Skip to content

v0.2.3

Choose a tag to compare

View all tags
@goshitsarch-eng goshitsarch-eng released this 13 Feb 19:04
· 24 commits to main since this release
v0.2.3
60aed78

Fixed

  • Large file downloads failing at ~400MB: the reqwest HTTP client was configured with .timeout() which sets a total request deadline (default 60s), not a per-read idle timeout — downloads taking longer than 60 seconds would be killed mid-stream; switched to .read_timeout() which resets after each successful read
  • Segment failures killing entire download: wired the existing RetryPolicy into the segmented download path — each segment now retries with exponential backoff and resumes from the byte position already written instead of failing the whole download immediately
  • Rate limiter silent overflow: speed limits were cast from u64 to u32 with truncation; values above u32::MAX now clamp instead of wrapping to zero
  • SQLite busy errors with concurrent readers: added busy_timeout(5s) so external tools (GUI, CLI monitors) reading the database don't cause SQLITE_BUSY failures
  • Segment persistence not atomic: wrapped save_segments() DELETE+INSERT in an explicit transaction to prevent corrupt resume data on crash
  • WebSeed unbounded memory growth: replaced unbounded event channel with a bounded channel (max_connections * 2 capacity) to apply backpressure when the consumer falls behind
  • WebSeed timeout same as main client: applied the same .timeout() → .read_timeout() fix to the WebSeed HTTP client

Full Changelog: v0.2.2...v0.2.3

Assets 2
Loading