In my #AirCraftCarrier #Sim #game you play as the commander of a carrier who's goal is to survive indefinitely. You build cities, develop space programs, and expand island to island, competing economically and militarily against #AI and other #PVP rivals. https://commslink.net/carrier00
Players start with a single carrier and 3 planes. They earn a steady income every 2 minutes to buy more planes or found a city on an empty island. Cities can produce taxes to raise your income. Other players can destroy your city if you do not defend it.
The red cities on the horizon are enemies. Fishing boats are what provide cities with food, but they require lots of room to find fish. Fishing boats competing for fish drive territorial competition in this game. Your city cannot produce more taxes if it cannot feed it's population.
The game is more like a social environment. The leaderboard ranks cities not by how large they are, but rather by how long they have survived in the game. Surviving for a long time, even when you are offline means having allies to defend you from sneak attacks, and a support network. Trade will be an upcoming feature.
My Carrier00 game is inspired by the text-based game https://area00.com
Please help me test it: https://commslink.net/carrier00. I will give testers permanent free premium status someday when that means something
Sounds like Boom Beach.
I trust you'll beef up the models. Get some nice looking low poly jet models, proper looking aircraft carriers, trees, houses and buildings. Water simulations we can usually cheap out on, but if the sea moved that would be more immersive.
But I know, examples and game mechanics. you're cooking, i know.
Sounds like Boom Beach.
I trust you'll beef up the models. Get some nice looking low poly jet models, proper looking aircraft carriers, trees, houses and buildings. Water simulations we can usually cheap out on, but if the sea moved that would be more immersive.
But I know, examples and game mechanics. you're cooking, i know.
The game is a MMO that runs in a web browser. I cannot go heavy on graphics or it will blow up your CPU
By going cheap on the graphics, I am able to make the planet a true globe with orbital mechanics like Kerble Space Program so the world is immense
try it free: https://commslink.net/carrier
last edit on
6/25/2026 8:28:57 PM
javascript graphics libraries are rough, probably between the node.js stack on both the dev and browser ends + the lack of low-level system calls. the isometric stuff like Command and Conquer with limited field-of-view works pretty well, open world is tough
i imagine you've been enjoying AI tony, have you used the MCP connector for blender + cluade? i've also been seeing trading harnesses for models trending on github, but i've been too swamped lately. also i've heard a fair amount of people saying that using models for trading is silly. but then what's a quant doing that an ai cannot
javascript graphics libraries are rough, probably between the node.js stack on both the dev and browser ends + the lack of low-level system calls. the isometric stuff like Command and Conquer with limited field-of-view works pretty well, open world is tough
i imagine you've been enjoying AI tony, have you used the MCP connector for blender + cluade? i've also been seeing trading harnesses for models trending on github, but i've been too swamped lately. also i've heard a fair amount of people saying that using models for trading is silly. but then what's a quant doing that an ai cannot
This might surprise you, but I actually don't know how to use Blender very well. I did help a friend with Claude's MCP connector for Blender and he wasn't satisfied with the output he was getting with Claude.
It was annoying cause he was complaining about how the CG looks like shit, and I was like dude, CG will look like shit before it's finished. I used Gemini to write a prompt and sent it to him.
I do have Blender but only use it to.... Actually I don't need it for converting frames to mpeg4 anymore.
I don't see much of a reason to use 3D anymore, or Ai and 3D while generative Ai is superior. Though theres a method to use 3D for camera movement accuracy for generative Ai, it's kinda niche.
As for trading with Ai.
Ai will have a "can do" attitude when asked to do this but won't address the downsides or challenges unless asked. Not Claude. It will straight up address timing or speed concerns. To an Ai, being fractions of a second late to move on a trade is disastrous. This is because whatever it trades for whatever reason, there'll be a million other bots playing on the same item, impacting the price. For example if your bot tried to snipe an asset worth a penny, it'll try to perform the trade but the value already increased due to the faster bots, the trade then fails.
It helps to instruct the bot to pay a little extra which pushes the price up and highest bid get's priority, that's whats happening with price action. At times a bot may push a price up 40% then sell when a satisfactory price comes in. Some bots share and trade with other participants so it'll see a lot of price action, it'll sell first along with participants crashing a value faster than other trading bots can react.
On Solana ( best and fastest chain for botting I think, cheap gas fees) Jupiter is good for fast signals that will wake up the bot to assess the situation and move.
Even if using an agent on a harness like Openclaw or Hermes it's recommended to use a brain running off a VPS. It'll be faster.
Still there's 1 more problem, the bots that are co-local with the exchange will always be fastest. How to get a bot running on the same server as an exchange, I don't know. It'll be highly unlikely to have a co-local server.
Set the bot to use MACD and RSI indicators and let it take profits on what might be a fake out. Also need to add specifics like the bot staying up to confirm and renegotiate the trade before awaiting the next signal, along with a list of other attributes.
Currently a bear market so making a 2× 100% gain isn't easy, so once the bot is tried and tested successfull, you'll want to give it more money. While it doesn't matter how much money we use, $500 minimal is usually recommended for this to be worthwhile in the current state of the market.
Last bullrun was 2024, Next bullrun will be in 2028. Already halfway there. Scarry how fast 4 years goes by.
didn't you do your 3d modelling work in blender?
you said you don't see a use for 3d rn, but maybe people vibe-coding video games will need 3d assets
the trading stuff here is interesting
i am glad you have these ideas about server latency. i actually don't want to post too many ideas here. i'm tryp4 on discord. i have no idea about the trading end, but i know how to deploy servers, host agents (where they need to be fast), run cron jobs, run kanban with hermes and make a book when a boomer interacts with your agent on facebook and it has hermes + stripe backend to pay for it all. add me bro, maybe we can see something