As is traditional now, I’m taking the last part of the year to collect all projects I completed and pick out any themes or interesting results from the year.

Long-term, I think the biggest change in 2025 was taking the Platform Guides seriously. I took a break from posting to do them, and they were much more work than the posts themselves, but I’m quite happy with how they came out and word does occasionally reach me that they helped a would-be retrocoder get off the ground. So that’s gratifying.

Also gratifying is that quite a few of my articles this year were widely shared and popular. Most of the stuff I write here is primarily valuable as reference material, and so most years my top 10 posts are usually from previous years. Not this time: the top 4 articles this year were all from this year, and they did so well that in a single year they also found themselves within my top 10 articles of all time. The flavor of these “instant hits” is mostly different from the others (they tend more towards “grand tour” than “tutorial” or “reference”), and since my focus still hasn’t changed here, I think these are providing a rather different sort of service. At some point early next year I should probably include some kind of formal welcome page describing what this blog is all about, since “curious bystander” is now a meaningfully-sized group of readership.

Themes and Projects

At the end of last year, I had a list of unfinished business from 2024 that carried forward into 2025, along with some new things on my list to look at. How did those turn out?

• I’d just started revamping the Platform Guides in late 2024 and was unhappy with my pace with them. As I mentioned above, they’re now in a place I’m basically happy with. Now I just have to keep the indices updated when I do new things.
• I wanted to experiment with a bunch of new devtools, and in particular wanted to update my Z80 and Genesis toolchains. This ended up being a sizable part of the Platform Guides hiatus, and it bore good fruit.
• I wanted to revisit the Cyclic Cellular Automaton on the Genesis to optimize it and port back the cool stuff I did in the SNES port. This happened, and it ended up beating the SNES in the end.
• I had a couple modern-coding projects that were brewing and neither really came to fruition. I haven’t ruled them out, but they’re both firmly back-burner at this point. I did get some decent modern projects out this year anyway, though, so I’m not unhappy about any of that.
• My list of retro systems to investigate was pretty short: the PCjr, the TI-99/4A, and the MSX. I didn’t get the PCjr but the other two got their share of the spotlight.

Looking over my articles, there are three major themes that stand out:

• The most overwhelmingly pervasive theme this year was the TMS9918A graphics chip and its legacy. Starting with the TI-99/4A, and moving through into the SG-1000, the ColecoVision, the MSX1, and the Sega Master System, these systems and my experiments with them formed the spine of my articles this year. Even the Genesis sort of belongs in this lineage, if you squint.
• A secondary theme was the evolution of how we’d accomplish a task through time. I opened the year with a series experimenting with various ways of implementing linked lists in the face of changing CPU architectures, and near the end I had a series on OpenGL rendering as it evolved over its lifespan.
• Finally, there was a significant amount of cleanup and parity work. The Platform Guides required me to have a collection of toolchains worth recommending along with sample code and build instructions, and that meant that I had to learn a lot of new tools and put them through their paces. Along the way, this led me to revisit my older code that I’d built, and that showed me just how far behind my earlier versions of Rosetta Stone projects were compared to later ones. They could be better, and I improved a good many of them.

The Project Compilation

The 2025 Bumbershoot Software Compilation offers pre-built copies of all the full-scale programs I produced over the year. There were a good chunk this year:

• Four new ports of my Shooting Gallery Rosetta Stone, for the TI-99/4A, SG-1000, ColecoVision, and MSX. These all are powered by the same graphics chip and as a result are all extremely similar in appearance.
• Three new ports of the Lights-Out Rosetta Stone program: updates to the ZX Spectrum and C64 versions, and a new one for PICO-8.
• An updated port of the Cyclic Cellular Automaton for the Genesis.
• Three extra cartridges for the TI-99/4A: two basic Hello World programs, and a PRNG test program.
• A port of the “Sprites and Controllers” Test Program from the Sega Genesis to the Master System.
• A bunch of other small graphical test programs for the Master System, of which the most elaborate involves text with drop shadows over an animated rainbow background.

Download the full compilation here.

Looking Forward

My current retrocoding TODO list is pretty modest, and closing the gaps that are left after the end of this year is likely going to be a matter of various one-shot articles instead of huge projects:

• The PCjr and MSX2 are still basically untouched, but we’ve done quite a bit with closely related systems (DOS and MSX1).
• There are some aspects of TI-99/4A software development that I looked at only cursorily.
• I could push some of my Lights-Out ports even harder. I have some ideas for pushing the TMS9918 and VIC-II chips even harder than I have, and this is a good place to do so.
• I still haven’t personally replicated every technique described in the canonical VIC-II article.
• I have some toolchain testing left to do.

My list of things to do with modern code has been growing, too, but a lot of that is still tentative enough that I’m not willing to lay down any markers just yet. We’ll have to see what 2026 brings! Here’s to a new year!