Miner signaling is a vote on bitcoins rules

Miner signaling on protocol upgrades is a vote: flagged support in blocks is the decision-making process by which Bitcoin's rules change, and miners are the electorate.

For (strongest version as argued): The miners said so themselves, in public, at the moment it mattered. At Scaling Hong Kong the operator of one of the largest pools argued that Bitcoin is a proof-of-work system and there is no mechanism for anyone else to vote; in a roundtable the miners' collective position was that they have real businesses, invested real money, and produce the blocks — real power over the network, which the developers lack. The industry side agreed: a CEO's widely read blog stated that bitcoin "has a built in upgrade mechanism" and that if a majority of miners vote for an upgrade, "by definition this is the new version of bitcoin." And behavior matched the theory on every side: Bitcoin XT and Bitcoin Classic each staked activation on a 75 percent miner threshold, SegWit on a 95 percent one — whatever people said flags meant, every faction's path to change ran through them. The whitepaper's closing line, nodes voting "with their CPU power," circulated as the design's own endorsement.

Against (strongest version as argued): The same source records the opposing design history: the flag in the block header was a safety signalling mechanism, indicating readiness so that activation would not split the chain — and the record supports that reading. P2SH activated in 2012 on a roughly 55 percent threshold and non-upgraded miners went on producing invalid blocks afterward, which is why the 95 percent threshold was adopted; in July 2015, BIP-66's activation split the chain for several blocks because miners had flagged support without actually validating. A flag, in other words, is not even reliable proof of enforcement, let alone a franchise. The small-block position drew the line explicitly: gauging miner opinion is fine, but combining rule activation with voting over contested proposals was dangerous and inappropriate, because activation is a piece of network security machinery, not a ballot. The corpus's verified entries state the settled mechanics the vote theory runs into: nodes discard invalid blocks regardless of the work behind them (nodes-enforce-rules, miners-capabilities-limits).

Nuance: Two different things get called "voting." Using flags to measure support is uncontroversial — even the small-block camp accepted gauging miner opinion. Flags as authority to change the rules is the contested part. The threshold fight was really this dispute in numeric form: read as an election, 95 percent is an absurd supermajority that hands any small coalition a veto (the large-block objection, verbatim); read as a safety check, 95 percent is the entire point. The source's sharpest observation is that each camp assumed everyone else would behave as they themselves would — which is why each kept predicting the other's moves wrongly. See the fact soft-fork-signaling-thresholds for the activation-mechanics history this dispute sits on.

Sourcing: a single owner-research file — the owner's reading notes on The Blocksize War (Bier), a secondary account whose substantive coverage stops at February 2016 plus fragments. INGEST_PLAYBOOK.md caps a claim sourced this way at CONTESTED with confidence low, and nothing here was cross-checked against contemporaneous primary records this session. The episode that most directly tested the vote theory — the 2017 user-activated soft fork — falls outside the notes entirely.

Common misstatements: "Miners vote on Bitcoin's rules via signaling" (the corpus's verified mechanics treat signaling as readiness coordination; rule adoption rests with nodes and the economy — see miners-capabilities-limits). "95 percent agreement is required because Bitcoin needs near-unanimous votes" (the threshold was a chain-split safety parameter, not an electoral rule). "A miner who signals has upgraded" (July 2015 demonstrated signaling without validating). "Only miners can express a preference" (node operators express theirs in what software they run, which is the mechanism the vote theory omits).