Dev Notes 7/31/2017

tl;dr: Using Construct 3 to speed up prototyping and development work.

It's been about seven weeks since my last update, so here's a bit of info about what's going on...

  • User testing different ideas has made me really rethink the whole development process.  Basically, I can't go for 6-8 weeks between having a game idea and testing to validate if the idea works.  I'm using Construct 3 for building stuff out right now.  It's very, very fast.  At some point I may write a developer's guide for using Construct 3 for other folks that are making a similar transition.
  • I built and threw out a prototype for a pixel-art "lite" mobile RPG.  It seems great on paper, and folks actually liked the pixel art quite a bit, but I couldn't seem to "find the fun."
  • I built a prototype of a space shooter, something simple that uses a standard Xbox game controller, and in 30 minutes I had something more fun than the pixel art RPG.  ¯\_(ツ)_/¯
  • I have a very, very bad case of RPG-itus.  Most of my concepts rapidly seem to go from "here's a small fun idea!" to "let's turn it into an RPG" waaay too fast.  When I do the math on even a simple RPG, it's going to be many months before I have anything worth playing. That's probably fine, at least in the sense that I've always been interested in more mid-core titles than micro-gameplay, but it also makes things feel like they are taking forever.
  • Looking at my notes from seven weeks ago, I have built a decent very minimal playable, and also have pulled together some decent design doc material.  That's the good news.
  • The bad news is that it all just feels like it's taking too long, and I have too many distractions right now.  The weather is great right now, my son is home from school, and I am spending (sorely overdue) time on personal health (gotta lose some weight).  It's all good stuff, but dang.  When I get cracking, I'm moving fast, but still.
  • I'm spending a lot of time thinking about how [relatively] little coding there is in game development, especially if you are using a good off-the-shelf game engine.  Game Maker Studio 2, Unreal BluePrints, and Construct 3 all let you build great stuff with virtually no coding.  Sometimes there is stuff that is reasonably technical, but not coding - for example, a tiny bit of simple trigonometry or vector math using built-in functions.  
  • There are times when I think server-side is a lot easier than game development - writing test cases, doing perf work, and scaling out servers lives in the world of facts, whereas "finding the fun" lives in a world of opinion.  You can validate opinions with things like user testing, but it's very slow and time-consuming.  It also only answers questions in retrospect (effectively, a form of regression testing).
  • I think it would be easy (read: 3-6 months) to build some plugins for Tiled and libGDX that would give you a nice drag-and-drop system for quick game play, but I don't think there would be much of an ROI on that investment.
    • Maybe it's more about adding in some kind of system for automatically adding server-side/multiplayer features for games?  Hmm.
  • Instead of coding, I'd say that most of the heavy lifting for [client-side] game development nowadays is around the design and art assets.  There are some interesting fringe cases (like shaders, or very unique gameplay), but the game engines do so much...
  • Most game development tools have no idea what to do about version control/collaborative development.  :P


Will Iverson