Someone To Love – post mortem

Posted by
December 20th, 2011 6:22 pm

Okay, so it’s time for a post mortem of Someone To Love! There’s also a timelapse, if you didn’t see it. Let’s dive right in – what went right and wrong, and what lessons are to be learned from my experiences during this Ludum Dare!

What went right.

  • Graphics. I mean, I usually think very little about my art skills, but this time I feel that my approach was perfect (for me, at least). I’ll admit that simplicity wasn’t something I thought of at first, but at some dead end while coding, I figured I could as well create the graphics. I used a total number of 6 colors, the main being white, with cat being 10×10 pixels and human 10×24 pixels. Then I used to change color tint of the scene, depending on the distance between player and the human. I usually tend to overdo things, so this simple solution (achieved after many hours of experimenting with other options) is brilliant according to my standards. Like the last time, at the beginnig I used placeholder graphics, which is always a great idea and I recommend it to everyone.
  • Music. Wolfram Tones. That’s it. Go, check it out if you don’t know this. I’m actually really interested if it’s possible to use this generated content in a commercial game – anybody got any info on this? I mean, I understand the technology is copyrighted, but I don’t think that this also aplies to generated music (I mean, Wolfram doesn’t have rights to every CA there is, does it?).
  • Flixel. I have issues with this piece, so read on below. Here I can write that Flixel is a great framework, especially if this is your first contact with AS3, and especially if this is your first contact with game programming framework. Automaticly done collisions, very simple map loading, many options, Flixel Power Tools for GREAT effects and much, much more.
  • Mood. I like the overall design of the game, the colors and cat monologues, and the direction it was heading when it was time to finish. I like that there are endings, and that there’s a few of them (last time I didn’t have time to make an ending for my game). It’s definitely not flawless, though – read on, below.

What went wrong.

  • Over-coding stuff. I didn’t have any idea regarding the theme, so I started programming some kind of platformer. I wasted a lot of time on nice effects when player leaves world bounds, only to design level in such a way that it’s impossible. I prepared my object model to be very flexible, but in the end it was a waste of time, because I didn’t need this flexibility.
  • Flixel. Yeah, when you’re used to programming, doing it with Flixel may be pretty crazy non-intuitive. I mean, collisions are done automagically for you! I know this speeds up the developing process, but this wasn’t something I expected, and made me scratch my head a couple of times (what if I wanted to do something non-standard? – at the end, I stuck to the simple things). I think Flixel’s general design is best if you know what kind of game are you making, and the physics at least resemble something common, like platformer, shot’em up, top-down movement, etc. If you need something else, you CAN program collisions yourself, but this wasn’t clear for me at a first glance – however, this might be due to my idea of learning the framework on-the-fly, during LD event.
  • DAME. Well, I definitely want to use it in the future – it’s just that using it during this LD’s was way too much for me to learn. In the end, I did my level design in paint.
  • Design. I should’ve cut the platformer elements to the minimum, so that finishing the game would take more like 3-5, not 15 minutes. If someone wanted to see all five endings, he would need to waste way too much time, going again and again through the same story. Yep, didn’t think about that. The worst part is that the level is completely static – so if someone wants to try bringing more fish to the human, or find the special item, he always needs to go up the platforms, using the same path. This is much more boring when trying to see other endings.
  • No balancing. That’s mostly why it’s so unnecessarily long. I should know better – I played my game a few times. Now I’ve got some ideas how to make it better: have randomized parts of the story, change level design every time a fish is brought to the human, or at least make time between events much shorter. Also, replaying the game could be much easier, as everyone who will play it more than once, do it only to see other endings.
  • I’m not even gonna write about eating, drinking too much coffee and bad time management. It was bad.

Things to consider.

  • Are you doing art/ story game? Make it SHORT. As in “5 minutes short”. If you have more than one ending, or some additional things for player to uncover, it can be even shorter, because it can easily be played many times.
  • Is ANY KIND of gameplay good for your idea? Do you need that platformer-style gameplay, with enemies, bullets and things to do, if you’re creating a STORY game, or maybe “some kind of interaction” would be enough? Don’t overdo things!
  • Is learning stuff during LD event good for you? If you love it, like me – here’s a suggestion. Pick the stuff you want to learn – do some research before LD and check if the amount of stuff to learn suits you. It’s BAD to waste TOO MUCH time on learning, when you’re supposed to create a game.
  • Very low-res pixel-art is GOOD. Easier to create, but still recognizable. Manipulating color from inside the code can give great effects. Learn some simple graphical tricks to make your game even more appealing!
  • If you’re programming in AS3, learn to incorporate TweenLite/TweenMax into your framework. It’s GOOD.
  • MUSIC changes the mood of your game astonishingly. Use anything you can use – be it Wolfram Tones or any other generator, or whatever. For instance, I’m quite sure that without that ambient music, my game would be much, much worse.

Okay, that’s all folks. Big thanks to everyone who motivated me during my development, I’ve had a GREAT time and YES, I WILL PARTICIPATE AGAIN, and again, and again… forever! Last time was Flashpunk, this time – Flixel, so next time… don’t know yet!


2 Responses to “Someone To Love – post mortem”

  1. Aaron says:

    Hey, marudziik, I just played your entry. Really cool stuff for somebody who had trouble with the theme and was learning new tools on the fly. 😀 Nice writeup–esp the bit on making art games short, and I’m totally going to check out Wolfram Tones. I’ve been stumbling across it several times over the year, forgetting about it, and your game convinced me that I should definitely remember it. 😉

    • marudziik says:

      Gosh, thanks for the warm words! I totally recommend any kind of music generator – recently I also found out about GreaseMonkey’s script, but didn’t have the time to check it out. Judging from some games I managed to play, which were using it, it must be quite good. Here’s more info:

      But yeah, Wolfram Tones & generated music was “the great thing that happened to me” during this LD. I’ve found out about it last time, in August, and figured I had to use it this time. I mean, being mostly a programmer, I have enough problems with creating graphics, let alone music. Having neat ambient tunes in my game is really awesome. :)

Leave a Reply

You must be logged in to post a comment.

[cache: storing page]