https://hackyjs.com/posts/benchmarking-ernie-55-flash-base-separating-per-token-flops-from-kv-cache-bandwidth-at-1m-context Hacky JS en 2026-11-06T00:00:00.000Z Benchmarking ERNIE 5.5 Flash Base: Separating Per-Token FLOPs From KV-Cache Bandwidth at 1M Context ernie-5-5, moe-inference, kv-cache, benchmarking, long-context https://hackyjs.com/posts/designing-a-tested-failover-path-for-javascript-deploys-when-vercel-goes-down Hacky JS en 2026-11-06T00:00:00.000Z Designing a Tested Failover Path for JavaScript Deploys When Vercel Goes Down vercel, deployment, failover, javascript, redundancy https://hackyjs.com/posts/diffing-zyxel-gs1900-firmware-to-confirm-patch-3-actually-closes-the-root-rce Hacky JS en 2026-11-06T00:00:00.000Z Diffing Zyxel GS1900 Firmware to Confirm Patch 3 Actually Closes the Root RCE zyxel-gs1900, firmware-diffing, root-rce, patch-verification, embedded-security https://hackyjs.com/posts/electron-vs-rust-desktop-memory-measuring-working-sets-instead-of-trusting-290mb-headlines Hacky JS en 2026-11-06T00:00:00.000Z Electron vs Rust Desktop Memory: Measuring Working Sets Instead of Trusting 290MB Headlines electron, rust, desktop-apps, memory-profiling, webview