@iamcheyan: The task was to reconstruct zircon mir3 with Godot and port it to Linux. After two days, I successfully got both the server and client running.
Summary
The author reconstructed zircon mir3 with Godot and ported it to Linux, successfully getting the server and client running within two days. He also mentioned that the AI model gpt-5.6-luan efficiently completed multiple long-running tasks, with 62% of the quota remaining.
View Cached Full Text
Cached at: 08/09/26, 03:17 AM
The task was to rebuild Zircon Mir3 with Godot and port it to Linux. It took two days, and I’ve successfully got both the server and client running. https://t.co/eLhbOQlPQG
徹言 (@iamcheyan): gpt-5.6-luan is so economical. I opened several goals — one ran for 14 hours, two others ran for 9 hours, and another for 5 hours, and there’s still 62% of the quota left.
Similar Articles
@xiaomovps: After a company starts using AI, they quickly hit several hard problems: whether data can be externalized, whether costs can be controlled, and whether to build internal models themselves. This article documents a very real weekend operation—remotely connecting to the company's DGX Spark and hands-on running Ling-3.0-flash. Not stopping at 'can it run', but…
This article documents the complete process of the author remotely connecting to the company's DGX Spark server on the weekend to successfully deploy the Ling-3.0-flash model, including selection, deployment, performance testing, and integration with development tools, and shares insights on local deployment as a controllable intermediate state.
@zhixianio: After receiving the new machine, I began an 'ascetic' practice of forcing myself to use local models for common tasks. I thought it would be painful, but both speed and quality greatly exceeded my expectations: Model: Qwen3.6-35B-A3B-oQ6-fp16-mtp, Running: oMLX, with N…
The author uses the Qwen3.6-35B-A3B model and oMLX tool on the new local machine for daily tasks, finding that both speed and quality far exceed expectations, even outperforming remote LLMs in PA and coding scenarios, demonstrating a significant improvement in on-device AI capabilities.
@xingbugengming: I don't have the GPT 5.6-Sol model, but I spent an hour using up five hours' worth of quota to test the frontend capabilities of GPT 5.6-Terra. The results were very impressive, especially when handling image matting—it's amazing. I asked it to create a webpage: an interactive card of Guan Yu from the Three Kingdoms, with mouse drag...
The user tested the frontend capabilities of GPT 5.6-Terra, creating an interactive Guan Yu card webpage from the Three Kingdoms. The results were outstanding, particularly in image matting and depth-of-field perspective.
@SaitoWu: https://x.com/SaitoWu/status/2069076084495438186
This article describes using the Codex AI agent to automatically migrate terminal shell configuration from Oh My Zsh to Zinit + Starship + Rust toolchain, demonstrating the AI's ability to perform engineering steps such as backup, key isolation, and performance analysis, ultimately achieving an order-of-magnitude improvement in startup speed.
@AI_Community_9: Unsloth Desktop takes the experience of Qwen3.8 27B to new heights. Out-of-the-box + cross-platform, the core is optimizing VRAM for fine-tuning and inference to the extreme (VRAM saved by 70%, speed increased by 2 times). It packs 'training, inference, and Agent toolchain' into a free, open-source desktop application, …
Unsloth Desktop is a free, open-source desktop application that integrates training, inference, and Agent toolchain through extreme optimization of VRAM and speed, significantly improving the user experience of AI models like Qwen3.8 27B.