*

- start working on audio system*

- draw some more art and implement a few more entities for testing*

- add a few more QoL (quality of life) features to the editor for swiftness

- start fleshing out the first test-level with a puzzle, sound and etc effects and computer entity interfacing

no time for feelsbadman, let’s be useful... productive

- B

[bensiiiii]

u ever just do stuff?

i have made substantial progress on the engine, both editor and game client alike. also some improvements to subsystems.

as usual i’ve been finding myself improving a lot of things and gradually improving the overall structure of the engine. by documenting as much as i can about anything, everything and nothing, i seem to be making solid increments in the todo-list, and with high frequency.

in essence, i am answering big questions by giving myself smaller questions and iterating the process back on myself until the questions are so small there’s no way i can’t do them. by doing this i am also, with time, learning to simplify daunting problems, and avoid getting stuck

i have set up a rendering pipeline that is intuitive and easy to manage. i can do all sorts of stuff with custom shaders, dynamic lights and such.

there is state-shifting, a naive gui system, game events (scripting), naive resource management and saving/loading game-saves, levels, story states, and such.

immediately i am looking at implementing:

- game configuration (options)

- entity physics

- particle system

- audio system

the overarching goal is producing a playable prototype as fast as hellfire

- B

[bensiiiii]

fast and free of discomfort, the author has successfully implemented and introduced means to most of the items in the previous post’s todo-list. above you can see images exemplifying the progress.

there are now only two essential functionalities left to implement:

- selecting and editing the item

- completing the input dialog details for saving/loading/new level

- finishing implementation details around placing and editing light sources? *

(* denotes a less important task)

when these are done, i will start sharing actual game design and other engine programming progress, but first i’m going to space to molt, and will return henceforth fresh and lean

- B

[bensiiiii]

just some small features added to the editor. can now edit by layer and also toggle rendering by the selected layer. some internal additions such as level-specific handling (saving/loading) have been implemented, in concordance with a few more gui components, such as checkboxes, labels, dialog with input, etc.

i need one or two more things such as: 

- a drop down list for quickly selecting entities

- an object focus pane to modify attributes of entities and tiles

- selecting entities or tiles in the level pane itself, to modify existing objects

-  intuitive way of changing between editing tiles and entities

- making sure input doesn’t bleed into other elements ever by using elem-focus

when this list is old news we’ll be looking at moving back over to the actual game programming, and it’ll be extreme

- B

[bensiiiii]

first post, showing where my scene/level editor is at for a project i’m working on.

 i’ve invested some 1 year on and off into the engine development in total, and it’s finally starting to functionally and systematically come to a place where it’s possible to make some solid practical progress, that’s to say; actual interaction, producing content, so and such.

with time i aim to be sharing as much as is productively healthy and to be using this log as a sanity check for further development

- B