Startracker 🔺

How to make game UI design and implement from start to finish using Paint (lol), GPT Image 2, and @cursor_ai with Opus/Composer: Prompt in comments 1/n

@snapss @TheGrottoL1 @cavemoneyDEX I'm still debating if it fits my game. Booster packs as nfts would maybe

If you're on @TheGrottoL1 and you: -launched a successful game or -have an idea for a game but don't have funding you should use the crowdfund feature and launch a coin! earn creator rewards & attract capital attention! when tokens graduate, liquidity goes to @cavemoneyDEX !

@0xNeothon @avaxjesus @Wrathtank_avax @TheGrottoL1 @doodlelamahh That's a great idea! Should make that as a feature!

yeah those two @doodlelamahh and @startracker are so good! I can't beat my previous score either and they are in 5 figs. May be next week we can do a competition with prize! I don't know if Grotto supports token usage? That will be so cool. You bet tokens that you can beat a certain score and lose it all if you can't 😂 @snapss @Wrathtank_avax ?

wow, Avalanche Snowboarder is the 6th most trending game on @TheGrottoL1 after Wrath's games ^_^

@gneubig This so much. I always put Opus as the director. GPT sometimes does the job, but mostly against its will, will likely give silly tasks to subagents and largely prefers to get its hands dirty. I do insist that Opus is hired for any visual tasks though.

I tell GPT 5.5, you are a manager, not a coder. Find the issues to solve and delegate to other agents. Do not write any code yourself. It does so for a while. I think "good GPT" and log off, I let it do its long running tasks with its team of subordinates. I log on an hour later and check in. GPT 5.5 is coding alone, its sub agents diligently waiting for orders. No STOP, I say, you are a manager. You MUST NOT code. My bad, says GPT 5.5, got it, I must manage, not code. One hour later, GPT 5.5 is coding. But it's OK GPT, I get you. For I am also guilty. No matter how many times a coder is told they are a manager, in their heart of hearts, they are still a coder. So I tell Claude Opus 4.7...

@LayrKits We need to find a better way. Even the pixelart tiny ones have that film grain inside them. I still haven't managed to solve it. Kling is a dead end for now

I'm thrilled to see such positive responses to my 2D Sprite Sheet Pipeline write up! I'm really excited to see what you build with it! Here are some of the animations I've created with it to motivate you guys! 😁 If you already tried it out, post the results below 👇

@NekoAlliance Looking forward to it!

Next Space falls on Boys Day in Japan 🎏 Talking to @Startracker about vibe coding a tiny MMO that fits on a floppy disk. Also, the NEKO panel with a round up of Web3, AI, & NFT news in and around Japan. x.com/i/spaces/1qxvv… ⏰Mon 4th 7pm PDT 10pm EDT, Tue 5th 11am JST

@benln You hiring crazy obsessed Cursor power users? Fitting into a corporate title mold is difficult

Cursor is hiring for 70+ roles: cursor.com/careers

@0xNeothon @doodlelamahh fkn RNG snowballs. no true test of skill when rng is involved it's addictive tho, love the theme!

Wow @doodlelamahh is really killing it @startracker did you play yet? 😬

Avalanche Snowboarder Update 🏂❄️ • Ice boosts to push your speed • Shield orb for 5s immunity • Combo system with multipliers up to 4x • Dynamic scenery with dusk, dawn, and blizzards More speed. More risk. Higher scores. Less snowballs in the Avalanche 😅 enterthegrotto.xyz/games/avalanch…

@LayrKits image models are surprisingly good for specials and stuff that clearly stands out. Ordinary movement is almost impossible. x.com/bigheeno/statu…

Last one seems the most plausible right now. I have had success with generating sprite sheets with image models. They still had to undergo serious cleaning and lineup but it worked quite well. Until the walking / running animation. But idle and attack worked. I even got a decent jump animation.

For the last 2+weeks, I’ve been working on a pipeline for creating clean, production ready 2D sprite sheet animations. I’ve achieved really good results and I’m currently finalizing a full write up for the workflow. It will include a step by step guide you can feed to the coding agent of your choice to recreate the full workflow. And I’ll be including all the tips and tricks I’ve learned along the way. And yes, as you can see in the video it can handle the walk/run sequences as well 😄

I'm not a video guy, so I programmed a trailer in @threejs 🔊🔊

@LayrKits I think there's 3 ways to solve this completely: - better/faster/cleaner video model (not Kling) - a good image model that draws consistent spritesheets - a 3D -> 2D pipeline, but likely AI powered that would rig hair, clothes etc not just the base skeleton like mixamo does

Just ran into this when trying out isometric. It definitely takes more time. I was thinking to look into rigging next, which would also solve the outfit/weapon variation issue. Maybe combine the video model output with something like spline. But that’s just brainstorming, not sure how that would work. And I doubt it would keep the quality. Will have to try.

I think it's due to a combination of pricing and thought pacing. A lot of us actually enjoy leaving it alone for an hour or so, and no matter the complexity (unless you orchestrate a loop that's a huge token guzzler) it will likely take less than 2 hours to execute. I use fast sometimes for GPT though. It's sometimes too slow by default.

We have trained ourselves to believe slow=smart. So much so that I keep meeting people who don't use the Fast variants of Composer, GPT or Opus because they expect them to be dumber. Honestly not sure how we could fix this. Maybe the Cursor UI needs to render it as "Fast 🤑".

My issue with Kling is simply the time spent to orchestrate everything. And also, if you're doing higher res sprites (for 2D anime fighting) you'll notice artifacts from unnatural frame skipping (when you run it at 12 frames). I'm also doing isometric so it's a PITA to do all 3-4 directions. Trying to find a faster, more automatable way. Perhaps another video model would do better.

@startracker Still pure 2D. And my workflow starts much like yours with an image and Kling 😄. Excited to share it! Your work looks awesome too. Once I’m done with the article I’ll be sure to experiment with it.

@Wrathtank_avax @samtechcoded This is starting to look really really cool!

oh man a whole ass hub world @samtechcoded

@therainbowbun Mobile games want to have a word. This is silly cheap compared to what's out there on the market.

The fact that this is the price of an expansion or brand new game is baffling. I'm sorry but I cannot defend it.

@samtechcoded @threejs This is a great idea! Make a camera system and let the AI direct. I think you're on to something here. This might be the foundation of automating video once we can get UE5 level done with AI.

@Matisdottir @cursor_ai Pretty much! And they're being upgraded with every new project

@startracker @cursor_ai tools aging like fine wine

The engine built in @cursor_ai in February is really helpful now for a completely different game. If you remember some of my old videos from Jan/Feb, you've seen me paint cliffs and shrubs in the wasteland. Same engine. Different game. It just works. (with a few bugs) 🪲 Still hoping to release first half of May - most features are in, some UI/UX polish is left and then setting up the server for everyone to access!

@bread_ @willisdeving @megaeth Interested too if possible!

@willisdeving @megaeth Feel free to join the builder telegram. Can send a DM if interested.

11 minutes into jumping into @megaeth and I have already seen more winning than in 4 plus years on avax. Now to find the builder community.

@0xSmitty @megaeth It's really fast. Block times are really that fast. It opens up possibilities. Now if i could figure out where do builders hang out

I feel accidentally responsible for lots of friends buying @megaeth NFTs so want to lay out my thinking here First off, I'm not sure if the chain will do well or not, but I don't really care All my interactions with the team have been pleasant, and they seem to be placing a high priority on acquiring crypto-native users (that's us!) Personally I am happy to explore a chain where this is the case, as most of crypto is moving away from CT and towards "serious" or "institutional" applications I don't really mind taking losses if it feels like the chain's mission is to give users like me things to do Sure, maybe we are buying the top here like usual, but I'm willing to give the ecosystem a try since at the very least, CT seems to be their target audience I think you're doing yourself a disservice if you don't try to find more opportunities. It doesn't mean you have to buy expensive NFTs, just start paying attention and look for openings Follow people on other chains, respond to their tweets, expand your bubble! As a wise man once said: "Don't be a chain maxi, that's a slur"