Morning report - why development gets slower the longer it goes on
Have you ever wondered how games can be shown in development, seemingly nearly complete, and yet the game doesn't get released for ages? It's pretty common, with the worst games being what we term 'in development hell' and not releasing for years and years. The problem (one of them, anyway!) stems from building on top of old code, and 'unwieldy' code, and the longer development goes on for, the worse this problem gets.
When you first start a new project, a new computer game, you have an idea, an exciting spark of inspiration. You sit down and over a couple of days of solid development, you have a working prototype that plays well and is full of potential. At this rate, it'll be finished in four months! But it never pans out that way. As you start adding content and new ideas, you have to connect new stuff with the old stuff and old code you already have. Quite often this can't be done in an ideal way because the old code wasn't designed to support the new ideas, so you use a bit of fudge. It works, it's fine, it's all good. But as you add more stuff, you have to connect the new fudge to the old fudge, and it starts to get a bit messy. Finding where to add the new idea can become a bit of a nightmare with 'spaghetti code'.
ionAXXIA is progressing and the next iteration is out ever so soon, still. Development is frustrating at the moment because I'm working through old code. ionAXXIA actually started life as ProteanWars some three years ago, where you could design your ship and its weapons in their minutiae. Every component had its own parameters and every parameter was exposed for adjustment. It was about a year in development when I shelved it as just too complicated. I picked it back up a little over a year ago after thinking I could easily turn it into a simple arena shooter over Christmas and just get it out there.
At this point, there was a new Unity Asset providing a proper database for my ship components. So I added that, and fudged a quick connection to it as I was just making a simple arena shooter. Then I spent months perfecting console-like controls on touch screen to make it play well. Then I added simpler ship designing. At the end of the day, I guess didn't really want to make just a simple arena shooter. ;)
What I've found this past week is that a few bugs and issues that are easily solved are hard to trace. I have systems connecting with systems connecting with systems and a fair bit of legacy code that isn't ideal. As a dev you are faced with two options - 1) Rewrite significant bits of code so it's better and easier to maintain going forwards, at notable delay. 2) Carry on with things as they are because you're nearly done. Both suck!
