tokio-rs/tokio

Prefer this fork only if you must stay on an old Tokio lineage for compatibility. For new work or active maintenance, upstream Tokio is the clear choice.

Prefer this fork if you need Solana-tailored Tokio behavior and are prepared to carry a significant downstream maintenance burden. Prefer upstream if you want the broadest feature set, freshest fixes, and easiest long-term maintenance.

Prefer this fork if you specifically need stall-detection instrumentation plus Anthropic’s publishing workflow; prefer upstream if you want the newest Tokio features, platform fixes, and lower long-term maintenance burden.

Prefer upstream Tokio for almost all adopters. Choose this fork only if you need one of its specific runtime or macro adjustments and are prepared to own an aging, heavily behind codebase.

Prefer this fork only if you specifically want the Semgrep workflow and are comfortable carrying a very stale copy of Tokio. For general adoption, upstream is the better choice because this fork adds almost nothing beyond security scanning while lagging far behind.

Prefer the upstream Tokio unless you explicitly need this fork’s older behavior or custom edits and are prepared to maintain a heavily diverged runtime yourself. For new adoption, the staleness and missing upstream work are major negatives; for legacy compatibility, it may be useful as a frozen baseline.

Adopt this only if you specifically need the 2025-04 fork state. For production or active development, upstream is the better choice because this fork adds nothing and is substantially behind.

Prefer upstream unless you specifically need this exact older snapshot. This fork adds no visible functionality and is materially behind, so adopters inherit upstream lag without any compensating fork-specific benefit.

Prefer this fork only if you need its custom runtime/sync behavior and can own the maintenance burden. For most adopters, upstream Tokio is the safer choice because this fork is far behind and appears to remove or rewrite major runtime pieces.