FRIDAY, FEB27
1. European payments dead?, 2. Wuille linearizes, 3. AxeOS protects, 4. How to price agents
Supported by Proto and Bitkey - a part of the Bitcoin ecosystem at Block, Inc.
1. european
European payment processors are demanding full personal details, including names, birthdates, and home addresses, for Bitcoin Lightning transactions worth less than $10, despite no EU regulation actually requiring it. According to a detailed analysis of the practice, three separate EU frameworks (the Digital Services Act, anti-money laundering directives, and VAT rules) get routinely conflated by compliance departments, none of which mandate know-your-customer procedures for micropayments over pseudonymous networks. A real-world case study shows an $8.55 web hosting purchase through CoinGate requiring extensive identity collection before generating a Lightning invoice, turning one-second settlement into minutes of friction. "Risk-based assessment, the thing EU directives explicitly permit, gets abandoned in favor of blanket surveillance," the report notes, pointing to multimillion-euro fines from Lithuanian and Swedish regulators as the driver of institutional paranoia. Non-custodial alternatives like BTCPay Server bypass the problem entirely but demand additional technical sophistication most merchants lack. The result is a compliance moat that protects incumbents, marginalizes open-source alternatives, and degrades Lightning's core value proposition of fast, private, borderless payments.
-EDITOR·OP_DAILY SHARE TO X2. linearize
Bitcoin developer Pieter Wuille has published a detailed technical proposal for optimizing how Bitcoin's mempool orders transaction clusters, a key component of the cluster mempool project aimed at improving block construction and fee estimation. According to Wuille's post on the Bitcoin development forum, the algorithm builds toward finding optimal transaction linearization by iteratively identifying the highest-feerate subsets within a cluster, a problem he conjectures is NP-hard. The approach uses a branch-and-bound search with several pruning techniques, including "potential-set bounding" to discard unpromising search paths and a "jumping ahead" optimization that reduces worst-case complexity to an empirically observed O(1.6^n), down from O(2^n) in naive implementations. "We plan to set the cluster size limit so that even at the limit, the ancestor-set based linearization algorithm completes in a reasonable time," Wuille writes, noting the design guarantees results at least as good as current ancestor-set methods even under computation limits. The work represents a foundational piece of cluster mempool, which would replace Bitcoin Core's decades-old transaction sorting logic with a more principled framework for block template construction.
-EDITOR·OP_DAILY SHARE TO X3. protects
Community researcher mweinberg has published definitive proof on GitHub that two popular solo mining pools, LuckyMonster and zsolo.bid, were secretly redirecting home miners' hashrate to mine Bitcoin Cash while advertising BTC solo mining. According to Matt Howard at Solo Satoshi, the evidence was collected by monitoring 16 mining pools simultaneously and comparing the previous block hash each pool referenced, revealing that all three scam endpoints were locked to the BCH blockchain while 13 legitimate pools tracked Bitcoin's chain tip in unison. The AxeOS development team responded with firmware version 2.13.0, which introduces a coinbase transaction parser that decodes payout addresses embedded in pool work assignments and flags mismatches directly on the miner's dashboard. "This gives every Bitaxe owner an automated verification layer built directly into the firmware," developer mutatrum explained in the ESP-Miner GitHub discussion. The release also adds TLS-encrypted stratum connections and support for new hardware models. With open-source home miners having claimed over $1 million in known block rewards, the update converts what was once a manual forensic process into an automated defense available to every device on the network.
-EDITOR·OP_DAILY SHARE TO X4. agent
SaaS incumbents are trapped between protecting per-seat revenue and adapting to a market where AI agents eliminate the users those seats were sold to. According to Shawn Yeager, a Bain analysis of more than 30 SaaS vendors found that 65% are layering AI usage meters on top of existing seat pricing and 35% are raising per-seat prices with bundled AI, but zero have fully transitioned to outcome-based models. Salesforce illustrates the tension directly, shipping MCP protocol support across its platform while simultaneously running three separate pricing models for its Agentforce product. Meanwhile, startups built natively around outcomes are moving fast: Sierra charges per resolved customer interaction and reached $100 million in annual recurring revenue within 21 months, a model Yeager calls "nearly impossible to retrofit onto a seat-based business." Distribution dynamics compound the pressure, as protocol-compliant agents get discovered by every MCP-enabled client without a sales call, with one security startup signing eight unicorn customers in four months through this channel alone. The most durable opening, Yeager argues, lies in vertical agents that clear domain-specific regulatory barriers where horizontal AI tools cannot operate.
-EDITOR·OP_DAILY SHARE TO XIf you love OP_Daily, consider subscribing and forwarding it to your community.