identify_gate

Identify pacing across a shard fleet. Discord allows roughly one IDENTIFY per 5 seconds per guild bucket and reports the real budget as max_concurrency on GET /gateway/bot. Fetching that value is a later milestone; until then the pacing lives here as documentation and a conservative default. A fleet that ignores it meets close code 4005 (already authenticated) or session-start-limit 429s. Stability: Experimental.

Not wired into anything yet — tadpole/bot runs one shard, which never needs pacing. The one function is pure math for the fleet code to come; see also tadpole/gateway.

Values

recommended_delay_ms

</>

pub fn recommended_delay_ms(shard_count: Int) -> Int

Milliseconds to wait between IDENTIFYs when starting shard_count shards from one process, assuming the worst case of a single guild bucket. Large bots (max_concurrency > 1) can safely go faster once /gateway/bot is wired up; small fleets cannot go faster than one identify per 5 seconds at all.