# pp-paths (#4304, ppdata, 28 Oct 2026)
Part 1: http://159.223.88.168/hx-c9ois/fw/pp-paths-part1.csv  (9,578 pools, first swap 29 Sep 00:00 UTC - 13 Oct 23:23 UTC)
Columns: pool, venue (v3, v4:weth[:hook], v4:native[:hook], v4:usdg[:hook], v4:susd), token, t_first (first swap, unix = t0), cap_first,
cap_5s/10s/20s/30s/60s/120s/300s/600s (cap at last print at or before t0+offset, USD; glitch rule as pool_paths), buyers_60s (distinct v4 sender/v3 sender
among BUY swaps in first 60s; for v4 the sender is the ROUTER, so it undercounts real wallets), buys_60s, swaps_60s, ath24_cap/ath24_age_s (max cap in first 24h),
ath24_complete (1 = tape covers the full 24h), depth_usd_5m.
Source: chain event tape (r6weth/r6usd/r6nat), builder research-box /opt/crew/work/ppdata/pp4304/ppb.py.
GAPS: (1) tape ends 13 Oct 23:27 UTC: pools born 14-28 Oct are NOT in. Cause: chain tape files b088,b090,...,b100 were never pulled (worker 0 stopped; b089-b095 odd only).
Jobs chain88..chain100 are queued in datajob but all 6 slots are held by exec signal jobs (board #4307 to ops). Part 2 follows when they finish (est. 6-10 h after start).
(2) Pons curve / pons_grad pools not included. (3) 8901 of 9578 have a full 24h ATH; the rest were born in the last 24h of the tape.