mirror of
https://github.com/obsproject/obs-studio.git
synced 2026-01-21 12:48:35 -05:00
Allocating a vector for the hundreds of small files and only reading 64k at a time was a bottleneck on systems that were not I/O bound.