site banner
ArjinFerman  Tinfoil Gigachad  3d ago (text post)   394 thread views

Tinker Tuesday for October 13th, 2026

This thread is for anyone working on personal projects to share their progress, and hold themselves somewhat accountable to a group of peers.

Post your project, your progress from last week, and what you hope to accomplish this week.

If you want to be pinged with a reminder asking about your project, let me know, and I'll harass you each week until you cancel the service.

1
Top Bottom Old Controversial Comments
Jump in the discussion.

No email address required.

I decided to vibecode an app that I've been wanting to exist for a while now. It's a tool that lets you look through map files for the game StarCraft Broodwar. The game currently has an issue where if you have too many map files you lag out when trying to join a game. Blizzard made an update several years ago that introduced this issue and they haven't bothered to fix it ever since. The solution is to move the maps into a separate folder but you used to be only able to browse maps from inside the game client, so this created an annoying loop where you have to know what map you want to move it, but if you want to browse maps because you don't remember the map name or you're just browsing looking for something that catches your eye, you have to move the maps back. This is very annoying when you have tens of thousands of maps like me.

I ended up spending about 12 hours on it as I kept adding more and more features but it required 0 coding from my end. I'm still adding features and I don't think I've spent so much time on a single task in a long time. I quit learning software engineering back in 2017 because I didn't find programming to be fun and haven't built anything since then until now.

Turns out vibe coding is way easier than expected and the result is very impressive. All this for just $20 a month with Claude. A software engineer I know said after seeing my app if you were to pay someone to build this app years ago they would likely charge $10,000 for it. So I'm getting something that would've cost me $10,000 a few years ago for only $20 a month and some hours of my time instead.

I'm still working on some quality of life features (currently you have to manually extract specific files from the game to render the images from StarCraft, want to change it so the app can detect if you have StarCraft installed already and then it will automatically extract the files) but I'm pretty happy with the result so far.

Next I want to work on building out a tool that can translate StarCraft maps between languages more easily. If you didn't know, StarCraft is immensely popular in South Korea so there are many custom maps that are entirely in Korean and thus very difficult to play for people that don't know Korean. I used to translate StarCraft maps but it was a slow and tedious task. It should be fairly straightforward and faster with AI.

I also want to build out my own overlay tool for auto translating foreign games as you play. I know tools like this exist already but I want to try vibe coding it myself.

If you're a late adopter like me, I highly recommend trying one of the paid subscriptions and have it build an app for you. I don't know why I didn't start earlier. I am still learning what is the best workflow for using AI to vibe code and will try a different approach for my next project but yeah, pretty exciting to actually create something for the first time in a long time even if it's just getting AI to build the features for you.

120 t-shirts, 80 hats, 100 keychains, 250 marked glass mugs, and 50 electro-etched brass hat pins, and a town festival that went well enough. Sold enough to be happy with it, didn't sell enough to be out of anything, stick a fork in it... except now the brewery wants laser-etched club member cards.

Working on a limited debug mode (and precursor for a DAP) for an RP2350+ESP32 combo. I'm not quite to the point where I'm comfortable just pointing the slop cannons at this project, so it's mostly been a matter of using the LLM for sanity-checks of the implementation, but there's been a few points where it's tempting. Still, it's a toy version of the real problem, and if I can't write this code, it's not going to be practical for a student to expand it.

I made a video showing my minecraft mcp in voice mode: https://rumble.com/v7guxr6-minecraft-mcp-in-voice-mode.html

It's kind of fun. The topographic maps are new.

Trying to vibe code an eastern cultivation mod for the beaver colony management game timberborn. Why? partly shits and giggles. But also see where my former job of programming has gone. I've only had one 3 hour session so far. Its been very informative.

Things I discovered:

  1. Humans are lazy programmers, that is their worst quality. (knew this already, its why I started using AI)
  2. AIs are the opposite of lazy programmers. That is their best quality.
  3. Humans are lazy programmers, that is their best quality.
  4. AIs are the opposite of lazy programmers, that is their worst quality.

I had a plan of battle going in, hashed out with free version of chatgpt. Start with an initial proof of concept mod to see how well AI can mod and change the game. I was gonna stick to mechanics that are already in existing mods. Chatgpt helped me setup a local programming space that would use gemini as a local coding agent.

Was all going well, but then we sent gemini to start on the initial coding task. And it kept looking at a bunch of files in the game library. Running a bunch of read commands and asking for permission to do so. I checked in with chatgpt 'hey whats this thing doing why is it reading all these files' chatgpt explains this is entirely harmless and its just trying to read the files to understand what it can do modding wise. "wait" i think, there are examples of how to create basic mods online. I talk to chatgpt about using those, chatgpt realizes oops ya its not really using those as an example its trying to basically reverse engineer the whole game. We stop gemini from going further down the rabbit hole, redirect it to a more focused task, it then runs out of tokens. Fuck. 6d wait for more tokens. double fuck.

At this point I switch to being computer -> human -> AI middle man for chatgpt. I am the agent part of a coding agent. but I'm exercising more caution at this point because it just wasted a bunch of time trying to decipher the whole game when it could have just looked up the simple modding guide. I stop chatgpt from overthinking or overanalyzing things a few time "yo have your tried using the internet to see if that resource exists" gets met repeatedly with "oh yes! great idea! i found that exact resource I was trying to reverse engineer to be easily available online."

Anyways I ended up with a basic mod at the end of the night. It runs, the game doesn't crash. Minor first success i suppose.

Yesterday I began sending out query letters again. I know I'm setting myself up for disappointment and frustration, but I have to be able to say I gave it my best shot.

If you haven't, you might look at Ark Press, though it does have the same parent as Passage (Lomez / Jonathan Keeperman).

Thanks for the tip, it seems they're not accepting submissions at the moment.

I'm working on a highly vibe-coded project that I've wanted for a while.

I'm an active GM for the D&D spinoff Pathfinder 2nd Edition, and one of the selling points is that all the mechanical components are available online for free: every class, monster, feat, item, etc are freely available on a site called Archives of Nethys as soon as the volunteers get to it. This is nice for budget-minded players, but is also great for researching builds and such since you don't need to worry about paywalls gating your ability to check the content.

Unfortunately, I'm a little disappointed with the structure of the site: the pages are bulky and while they have a very nice search system, it still doesn't feel easy to browse. There was a much nicer site, PF2etools, that is much more responsive and easy to navigate and compare. But they stopped updating the site years ago, so it's wildly out of date with the latest books. But the bones are good, the site is entirely open-source, and it just needs properly-formatted JSON to run.

So my project scrapes HTML from Archives of Nethys (moving at a glacial pace as I really don't want to make this a whole discussion with the Nethys volunteers, especially given the wildly AI-hostile moods in geek-related habits these days), parses it to PF2etools-formatted JSON, and merges it into the existing PF2etools data corpus, with the intent that I can bring the codebase up to date with all the books, and potentially rerun it periodically to stay up to date. Since it's a relatively small static site, I'm hoping I can just host it myself, possibly behind some authentication check so I'm not having discussions with the Internet about this, and have an up-to-date copy for myself and my players.

Long-term, I've got some other plans:

  • Integrating errata overlays: every GM has to make some house rules, and I'd love to have them integrated right into the text of the site so anyone can quickly look up "how it works in my game". This would ideally allow me to do bans and homebrew at the same level.
  • Supporting more formats: we don't really need Nethys, the dream would be just pulling directly from a book PDF (which would also let me grab third-party stuff that wouldn't be in Nethys). We could also potentially output in other formats like Foundry-compatible JSON, easily-readable text, or anything else that might be useful.
  • More integrated tagging: PF2etools has a few unique-to-it spell tags like "forced movement" that make a lot of sense to group and tag game pieces with even if they don't have any official game designation. I'd like to be able to standardize this process and add a lot more.
  • Ratings support: being able to load a content creator's evaluation of something like spells and have it mark up and filter spells based on that.
  • Full Scryfall-style query syntax
  • Going really crazy and trying to build a full character builder off of the existing JSON database. The hard part is getting all the options encoded properly and I've done that, so could I just build something to compete with Pathbuilder?

Currently, I've got a full crawl of Nethys that's translating into PF2etools-compatible JSON and merging properly into the PF2etools corpus, but I keep running into minor fixes when investigating that make me roll back the whole merge: the latest two are that it was treating heritages for any ancestry as their own ancestry, and that Nethys randomly decided to treat individual chapters of some of the latest books as their own sources (a change that is making me more and more confident that making my own database with blackjack and hookers is the right choice).

It's getting very frustrating to try to check these things. I'm floating the idea of trying to get ChatGPT to build some kind of review queue for me where it points out things that might look unusual or wonky and I can see if it looks the way I'd expect it to or if it might point to some error in the parsing progress.

I added the overlay HUD for health and ammo (along with an actual crosshair), it does feel a lot easier to keep an eye on them now, so mission accomplished. I also almost finished the power up system along with spawning. Two issues I have with it right now is that the power-up's beacon lights are too short to be useful, and there seems to be something off about the random number generator, as the frequency of the same type power-up spawning a few times in row seems to be way above random chance.

I'll probably spend this week just tweaking these details, maybe I'll look into limiting the size of the map as well.

How have you been doing @Southkraut?

Wired together my visual perception logic and...ended up stumbling over myself because my design for it wasn't thorough enough, and by attempting to simply wing it on the details I ended up making a hot mess. I need to take a step back and draw up a more rigorous design with fewer blank spots.

I added the overlay HUD

It works, but badly occludes the space around the player character.

It works, but badly occludes the space around the player character.

Is that a big deal? By the time they get this close you'll be scoring hits by shooting blindly... I suppose I can turn it into a horizontal bar, which should be as legible and cover less area. I can even make it configurable for the player.

Can't speak for others, but it bothered me. Yes it doesn't hinder the player much, but I'd find it mildly annoying to not have a clear view of the enemies that are closest to taking a bite out of me.

What is this place?

This website is a place for people who want to move past shady thinking and test their ideas in a court of people who don't all share the same biases. Our goal is to optimize for light, not heat; this is a group effort, and all commentators are asked to do their part.

The weekly Culture War threads host the most controversial topics and are the most visible aspect of The Motte. However, many other topics are appropriate here. We encourage people to post anything related to science, politics, or philosophy; if in doubt, post!

Check out The Vault for an archive of old quality posts. You are encouraged to crosspost these elsewhere.


Why are you called The Motte?

A motte is a stone keep on a raised earthwork common in early medieval fortifications. More pertinently, it's an element in a rhetorical move called a "Motte-and-Bailey", originally identified by philosopher Nicholas Shackel. It describes the tendency in discourse for people to move from a controversial but high value claim to a defensible but less exciting one upon any resistance to the former. He likens this to the medieval fortification, where a desirable land (the bailey) is abandoned when in danger for the more easily defended motte. In Shackel's words, "The Motte represents the defensible but undesired propositions to which one retreats when hard pressed."

On The Motte, always attempt to remain inside your defensible territory, even if you are not being pressed.


New post guidelines

If you're posting something that isn't related to the culture war, we encourage you to post a thread for it. A submission statement is highly appreciated, but isn't necessary for text posts or links to largely-text posts such as blogs or news articles; if we're unsure of the value of your post, we might remove it until you add a submission statement. A submission statement is required for non-text sources (videos, podcasts, images).

Culture war posts go in the culture war thread; all links must either include a submission statement or significant commentary. Bare links without those will be removed.

If in doubt, please post it!


Rules


Recommended Posts And Communities

Recommended Realtime Chats