dinomelt:

More than once (and specially on a looong project), as one grows and develops better and more efficient ways to make things (an evolving artstyle, a cleaner way to program, new approaches to world and level design…), it becomes reaallly tempting to just go back and re-do the whole thing.

During the 2+ year development cycle of Dinomelt, this has happened more than once: The protagonist, Gwrep, had a complete sprite-re-drawing halfway through the project. Also, recently, plenty of the main background assets and platfoms got a color, line and overal polish update, so that most aspects of the game had a more consistant line.

As an example, take a look at this gif from version 0.4 (late 2015!)

It was looking and working fine, but the animations on the protagonist, Grwep, weren’t the most convincing, they were stiff, and the ‘run’ was a bit off… Months later I went over its whole group of animations to make them smoother with stronger poses, and a better sense of weight (specially on the runcycle).

---

In that sense, my current goal is to get this release DONE before I get the ‘itch’ of remaking everything again, otherwise, you may expect the game to take another year or even more!!

I can confirm this is a super serious issue. I worked on a huge project back in college called “Recession”, a metroidvania about accumulating wealth while the game’s own economy collapses. The project failed essentially for this reason.

I worked on this for several years, and during a formative time where I improved a lot in both art, programming and design, so I was constantly going back and remaking, refactoring and re-designing.

Had to give up on it at one point, because not only was it getting nowhere close to being finished, but the constant re-working and re-designing of whole areas made the project grow into something less and less cohesive, eventually feeling so different from how it started that it was pretty much unsalvageable! New areas and levels had been haphazardly added onto the core game and it just grew with no rhyme or reason over time, it was a complete mess.

I did learn a lot while working on it, its probably the time where I grew the most as an artist, designer and programmer, but its still a huge bummer that so much of this work was essentially wasted. Devs gotta be aware of this problem and feel it coming before it really sets in, its a particularly high risk for any hobby project where you have no outside influence to give you pressure and a deadline.