I have this legacy WordPress theme (Arras). Over the years I’ve made several attempts at refactoring it. My first attempts were back in 2013. Those early attempts fixed a few of the pressing issues, but were, I confess, less than satisfactory. The code-base was still pretty messy, and my changes broke as many things as it fixed. It happened that the things that broke weren’t parts of the theme I myself was using, so the new mess was, at the time, more acceptable than the old mess.
Eventually, the mess became too much and I consigned the project to “no longer actively maintained” status. (I’m pretty sure the difficulties were what led the original developer to abandon it as well.)
This week I spent a bit of time with it again, coming back to it after a long hiatus with fresh eyes and a couple more years of experience. I started in on the project from where I’d left off before and made a bit of headway. The experiments this week have given me some hope that the mess can be cleaned up.
That said, I’m going to start back again on the project from where I first inherited it. With any luck, the refactoring from the beginning can become a kind of learning tool and the theme can be restored, bit by bit, to a useful state. Starting again from before my previous refactoring will also make it easier, this time around, to retain the functionality it had at one time that was lost in my previous attempts.