500 commits

posted by Jeff | Tuesday, April 30, 2019, 9:25 PM | comments: 0

Today was a quasi-historic day, when I made my 500th commit to POP Forums. Sort of. Actually, it's just 500 commits since March of 2013. The history prior to that is not clear. I think (and my friends who worked on CodePlex might remember) that prior to then I was using Mercurial for source control on CodePlex, and there was some point where they were able to convert it to Git. While I had been publishing zip files of the source to my own site for years with every release, I didn't really start developing it in the open until I moved it to CodePlex in the summer of 2010. When I learned it would be shutting down, I moved the Git repo to GitHub in the summer of 2014.

I mentioned at the beginning of the year that I was getting back into working on this project, and I'm pleased to say that I've generally stuck with it. I've been streaky, but committed to at least make one commit a week. Things really picked up when I decided I would commit to having a non-trivial understanding of a new thing, specifically Vue.js, and that's going really well. It's the most fun I've had with a new framework in, well, I've never had fun learning a front end framework, so yay. I haven't adopted all of the best practices around the "right" way to deliver it, but it's fun getting in the weeds. I have a feeling I may try at some point this year to dive into various federated authentication mechanisms, which would be helpful for work, or a liability to the people I'm trusting to work on that stuff (#managerproblems).

I don't have a roadmap for the app, but I imagine there's some approach I'll take to sprinkle some front end, componentized goodness into it, in a way that doesn't break the Google juice that it benefits from. Seriously, I could never add another topic on CoasterBuzz and PointBuzz and I'd get organic traffic for decades because it's so well indexed, so I don't want to break that. But there's so much room to clean up all of the scroll loading, templating, login process, etc.

If I stick with it, I think I could definitely add another few hundred commits this year. I do plan to ship a release when I'm done rebuilding the admin side (that's the front end project), even though there's still a lot of mess and need to refactor. There are still remnants of bad decisions made 15 years ago in there, but those bad decisions are largely an issue of organization and testability, not (generally) one of performance.


Post your comment: