2026-11-15
Trench lessons — 2026-11-15
Key takeaways
- Why SIMD-0460 keeps slipping: Anza paused feature activations until after Breakpoint
- Helius Historical Accounts API (beta, Nov 14): point-in-time token balances
- Nov 14 day: launch-block buyouts are falling sharply; grads are slower
- Two public filter examples from the same day
- Smaller items
Why SIMD-0460 keeps slipping: Anza paused feature activations until after Breakpointevidence: B: SolanaHub recap; A: on-chain RPC
- What: on Nov 10, Anza published the Agave v4.5 schedule and paused feature activations until after Breakpoint (Nov 15). The v4.5 mainnet-beta upgrade candidate targets Nov 27. That's why SIMD-0460 skipped both epoch 1086 and 1087. An on-chain check ~10:15 AM CT still shows the feature account System-owned with 0 data (not active). Current epoch is 1087; ChainReport says "queued", but given the pause, a flip at epoch 1088 (~11:36 PM CT tonight) is unlikely. Also watch the Agave v4.5 rollout around Nov 27 and the rent stages (SIMD-0437-3/-4: lamports/byte 5,080 → 2,575 → 1,322). Lower rent shrinks ATA-create cost.
Helius Historical Accounts API (beta, Nov 14): point-in-time token balancesevidence: B: Helius docs
- What: getHistoricalAccountInfo and getHistoricalTokenAccountsByOwner return account state at a past slot, or right after a given tx index in a slot. History goes back to slot 444,959,999 (Sep 7). 0 credits during beta, but only on Business plans and above. Rate limits: 100 rps and 10 rps. Useful for reading a wallet's token account right after its last sell, or a bonding-curve state right before a buy (true entry mcap).
Nov 14 day: launch-block buyouts are falling sharply; grads are slowerevidence: B: pumpdata
- What: 40,767 launched, 803 graduated (1.97%), 3.09M SOL traded (−13% vs 7-day), 497,695 wallets (−12%). Launch-block buyouts were 128, 65% under the 30-day average (Nov 13: 225, −41%). Graduations in the first second fell to 73 (9.1%) from 165 (18.2%). The median time for the rest went up to 8.2 min (from 4.7). 95.4% of graduates closed below their graduation mcap. No week-old coin held all day. Macro backdrop: SOL ~$75, down 16% over 7 days, BTC < $67K, $330M in liquidations (C: bitcoin.com).
Two public filter examples from the same dayevidence: B: pumpdata
- Wash volume: PCULT did 23,249 SOL of volume with 30 traders (~775 SOL per trader) and peaked at only $249K. A practical public filter is to treat very high SOL volume per unique trader (for example above ~50 SOL) as wash-like churn rather than organic flow.
- Grad-farm deployer: one deployer launched 71 coins and graduated 12 the same day (a lagging all-time record still showed 28/1). Same-day deployer graduation counts from fresh data are more timely than lagging third-party records.
Smaller itemsevidence: no action
- Venue programs: pump.fun bonding curve, PumpSwap, Raydium CPMM/CLMM/AMM v4, Jupiter v6, LaunchLab, Meteora DBC, and Token-2022 are unchanged from Nov 14 (curve still on the Nov 13 redeploy at programdata slot 469,055,660).
- Pump.fun sent another 64,666 SOL to Kraken (lifetime ≈5.62M SOL sold). Team/investor wallets moved 1.458B unlocked PUMP to Coinbase Prime. This is general selling pressure.
- Jupiter Pulse (token discovery by organic volume / popular traders) and Portfolio V3 are rolling out before Breakpoint.
- Draft SIMD-0723 would remove durable nonces, with no mainnet activation before May 5, 2027.
- Breakpoint is today (Nov 15). Tomorrow is a natural check for any pump.fun, Jito, or Alpenglow announcements that touch transaction landing.
Sources
- https://solanahub.de/en/news/solana-news-wochenrueckblick-2026-11-15/
- https://chainreport.co/fork-desk/solana-feature-gates/
- https://www.helius.dev/docs/rpc/historical-accounts
- https://github.com/helius-labs/helius-rust-sdk/releases
- https://pumpdata.fun/stats/2026-11-14
- https://pumpdata.fun/stats/2026-11-13
- https://news.bitcoin.com/crypto-news/pump-fun-sold-888m-sol-kraken-deposit/
- https://defiedge.com/defi/pump-fun-vesting-wallets-transfer-november-tranche-to-co-4826c71a
- https://solanahub.de/en/news/jupiter-aggregate-everything/
Evidence grades: A = large dataset, B = official docs, C = vendor/thread/small sample. Research notes for education only. Not financial advice.