I’ve been back in Phoenix since Saturday afternoon, after a relatively uneventful series of flights that included another automated landing at a fogged-in Charles de Gaulle.
I haven’t posted much in the past several weeks because, for the most part, I’ve been keeping myself busy with the blog software rewrite project I mentioned last month.
Initially, the plan was to rewrite the software using a development framework that more closely resembled what I use in my day job. The thought was that I could enjoy a spillover effect in both directions, with the research I did for one project improving the other.
The reality, though, was that after about six weeks, the spillover was only going in one direction. The framework I typically use for business application projects works well because business rules are rigid — or at least they should be. The internet, on the other hand, is messy. The rewrite project was stalling because I was trying to parse incoming messages into rigid data structures that couldn’t reflect their messy nature. By contrast, the framework of the current software is better suited to handling loosely defined data structures idiomatically, which was part of the reason I chose it in the first place.
Thanksgiving morning, I made the positive decision to stop trying to move forward along a blocked path, take a step back, and start a brand-new rewrite. This time, I’m using a more recent version of the current framework, rebuilding the foundation from scratch, and bringing over what was salvageable from the current software.
It was the kick in the pants I needed. Already, I’m where I was with my first attempt in about a third of the time. However, I’m still far from done. The biggest feature I want for myself is a check-in process that integrates cleanly with OpenStreetMap data. The biggest feature requested by readers is the ability to react to or to comment on posts. When those two items are ready, I’ll start migrating. I’m not even going to try to predict a date.
