My Peculiar Postmortem

Posted by (twitter: @brantkings)
May 10th, 2013 4:02 pm

This is the postmortem for my entry: My Peculiar Car! If you didn’t played it, please do!

shotPost

What happened

Conception

Well, when the theme went online, I was on a party. I got home drunk so I just decided it was better to drink water, have an inspirational sleep and a shower. After all that, I kinda had the idea ready (I was lucky this time). What I thought: so, minimalism was doing the minimum to express the most? And combining that with an idea I had some time ago that changing a gear would be a puzzle while driving cars, I had maybe a good idea, and hopefully I wasn’t too far behind.

Programming some squares

So I went programming right away. Flixel is my engine of choice because it helps you a lot with basic stuff like States and animations. I decided to do first 5 discrete lanes and have a car being able to change between them first. I opted to not make them continual because it would be hard to make a parallel between a gear change and a car move. OK. With that ready, I had my first problem: how should I give the the impression of speed? I really thought about doing a top down view with parallel lanes and all, but opted for a top-gear like experience, using perspective.

I actually tried to implement these darker colors that made the impression of speed here, but it bogged down the fps so much, I left it away. Maybe I did it in a dumb way.

Ah, Top Gear. I actually tried to implement these darker lines of colors on the road that made the impression of speed here, but it bogged down the fps so much, I left it away.

While the top down would be really easier to implement, I wonder if in the end the game would feel less like an racing game and more like a double puzzle-game kinda like Frogger with more restrictions. I spend a lot of time implementing the perspective, mainly because with Flixel, use scale and position at the same time and you are never really sure of what is going on. I had to use a lot of magic numbers that mysteriously make everything work and I hate that. That is really one of my few problems with Flixel.

So, with the lanes, cars and obstacles ready, I had a silly little game hihi. But I still had to implement the gear. That was way easier than the perspective lane, and everything went pretty straightforward. I finished implementing it, and when I finally had the two together  the gear blocking the movement of the car, I actually thought I had a neat game idea. The problem was with the gear map generator: it was super random and sucked. Always guaranteeing that, with each new map, you can go left and right at least once made it a lot better, but in my opinion, it still lack something more.

Squares race! You can see how the first gear map generation algorithm sucked horribly here.

Squares race! You can see how the first gear map generation algorithm sucked horribly here.

 

Another thing that I did very rushed: the cactus positioning. It was completely random. It maybe still is. I tried to making it better by making the chance of cactus appear bigger in less crowded lanes, but that was it. It is still bad. If I ever expand this game, the level design is the first thing that needs to get better.

Drawing some stuff

On to the graphics, I really wanted to draw everything as simple as possible, without shades or anything. Then I completely threw that from the window while drawing the car, because I shaded and everything. It took a lot of time (it looks kinda cool). For everything else (except the fish because it is important), I tried following simplicity.

For most sprites, I drew a fast silhouette on Graphics Gale, changed it’s colors, and that’s it. I tried to exaggerate the man’s animations on the car to make the impression he is trying really hard, I don’t know if it passed that message. But if it ended up not terribly ugly, I guess it is hooray.

I concluded a pretty obvious thing: if you make it smaller, you draw better. So what I did was drawing the intro, game over and the outro images as 80×60 sprites, and then scaling them 10x in the game. It looks rushed, but it looks like something.

I tried to make the color palette like a really sunny desert, that goes well in the eyes, and I guess that went awesome.

Sound

Then, finally was the part of putting some sound in the game. This is where the game got downhill.

See, sound is a very important part of the game. But it ends up that flex only accepts 44000hz mp3 files. And wasting some time editing, recording and filtering to discover that your sound fx wouldn’t be able to be used in flex is so frustating it almost hurts. Some sounds did work though, like the motor one. I don’t understand why sounds recorded with my mic didn’t work, if it worked before (including on my last LD entry, Just Politics).

But that led up to a good thing: after failing 3 different sound effects for the gear, I stumbled upon a free sax solo on the internet. I said “why not” and that turned out to be one of the best features in the game. Way better than me screaming on your headphones when you achieved the sweet, blue spot. Haha!

I couldn’t compose any music, I wish I could, I’ll try to learn that.

Finally

I ended up just moving the GUI, asking my friends for a opinion (something that did help a lot!), and trying to polish the game the most I could. In the end, I was finding difficult clicking with the mouse. But then, the reception was great, and I liked it.

I didn’t really do it all in that chronological way depicted here, but more or less so.

So, to sum it up:

The final game!

The final game!

What went right

  • Well, I was for five minutes browsing through a certain free sound page on the net, and while I couldn’t find any honks I have been able to find a snazzy sax solo. Downloading it ended up being my best decision in this compo, as the sax sounds really got praised.
  • Simple, without shading pixel art with exaggeration is kinda beautiful, passes the message and it is easy to do. Don’t have time? Do the sprites on a smaller grid and then enlarge it.
  • It was really easy to screw up with this idea. Early playtesting people were just asking themselves why the car wouldn’t move. Adding a little trembling on a locked gear handle did the trick, maybe? Or the tutorial? Anyway, people didn’t ask that in the comments.

What was wrong

  • I really wish I knew what went so wrong about sound. This Audacity vs Flex fight was really, really frustating, and I ended up giving up a lot of sounds that could make the game so much better.
  • The cactus generation is of course the worst part of the game: a little less procedural generation, and more well-thought position I guess would be for the best.
  • I don’t like how the gear HUD look, I could do better (and the handle don’t look like part of the game at all).

PS.: Yeah, sorry for any engrish.


Leave a Reply

You must be logged in to post a comment.

[cache: storing page]