Aniflow 1.3

Multi-GPU fix and quality-of-life release.

Fixed

• On laptops with an integrated + discrete GPU, the shader chain, AI upscaling and RIFE all ran on the integrated GPU (device 0) while the discrete card idled — up to 10x slower encodes. Aniflow now detects the Vulkan device list at startup and picks the discrete GPU automatically; the GPU field in Settings (0 = auto) overrides it everywhere, and the detected devices are listed right under the field.
• RIFE jobs no longer break when the app is installed under a path containing special characters (e.g. the %LOCALAPPDATA% one-line install with an unusual username) — every path on the internal command line now travels via environment variables.
• Long AI jobs no longer slow the UI down: progress counting probes only new frames and stderr is read incrementally instead of re-reading a multi-megabyte file twice a second; Real-ESRGAN's percent spam is filtered out of the log.

Added

• In-app update check: on startup Aniflow quietly asks GitHub for the latest release and prints a log line when a newer version exists. No popups, silent when offline.
• The job log now shows which GPU was selected on multi-GPU systems; -SelfTest prints the full device list.