August is loosely about video games, but really itās about three big things in particular.
-
book: Wrangled LaTeX some more. Came up with a new style for admonitions (little set-out boxes) that I really like. Drew some icons for a few of them. Started on another chapter, for reasons; see below.
-
veekun: Regexing machine code for addresses was getting really clumsy, so I went one step further and wrote a disassembling pattern matcher thing. You write some assembly with some variables in it, and it finds occurrences of that code and tells you what the variables are. I can pretty much paste in entire functions, massage them slightly, and find matches. Itās pretty slick.
The upshot of this is that loading original Japanese Red and Green now works! But Yellow doesnāt. So I fixed that, and now Japanese Blue is broken. Or maybe I fixed it and that broke Yellow again? Iām not sure. There were some tiny changes to core code between some of these games, and the pattern-matcher has no way to express alternatives. I donāt know if Iām better off inventing one or just fudging it.
Anyway, pretty close to having all of gen 1 dumping PokĆ©mon reliably. Still need to actually dump other stuff ā moves, items, encounters, and the like ā but thatās much more straightforward.
-
hax: I was still in a mood to dink around with Game Boy stuff, so I added Python 3 support to some relevant tooling and wrote a proof of concept for storing PokƩmon maps in Tiled format.
-
blog: I wrote a thing about writing tests.
-
twitter: I taught @perlin_noise a few new tricks.
-
art: I drew a friendās lizard pal based on a reference photo, which isnāt something Iād seriously tried before. Value-only, only one layer, only one brush. It came out surprisingly well.
-
gamedev: I participated in Ludum Dare 36, a 48-hour game jam. Iād never done LD before, and naturally I picked the only one that has no ratings round (for administrative shuffling reasons). Oh, well.
The result was Isaacās Descent, a short puzzle-platformer for the PICO-8. You can play it via the web (source code included), and I also wrote a post about it.
So! There are a few days left, but itās pretty much the end of August. Letās see how I did.
-
Draft three chapters of this book, August: one chapter
Well, I didnāt get a chapter done. I did make huge progress on the chapter I started, though ā plus I began a second chapter, and generated enough notes for the entirety of a third. I spent a decent amount of time wrangling Sphinx and LaTeX, too, which I wouldāve had to do sooner or later regardless.
So I didnāt do quite what I wanted, but I did do far more than Iāve put into any previous harebrained book idea, and it was a pretty decent chunk of work. Iām okay with that.
Just what is this damn book, you ask? Ah, perhaps you should read that Ludum Dare post.
-
Get veekun beta-worthy, August: basics of the new schema committed; basics of gen 1 and gen 6 games dumped; skeleton cli and site
Haha, no. I got gen 1 almost working for PokƩmon only. It turns out that while gen 1 has the simplest data, it probably has the most convoluted storage.
On the other hand, the detours taught me a lot about Game Boy architecture, which was interesting and helpful for making the dumper fairly robust thusfar. I also made some breakthroughs on architecture that had been haunting me for a while. Iāll have to move my ass in the next week or two to catch up ā hopefully finish gen 1 and get a few other generations dumped real soon ā but I think this is still doable.
-
Finish Runed Awakening, August: working ending; at least one solution to each puzzle; private beta
Whoops! I did basically squat on Runed Awakening. I figured out most of the ending, which had been my major roadblock, but I didnāt touch the code or run the game a single time. Dang. Itās not like I was goofing off all month, either; I just didnāt have a big block of time to devote to the weird mishmash of writing and planning and programming that IF requires.
I really want to finish this game, but end of October is not looking too great. I donāt know why itās proving so difficult; itās not that complicated, and I started on it almost two years ago now. Iāve made multiple other games just so far this year! Argh.
If itās any consolation (to me): I picked November as a target because Mel wanted to embed Runed Awakening on Floraverse as an update around that time. But Isaacās Descent takes place in the same universe, so it works just as well. Goal accomplished!
Onwards to September. The only thing on the list with a real solid deadline is veekun, since the new games will be coming out. Itās a bit behind, but Iām pretty sure I can catch up. Gen 2 shouldnāt be too different from gen 1, and Iāve done gen 4 and onwards before.