Phew, another awesome Ludum Dare has ended, I slept for a bit, and means that I can finally catch my breath & write a few words about “what went right/wrong”. If you haven’t played my entry – What The Sock – yet, I strongly recommend doing that since I’ll be referring to certain parts of the game throughout the post (just click on the images, these are also a tutorial for the game):

What went right:

• Working in a semi-familiar environment: this one’s a no brainer, but having ~8 months of prior experience with Unity and several years of practice in Photoshop helped me immensely. It was the first time I’ve played around with Unity2D though, so while I encountered some issues along the way (especially ones related to the 2D physics) I also learned some stuff – which is always good.
• Unity editor: seriously, it makes a lot of tweaking around much easier.
• Simple idea for the game: even though it took me some time before I got to the prototyping/tweaking stage, the concept was simple enough that I was able to deliver most of the planned content. I had to cut one mechanic and a bunch of assets, but the final result is satisfying, IMO.
• The soundtrack took me 15 minutes to create and implement: I kid you not, Soundation Studio is really great for that. You can’t make A-class music with it, but it’s a good-enough tool for compos.
• Unity animation curves were much better than expected: I used them for my intro/outro and a couple of other things

What went wrong:

• Unity2D is not yet ready for primetime: while great folks at Unity have delivered a bunch of great 2D tools that make creating two-dimensional games waaaay easier than before, they’ve also failed to implement some of the more no-brainer features found in other software. For example, there’s no easy way of doing tiled graphics or infinitely stretching 2D backgrounds in Unity – the easiest way to do that is applying a Repeat Texture on a 3D Quad, but this comes with a bunch of other problems. Another thing is the aforementioned 2D physics implementation – it’s not easy to work with, and a couple of things (like OnTriggerExit2D, which would’ve saved my ass a couple of times) doesn’t work at all.
• Too much time lost on fixing assets: alright, this is entirely my fault. I was semi-obsessed with delivering good animation (especially the walk cycle, which looks kinda crappy) and spent a lot of time on fixing that – when instead I could’ve created some new assets like different background buildings, or, I don’t know, hydrants to stand behind.
• Too much time lost on coming up with a game idea: Ludum Dare started at 3AM (CET), I woke up at 7AM, but it was 2PM before I really started working on the game. I had a couple of ideas, but it took me a lot of time before I had an “Aha!” moment. I need to somehow make this process fasterin the upcoming LD’s
• Unity Mecanim state machines are not so okay: seriously, it would be easier to switch animations from code if it wasn’t bugging me about having to change animation types to “Legacy”.

TL;DR:

• simple jam game = (quite possibly) better jam game
• don’t learn your tools during the compo! (although I did that with Greyspace and it turned out okay)
• make sure to include music in your game!
• gif-based tutorials are the best thing

Stuff to do before the next LD:

• learning how to do better pixel art and animations
• learning to “how do I music”
• ???
• finding peoples for a jam (maybe?)

Tags: LD28, postmortem, spierek, what the sock