Weeks into developing the prototype we realized that it's slowly getting more difficult to change very simple things; and not being able to add more characters without a major revamp was the last straw!

So what happened?

Bad, undocumented, and spaghetti code filled with band-aid solutions happened. I remember making it clear back then our aim was to make a prototype where the mechanics felt good. Fast. So I chucked all the decent programming practices™ I’ve learned and focused on making it work. Needless to say, I overdid it. Here are some of the stuff that made me want to flick the ‘me’ two months ago with a rubber band:

• Core mechanics heavily reliant on GUI code

• Parallax Background loop used poorly made formula based on trial and error on the furthest background that lead to inconsistent instantiate positions

• Too many temporary variables and unplanned scripts made from quick patches/changes on the mechanics when we were deciding on what felt good

• Accidentally imported multiple Firebase packages and couldn’t figure out which ones to delete without ruining the ones that I need

• GUI logic were pretty much enable-disable scripts that had the possibility of overlapping with each other

• And quite a few more not worth mentioning

So it felt really worth it to just start from scratch. I managed to catch up to around 70% of the entire progress in less than a day so it felt pretty fulfilling anyway. Plus, this time it’s much better prepared for having a back-end! But one of my favorite parts is the fact that the background looping feels satisfying now.

Check out the entire story: https://dandangames.fun/?p=248