JustJoeFX

@VukRosic99 Its ist only useless if you dont publish it was useless.

do you think most of the research is useless?

@blader Not working with root.

--dangerously-skip-permissions isn't just great for starting claude, it's a great way to live life

@zaimiri Comparing qwen coder 2.5 with opus 4.6 is beyond everything...

I stopped paying for Claude Code. $200/month for an API subscription to write code. Then Ollama dropped Anthropic API compatibility. Now Claude Code connects to free, local models on my machine. Here's the exact setup (took me 10 minutes): 1. Install Ollama → curl -fsSL https:// ollama .com/install .sh | sh 2. Pull a model → ollama pull qwen2.5-coder 3. Point Claude Code at localhost → ANTHROPIC_BASE_URL=http: //localhost :11434 claude That's it. Claude Code thinks it's talking to Anthropic. It's talking to your laptop. Best models I've tested: • Qwen 2.5 Coder - best all-around for code generation • DeepSeek-Coder - strongest at debugging and refactoring • Llama 3 - solid general reasoning Local models aren't Sonnet or Opus. Complex multi-file refactors still stumble. Long context windows get messy. But for everyday coding - scaffolding, tests, quick edits, boilerplate - they handle it fine. Your code never leaves your machine. Your bill goes from $200/month to $0. Your API key stays in your pocket. (Save this for later.)

@OpenRouter IT worked a few hours later.

@JustJoeFX Can you dm your openrouter email to us?

@vxdb pen & paper.

What password manager do you use? I wanna see if anyone is moving away from 1Password since the price increase

@renkoblocks You forgot Inflation.

@Xianbao_QIAN Opus 4.5 Open source...

Oh! What's next?

@virasana_ni Crypto is wet dream of 1984.

@0xSero 1. clear context 2. review code manually before prompting 3. make a clear prompt that a experienced coder without business logic knowledge understands and do it in plan mode. 4. opus makes a one shot solve. 5. test manually and review changes manually 5. git commit

Note to self, don't use subagents. 20% in an hour

When no one Cares, why should you Care?

@AnthropicAI why is Opus so retard since yesterday??

@bcherny please add a Feature for global settings to start in specific Mode. Plan mode for me.

@trikcode I code since 1995.

I am doing a project about elderly programmers. If you are a programmer and over 25 please DM

@bcherny @theo can i run a claude code instance in a vps and 24/7 with --dangerously skip premission code while i am in train, plan, car?? like "start a new project in subfolder worldpeace and develop a strategy to bring peace to the world" with ralph wiggum plugin to work till reached??

@theo Yes but then go to the Code sidebar and you can use Claude Code from your phone This has been launched for a bit but I was surprised so few people knew about it

Don't get too excited, just a download link for the mobile app 🙃

@TismStrategy @gemchange_ltd @Polymarket @PolymarketTrade So vps are allowed?? Thx will try.

@JustJoeFX @gemchange_ltd @Polymarket @PolymarketTrade I tried Netherlands and got good latency. Still not enough to compete with the successful bots though

devs are making $10k-200k monthly on polymarket no they're not lol here's what those threads don't tell you zero fees = zero friction for HFT shops to eat your lunch. In an efficient market, these simple arbitrage opportunities are fleeting at best. High-frequency trading bots and sophisticated arbitrage algorithms continuously scan for such discrepancies, capturing profits within milliseconds. You're not competing with other python hobbyists You're competing with rust bots running on dedicated polygon nodes with sub-millisecond execution. That "simple python script" from twitter? For Python systems, these numbers were more in the 250-500 microsecond range per message, meanwhile actual HFT systems in rust/c++? Processing time per quote message is about 12 microseconds. that's 20-40x slower before you even place an order. what you actually need: - dedicated polygon RPC nodes (public ones rate-limit you into irrelevance). 99.99% uptime, 100% healthy nodes in production, and average latency max 86.5 ms for premium infrastructure - VPS colocated near polymarket servers. Ultra-low latency and physically close to Polymarket's servers - Code written in rust. There's literally a polymarket-hft crate optimized for trading scenarios requiring fast execution - Proper quant algorithms you can't vibecode because garbage collection is extremely useful during development but often sub-optimal for certain high frequency trading strategies so those "$200k/month" traders? They exist. They're the top 0.5% and they're running infrastructure that costs more than your yearly salary.

@gemchange_ltd @Polymarket @PolymarketTrade Tried German,Mexico,Japan....

@gemchange_ltd @Polymarket @PolymarketTrade Not working. Polygon/cloudflare Blocks Datacenter ips. Tested with multiple aws Locations. And some smaller Providers. All are blocked from Post requests.... You can read but Not trade....

Best % in the world are 5% from Beer.

@virasana_ni Why do experiments with Prop Account? Why not Paper Trade on Tradinview or Sim in Ninja?

@virasana_ni Gc long only?