TL;DR: llama.cpp now allows configuring the output stream for build information, improving flexibility for developers.
Summary: The latest llama.cpp update (b10798) introduces a configurable output stream for build information. The llama_print_build_info function can now write to a caller-provided FILE* instead of hardcoding stderr, with stderr remaining the default for backward compatibility. This change allows tools like llama-app to direct version output to stdout as expected by users.
Why it matters: This enhancement provides greater control over logging and integration for AI developers building on llama.cpp. Builders should review their build processes and consider leveraging this new flexibility for cleaner output management.
Source: github_releases