AF Translate – AI Translation Editor, PO Editor & POT Generator for WordPress
Changelog
4.0.0
Major architectural refactoring of the Core AI Engine Subsystem Layer, consolidating multiple LLM and NMT request loops into unified central helper utilities.
Introduced standardized, deeply-sandboxed execution methods for remote API requests, credential retrieval, and system prompt formatting to simplify long-term code maintenance.
Resolved a critical synchronization bug on production Linux servers where bulk AI-translated strings failed to load back into the workbench editor rows after saving.
Engineered a proactive multi-tier Cache Invalidation Layer to flush WordPress core textdomain performance registries and persistent object caches immediately upon asset compilation.
Integrated automated LiteSpeed reverse-proxy header purging to ensure live server web servers drop stale translations instantly.
Hardened the JavaScript AJAX staging interceptor by deploying dynamic cache-busting timestamp parameters to completely bypass aggressive browser and edge proxy cache states.
3.0.0
Introduced the brand-new WordPress Core translation engine, allowing users to localize and automate AI translation pipelines for core system software and WordPress updates.
Completely restructured and re-ordered the admin sidebar submenus and horizontal navigation tabs to prioritize high-volume user workflows (Plugins and Themes).
Simplified administration terminology (e.g., Dashboard, Site Languages) to reduce cognitive friction for intermediate users.
Performance-optimized the dashboard activity log by introducing a strict 20-entry cap and disabling database option autoloading to eliminate front-end memory bloat.
Resolved an internal data architecture mismatch between the AJAX event tracking logger and the overview dashboard interface keys.
Fixed a Unix Epoch time calculation bug (“57 years ago”) by correctly binding MySQL datetime stamps to human-readable localized difference utilities.
2.0.0
Complete structural and design overhaul optimized for fast, zero-delay production environments.
Integrated the centralized AF Translate AI Settings Hub supporting 8 elite cloud translation engines and advanced large language models.
Implemented background bulk automation pipelines to process dense layout text records safely without crashing local server partitions.
Added an interactive staging sandbox to let users review automated AI translation cards before saving changes permanently to the server file structure.
Hardened internal directories by routing all output tracking variables to a highly secure path inside the main uploads directory.