Q2RE Map Trainer v1.2.1

📝 Changelog

v1.2.1 (2026-06-20)

• Spawn Trainer bot placement — bot stands on spawn points instead of hovering in a jump pose; ground snap runs after final spawn placement and is limited to 32 units so raised spawn platforms are not bypassed to the floor beneath

v1.2.0 (2026-06-20)

• Settings persist across game restarts — archived trainer_* cvars restore menu options after exit/relaunch (map-change persistence from v1.1.0 unchanged)
• Bhop timing uses real seconds — chain timeout and perfect/late feedback are frame-rate independent
• Spawn Trainer — resume retries when a client slot is busy; killing the bot no longer awards DM frags; README documents slot usage
• Path training — welcome prompt re-shows after respawn before first pickup; item categories use registry IF_* flags (custom items with e.g. IF_HEALTH follow the Health toggle)
• Performance — path trainer caches item edicts instead of scanning entities each check
• Robustness — trainer_mode_t centralizes path/timing exclusivity; TagMalloc fail-soft; trainer.log appends with session banners
• Developer — trainer_logic host unit tests (test.bat), CI test step, menu row index enums, g_local.h prototype cleanup

v1.1.0 (2026-06-20)

• Settings now persist across map changes — trainer configuration is retained when the map changes, and active trainers auto-resume on the new map (path training rebuilds the new map's item list, the spawn bot re-spawns, timings start fresh)
• Build environment overhaul — vcpkg is now a git submodule with a setup-vcpkg.ps1 bootstrap; build.bat/play.bat moved to the repo root (build and deploy are separate steps)
• Consolidated version constants into src/trainer/trainer_version.h (single source of truth, alongside the repo-root VERSION file)