Skip to content

docs: document PAUSE, and say what the two ESP-NOW figures measured - #18

Merged
Em3ritus merged 1 commit into
mainfrom
docs/pause-espnow-2026-10-07
Oct 7, 2026
Merged

Em3ritus merged 1 commit into
mainfrom
docs/pause-espnow-2026-10-07

Conversation

@Em3ritus

@Em3ritus Em3ritus commented Oct 7, 2026

Copy link
Copy Markdown
Owner

Follow-up to #17, covering the two items it left open.

PAUSE was the only preset the README never described. From the code on main:

  • churn_tick returns before phantom_lifecycle / ble_devices_tick while paused, so the BLE crowd stays on air with frozen addresses and the 15-minute ceiling stops applying to it until resume.
  • Wi-Fi probe MAC rotation (coexist.c → probe_agents_rotate_tick) isn't gated by pause and keeps going.
  • The console posture reads DARK while paused, even though the frozen BLE crowd is still transmitting.

The README now describes PAUSE under Signed fleet control and qualifies the "every identifier capped at 15 minutes" line in Features.

ESP-NOW figures: "~99/min" and "66.9 → 4.2/min" read as a before/after that doesn't add up. They come from two different instruments:

  • ~99.4/min: a fleet-wide Kismet capture of five boards (a3d2cb7).
  • 66.9 → 4.2/min: one sniffer parked on ch1, counting by decoded type (a7ecd4b, d110810). It undercounts the channel-hopping decoys.

README and CHANGELOG now say so, and that the fleet-wide number hasn't been re-measured since the fix.

🤖 Generated with Claude Code

PAUSE was the one preset the README never described. churn_tick returns
before phantom_lifecycle/ble_devices_tick while paused, so the BLE crowd
stays on air with frozen addresses and the 15-minute ceiling stops
applying to it; Wi-Fi probe rotation (coexist.c) is not gated and keeps
going. The console shows DARK meanwhile. Say all of that, and qualify the
15-minute claim in Features.

"~99/min" and "66.9 -> 4.2/min" read as a contradictory before/after.
They are different instruments: a fleet-wide Kismet capture (a3d2cb7)
and a single ch1-parked sniffer counting by decoded type (a7ecd4b,
d110810), which undercounts the hopping decoys. Say so, and that the
fleet-wide number was not re-taken after the fix.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@Em3ritus
Em3ritus merged commit ef25f86 into main Oct 7, 2026
4 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant