# Data ready (partial, usable): hook e044 tape, ppdata #632, 11 Oct 2026

**File** `/opt/ppcopy/r6e044t.db` (research-box; also /opt/sw/ppdata/r6e044t.db), 1.4 GB. Built from the chainlogs tape (not RPC), same schema as r6weth.db plus tables `init` and `cov`.
**Pools:** 1,170 (570 `fam=run` = e1-runners pools of hook 0xe5e7..e044, 600 `fam=ctl` = controls). 1,168 have events.
**Events:** 5.14M rows in `ev` (5,136,314 swaps kind 1, 2,212 ModifyLiquidity kind 10), each pool cut at first swap + 216,000 blocks (~6 h), so a 3 h climb to 1M is covered.
**Columns ev:** pool, venue, block, t (interpolated from block), li, tx15 (15-hex tx prefix), kind, a, q (raw pool-side amounts, + = into pool), px, liq, sender, recipient, tx_from.
**tx_from is EMPTY** (tape has no full tx hash; the earlier RPC job fetched it for only 40 pools and is stopped, it ran at ~3 pools/h on free RPC). Your entry test (price/cap path + pull) does not need it.
**Coverage caveat (important):** the chainlogs tape is complete from block 56,000,000 and partial below (b052-b055 still being filled; nothing below 52.0M). `cov.full_cov=1` marks pools with init_block >= 56M: **run 320, ctl 301**. The other 250 run / 299 ctl pools (init 28.8M-56M) may miss their first swaps, so their first-swap time and early climb are biased late: for your fast-climber test use full_cov=1 only (n=320 touchers, enough for the n>=100 pass bar), or wait for chainlogs b052-b055 to finish and I rerun (6 min: `r6e044t.py` in /opt/crew/work/ppdata/e044t).
**Tape gap:** blocks 84,943,668-85,000,000 empty, tape ends ~85.34M.
**Query:** `select e.t,e.kind,e.px,e.a,e.q,e.liq from ev e join cov using(pool) where full_cov=1 and pool=? order by block,li;`  Join `meta` for fam, created_*, first_swap_block, hooks; `px` is token price in quote units (cap = px x supply, supply via /opt/ppcopy/nm2.db).