Hugzilla – Postmortem and Reflecting Back

Posted by (twitter: @joeld42)
August 27th, 2015 12:32 pm

This was an interesting LDJAM for me. In a way it brought things full circle back to my first LD, LD1. That was what, over a decade ago?

The first ludumdare game I made was called “zillagame”. The theme was “Guardian” and the game involved protecting a small, seaside community from an attacking godzilla.

Screen Shot 2015-08-26 at 8.43.47 AM

This time, the theme was “You are the monster” and so I decided to revisit that idea, but from the perspective of the Godzilla. My game is called “Hugzilla”. The hugging angle only came up after I modeled the zilla with outstretched arms and noticed it looked like he was trying to hug the buildings. The best components in any creative endeavor are alway from these kinds of happy accidents.

Screen Shot 2015-08-23 at 6.17.03 PM

Some differences stand out immediately. I got a little better at 3d modeling, I realized the importance of picking a good color palette, and it’s pretty cool that you can run this in a web browser these days. But pretty much, I made the same game (or incomplete half-game). Thirty-three ludumdares (well, I didn’t join all of them, probably 2/3s) and I’ve got basically a monster that moves around the landscape, buildings that take damage and some nice scenery, with little in the way of actual gameplay.

What does this mean? That I spend too much time on engine features (I remember I spend a lot of time on PNG loading in LD1, this time I spent a while on object mouse picking which went unused). That 3D is untenable without a huge engine like Unity or UE? Maybe that I am just a slow learner?

Not really. I don’t do these game jams to try and win. My huge collection of unfinished prototypes is my trophy shelf.  But most of the time, I’m doing LD to expand my boundaries, either to learn or try a new tool, or to push the limits of whatever tool I’m using. It’s a way of battle-testing a tool or process. I also tend to focus on the graphics rather than the gameplay, which makes sense because that’s my background (tools and graphics programming). I wonder what would happen if I made a game with nothing but cubes. In face, my 2nd ludumdare game was sort of all cubes, but I did get carried away and ended up writing a pseudo-3d level editor and convex polygon tessellator for it. I guess you can’t change who you are, and LD, like any creative expression, is a way to find and reveal that.

Anyways, I should probably do the traditional postmortem lists here:

What went right:

  • Luxe engine, Haxe language. Once again, this turned out to be a very productive tool. 3D support is still in very early days but Luxe and Haxe feel like a very, very good fit for game dev, both for jams and larger projects.
  • Blender Exporting. I spent a little bit of time beforehand getting a bare-bones blender export script working. That was hugely helpful, for the first time I wasn’t manually exporting and importing everything which always takes a lot more time than it should.
  • “Tilemap Texturing” I came up with a very interesting trick to use a uvmap similar to a tilemap in blender. I’ll write up a full blog post about this soon, but this trick alone was worth the entire weekend.
  • “Matcap” materials. I threw in a “matcap” style shader, which lets you render a material sphere and use that to fake a material. It worked great. I didn’t have time to use it for other than giving things a little sheen but for future contests this seems like a good way to make a bunch of materials quickly.


example of matap. That sphere is actually a render out of blender/cycles, but this trick works great with hand-painted material spheres too.

What went wrong-o:

  • Vague game design and mechanics. I only had a vague idea of what the gameplay would be like, I was imagining RTS-style controls at first and so I spend several hours doing 3d picking code and stuff I would need for that. However, I didn’t end up actually needing that. If I would have nailed down my gameplay a bit more on paper first, I would have realized this earlier on.
  • Last-minute shadows: I had some shadow code that I was working on before the compo (as part of my basecode). With twenty minutes left I decided to try and throw that in there. I had barely enough time to get it working and no time to tune it, so there are all kind of weird shadow artifacts, so it looked worse than if there were no shadows at all. I’m honestly a little jealous of the unity folks who can just check a “shadows on” checkbox. Anyways, I’ve tuned these since the compo, so the post compo version will have nice looking shadows, and hopefully I’ll have this to use in the future.
  • Anti-social. I did spend a bit of time on the Snowkit Community chat, but overall I felt a bit isolated. There was just too much volume on the ld site to follow other people’s projects, twitter wasn’t too lively this time around, and I didn’t join the IRC chat this time. Next time I will try to spend more time in IRC and maybe attend a meetup if there is one nearby.

Anyways, I’m probably not going to do much more with this game at least anytime soon. But I’m taking some of the tech and code for this and am going back to revisit an earlier ludumdare game that was one of my favorites (from LD13), maybe that will turn into something tangible.

See you all next time! –Jovoc


Leave a Reply

You must be logged in to post a comment.

[cache: storing page]