July Blockchain Technology Update: Bitcoin's CTV+CSFS Debate, ETH & SOL Expand Blockspace
Written by | GaryMa, Wu Blockchain
The WuBlockchain summarizes key developments in the blockchain technology space for July:
Bitcoin
● OP_CHECKSIGFROMSTACK (CSFS): promise and controversy. Developers discussed how combining CTV with CSFS could accelerate the Lightning Network’s move to Point Time-Locked Contracts (PTLCs) and simplify protocol stacking. Progress is hampered by limited tooling and standardization, and may require trade-offs in on-chain efficiency. A public letter (66 signatories) urged Bitcoin Core contributors to prioritize review and integration of CTV+CSFS within six months, sparking intense debate. Points of contention include whether legacy script support is necessary; the limits of CTV vaults (e.g., address reuse and theft risks) versus the advantages of key deletion and static recovery. Core contributors criticized the perceived pressure from the letter, emphasizing technical consensus over public campaigns. In BitVM discussions, CTV+CSFS appears to shrink transaction sizes and enable non-interactive anchoring; some ideas have been validated in test software.
Ethereum
● Fusaka upgrade milestones. Tentative mainnet timeline: Fusaka release candidates ready September 1; public testnets upgrade September 15/22 and October 6; mainnet upgrade November 5.
● Glamsterdam planning. Core developers began scoping the next upgrade; the shortlist includes EIP-7732 (ePBS), EIP-7783 (6-second slot time), and EIP-7805 (FOCIL).
● Gas limit raised. Ethereum mainnet gas limit increased to 45 million.
● EIP-7983 (Buterin & Wahrstätter). Proposes a per-transaction gas cap of 16.77 million (independent of the block gas limit) to strengthen DoS resistance, mitigate load imbalance from high-gas transactions, reduce state bloat and validation delays, and improve zkVM compatibility and parallel execution. Status: draft.
Ethereum L2s
● Polygon PoS — Heimdall v2 hard fork. The most complex upgrade since 2020: Heimdall moves from Tendermint + Cosmos SDK v0.37 to CometBFT + Cosmos SDK v0.50, cutting finality to \~5 seconds and retiring legacy technical debt.
Solana
● Jito’s Block Assembly Marketplace (BAM). Co-launched by Jito Foundation and Jito Labs, BAM uses Trusted Execution Environments (TEEs) to enable verifiable ordering, programmable execution logic, and new revenue-sharing. It consists of BAM Nodes, BAM-enabled validator clients, and a plugin system that supports custom ordering logic; developers can charge fees, while node operators, validators, and stakers share revenue via a built-in model.
● SIMD-0256 activated on mainnet, lifting the block limit to 60M CUs.
● SIMD-0286 proposed, to raise per-block compute from 60M to 100M compute units (CUs).
BNB Chain
● Two hard forks in H1 2025 (Lorentz & Maxwell). Block time reduced to 0.75s, finality to 1.875s; daily tx peak 17.6M; average gas ≈ \$0.01. The Goodwill Alliance reports a 95% reduction in harmful MEV. In H2, the block gas limit will rise to 1 billion to support higher on-chain activity. For 2026, goals include faster confirmations, higher throughput, and native privacy features.
● BEP-593 (community proposal). Introduces a high-efficiency consensus path with: validator node ID registration for better operator management; faster direct message propagation; lower congestion and improved overall performance.
● FastFinality integrated with Binance. Confirmations drop from 15 blocks to 5 (an 83% speed-up), backed by ≥66% validator consensus, delivering quicker settlement for DeFi, NFTs, and on-chain games.
Base
● BaseApp launch. Framed by the community as a crypto-native blend of WeChat + TikTok + Alipay + an App Store, BaseApp unifies AI agents, mini-app collections, Farcaster feeds, and video streams for seamless on-chain social, payments, and app distribution.
Hyperliquid
● CoreWriter on mainnet. Marks bidirectional integration between HyperEVM (smart-contract layer) and HyperCore (high-performance trading engine). CoreWriter lets HyperEVM write to / mutate HyperCore state, complementing prior Read Precompiles (price, orderbook, positions, etc.), without off-chain systems or complex glue. This is a step from a pure perpetuals DEX toward a programmable finance stack.
Cosmos
● Cosmos Hub pauses EVM platform plan. The team will reassess, citing high VM build cost, user-experience trade-offs, and misalignment with Cosmos’s multi-chain interoperability advantage. The focus shifts to IBC-based services, strengthening the Hub’s role in the L1 services market.
Other Notables
● DoubleZero protocol adds native multicast. Today’s chains largely rely on unicast, redundantly sending identical packets to each peer — wasting bandwidth and adding latency. Multicast enables single-send, in-network replication, subscriber-only delivery, and deterministic fiber-path routing. DoubleZero introduces on-chain group configuration and permissions to bring multicast to decentralized systems, targeting TradFi-grade performance.
● Sonic Research — SonicCS 2.0 consensus. A DAG-based protocol with overlapping elections to parallelize the ordering of multiple blocks. It models voting as 0–1 matrices and uses SIMD instruction sets (e.g., AVX2) to vectorize elections, accelerating vote and aggregation. In mainnet tests over 200 epochs, average throughput doubled and memory usage fell 68%. The team plans client integration and a technical report.
● GMX v1 GLP exploit on Arbitrum. \~\$40M in tokens moved to an unknown wallet. SlowMist attributes the root cause to a design flaw: short operations immediately update globalShortAveragePrices, impacting AUM and allowing GLP price manipulation. The attacker leveraged a reentrancy path by enabling `timelock.enableLeverage` during keeper-executed orders, opened large short positions to skew the global average price, artificially inflated GLP in a single transaction, and profited via redemption arbitrage.
● Sonic Labs — SonicVM deep dive. Through a new long-format instruction encoding, hash-function optimizations, and instruction-scheduling refactors, SonicVM achieves \~1300 MGas/s on Sonic mainnet — about 6.5× faster than Geth. A bespoke conformance framework ensures compatibility across 14 Geth spec versions, with >16 billion test vectors executed.
● Anoma testnet live (July 15). Introduces an intent-centric decentralized OS with 3D visualization of cross-chain operations. Aims to reduce multi-chain app complexity; collaborators include Celestia, EigenLayer, Near, and RiscZero. Mainnet features will roll out across multiple L1s and L2s.
● Core Foundation — Rev+. A protocol-level on-chain revenue-sharing mechanism offering sustained incentives to developers, stablecoin issuers, and DAOs. Revenues from on-chain activity are distributed based on contribution metrics such as transaction volume and address activity.
● Tac L1 mainnet. A third-party Layer 1 connecting EVM dApps to the TON / Telegram ecosystem, enabling Telegram users to access DeFi protocols (e.g., Curve, Morpho, Euler) directly in-app, with no code rewrites for developers. Tac has raised \$11.5M, led by Hack VC.
● Mira Verify. An AI content-verification network using multi-model consensus to detect hallucinations and misinformation. It extracts factual claims from text, then independent AI models label results as “true,” “false,” or “no consensus.”
● Conflux Tree-Graph public chain v3.0 (Shanghai). Planned for August launch, targeting 15,000 TPS, with support for AI agents and RWA settlement. Built by a Tsinghua-affiliated team, Conflux already interoperates with several major chains and plans offshore-RMB stablecoin and cross-border settlement pilots in Belt and Road economies.
End of report.
Follow us
Twitter: https://twitter.com/WuBlockchain
Telegram: https://t.me/wublockchainenglish