0.88.0
vladmandic/human0.88.0Sep 25, 2026by takahirom
Upgrading?
AI Summary
Optimizes focus navigation speed on TV devices by reducing redundant view hierarchy fetches.
Key Highlights
- Focus moves are approximately twice as fast on Android TV (0.96s to 0.48s per press).
- Reads view hierarchy once per move and uses settled hierarchy to decide the next key.
New Features
- Improved focus navigation performance
Full Release Notes
Moving focus on a TV device is now about twice as fast per key press. On an Android TV emulator, each D-pad press in a focus move went from about 0.96s to about 0.48s, and from 1.1–1.4s to about 0.73s on a screen that is still loading content. ## Faster Focus Moves (#454, #455) - Each focus move iteration reads the view hierarchy once, and uses that frame for both the current focus and the target. A selector target that is not in that frame is still polled on fresh hierarchies as before. - The next key is decided on the hierarchy Maestro's settle already returned after the key press, instead of a fixed screenshot based wait and another fetch. The settle only returns once two consecutive fetches are equal, bounds included, so a scroll or focus change still in progress keeps it waiting. - `[timing]` debug logs now cover each part of a focus move (`focus.viewHierarchy`, `focus.pressKey`, `focus.waitForAppToSettle`). ## What's Changed * Read the focus and the target from one hierarchy fetch per focus move by @takahirom in https://github.com/takahirom/arbigent/pull/454 * Decide the next focus key on the settled hierarchy and drop the fixed wait by @takahirom in https://github.com/takahirom/arbigent/pull/455 **Full Changelog**: https://github.com/takahirom/arbigent/compare/0.87.0...0.88.0
Related Repositories
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
Open-source foundation of ToolJet AI - the enterprise app generation platform for internal tools, dashboards, business applications, workflows and AI agents. Build visually, from a prompt, or from Claude Code, Codex and Cursor over MCP 🚀
1,324-exercise fitness dataset — animation GIFs, 180×180 thumbnails, muscle-group & equipment data, and step-by-step instructions in 6 languages. The exercise data layer behind the LogPress app.
A complete web-based remote monitoring and management web site. Once setup you can install agents and perform remote desktop session to devices on the local network or over the Internet.
A general purpose HTTP reverse proxy and forwarding tool. Now written in Go!