Post-mortem: "Oh, sh*t!"
About
First of all, hi there and thanks for reading.
Let me introduce myself, I'm Berny a.k.a xPheRe and I've been working as a web developer for the last 10 years.
Now I'm looking to find a place where I've always belonged: videogames.
And that's where I am now, writing about my unfinished project Oh sh*t! featured at Ludum Dare 42.
I think it would be interesting to share my thoughts during the development of the project.
To me, at least.
(My apologies in advance for my English, it's not my first language)
Intro
Rewind to previous friday and picture the following: This is me, getting home late, after some beers. Now I'm turning my computer on and looking at my twitter timeline. Then a scream comes out loud from my throat... "OMG, LD#42 started already? OH, SH*T!"
Act I: Brainstorm
I was only one hour late, but here in Spain it was night time already. I decided to brainstorm some ideas before going to bed and let the pillow decide.
I wrote some ideas in little paper strips: - Space: Universe. Spatial concept. Keyboard key. - Running out: Reduce space with time. Fill space with elements. Run, literally, like Forest Gump. - Gameplay: Puzzle. Switches. Transformation. - Type: 2D platform. Top-down sokoban style. First person. - Environment: Computer memory. Schedule. Room. Wardrobe.
The next morning, after some good sleep, I carved some rules: - It should be simple. - I liked the computer theme. - A tetris-alike would be nice. - The game's name will be "OH, SH*T!". Everybody knows you need to start with the name.
Also I'd follow COMPO's rules in case I had something finished in 48 hours. Dude, I was so wrong about that...
Act II: I walk alone
Can you believe it? I've been promoted to memory manager! Awesome! A whole memory cell to manage and fill with bits from the bus storage... Wow. I hope to live up to the expectations...
That's the beginning of the story, now I want to convert this into a game. Where do I start? - Created some simple pixel assets in Gimp. - Drew the worker, the bits, the pieces, the factory floor... - Added some animations to the main character: idle, walking, talking, picking an object... - Made some plans with those assets, drawing mentally what I wanted the game to look like.
The worker must go to a conveyor belt, the input queue, to pick a package, then go to the main grid and place it. Bits placed in the grid can't be walked over, so each piece makes the next harder to put. In order to empty some cells there'll be an output queue with piece requests which you can fulfill. Also some packages are transformations that you can apply to a cell, like "clear all connected cells to this one that share color".
I opened Godot and start placing the assets here and there. Everything looked gorgeous and I had a walking player in a gray factory in less than 30 minutes. Then I started to code grid's logic and everything crumbled.
Cut-scene: The path to decay
It was almost a day in the jam and it came clear to my mind: Everything I've got until then were the secondary actors. The main character isn't the worker or the factory... it's the grid.
And since I had a hard time putting the grid to work with the rest, I decided to remove everything else. I even decided to pause development until the next day and let the pillow decide again. My pillow's a good adviser.
Break
But the pillow didn't help that much.
Sunday came full of social events and I wasn't able to touch a single line of code until late. But I had the time to think about the direction of the project. And that was the problem: It had no direction at all.
I've been working like a headless chicken, without planning. And also ignoring the first rule: It should be simple.
My next step should be to open a Trello and start organizing myself. I won't give up on the project, I'd be working until the last minute to ship it as JAM at least.
Back at home again, turned the computer just to look with dismay how my internet connection wasn't working at all.
Act III: Focus shift
The next morning my ISP fixed the connection problem, time I took to start a greenfield project just with the grid. I focused on the main concerns, placing pieces, avoiding collision, rotating... Making the grid work on its own then stacking the other pieces on top.
When I got it working with the mouse I decided to left it as this. No need for the player to move the worker here and there. Simpler.
The rest of the assets started to build upon the working grid. But most of those wasn't used, worker animations, grid connections, conveyor belts... Time spent for nothing.
As it became clear that I wasn't going to publish a full game, I began to polish some things. - Added some sound effects generated with the awesome "sfxr". - Brought back the main character and the game's title. - Created a simple site at itch.io to expose the little thing.
Everything was in place, but not as I would like to be. But I finally managed to make something for a Ludum Dare.
And that, that was the BIG prize.
The Good
- Engine choice: Godot feels like a glove to my ideals. It's lightweight, open source, with a great community and behind the closest development team.
- Breaks: Taking breaks to sleep and eat well is key. When stuck, go for a walk, do something else. The inspiration will come.
- LDJam community: Received good advices in the comments, also support and encouragement. This is awesome for a newcomer with an unfinished project on the table.
The Bad
- The house from the rooftop: When working solo it's vital to have your priorities sorted. Know what's the key point of your game, develop this. Then add the rest. I did it the other way around.
- Organization: Managing time is essential in a 72h race, use any management tool at your fingertips, but choose one and stick to it. From the beginning.
- Forcing the theme: It's hard to find novel ideas matching the theme. Just get some old ideas and mix them. I tried too hard to build an original idea and this munched too much time.
- Too many ideas: Once you get a good idea, stop overthinking. Mentally adding too much content creates a burden on expectations. I wanted to do so many things in 72h...
- Grow by little: Take baby steps towards your next goal, those should be reachable from the previous one, but enough to make you feel advancing. Bite the big picture in easy installments.
The Ugly
- Music: Had I know about Abundant Music there'll be some great midi tune in the background. Great tool.
- Mouse is for PC: try to find a way you can play your game in most platforms. Mouse location is not that portable (think console controllers or cellphones)
- Noise and interruptions: Using a pomodoro technique would've ease the focus on development and breaks, even with no one's around to tell.
- Compare and despair: I checked #LD42 hashtag at twitter too many times, watching in awe other people's games. But that's fine, this is YOUR game. Be proud of it.
Ending
At first this project seemed to me like a big failure. Then I understand it can be used as an experience towards better contributions.
Comments on this unfinished game made me think again if I should work further to finish the game or not.
It would take some time, but it will be worth it. I'll made a full game of this "Oh, sh*t!" experience.
Thanks for reading and enjoy Ludum Dare!
