Hash chains vs blockchains
Both link records with hashes. Only one needs a network, consensus, and a coin.
loomseal.com · Reading
Both link records with hashes. Only one needs a network, consensus, and a coin.
loomseal.com · Reading
Strip both to their skeletons and they are the same data structure: records linked by digest, where each entry commits to the one before it, so any alteration cascades forward and becomes detectable. That linked-digest core is decades older than any cryptocurrency, and it is the part doing the tamper-evidence work in both.
A blockchain is a hash chain plus machinery for a specific, hard situation: many mutually distrusting writers and no operator everyone trusts. That machinery is consensus, replication across many nodes, and usually an incentive token to keep strangers honest, and it is expensive in every dimension: latency, throughput, cost, operational complexity, and explanation. When the situation genuinely is many distrusting writers, the price can be worth paying. When it is not, the machinery is ceremony.
Ask one thing: who writes to this log? If the answer is one operator, and the problem is that readers do not trust that operator, consensus among strangers solves nothing, because there is exactly one writer to reach consensus with. What readers need is tamper evidence plus a way to catch the operator rewriting history, and that second part does not require a network. It requires anchoring.
Publish the chain's head, one short hash, somewhere the operator cannot rewrite: a timestamp authority under RFC 3161, a commit in a public git repository, a transparency log. Do it on a schedule. A rewritten history now has to disagree with heads that were published outside the operator's control before there was any reason to lie, which means a forger needed to be forging from the beginning, in public, continuously. That is the property most projects reaching for a blockchain actually wanted, at roughly zero cost and with nothing to operate.
Genuinely multi-party writes with no trusted operator and value moving between strangers: that is the honest home turf, and public chains serve it. What earns the skepticism is the private blockchain deployed inside one company, which is a database with extra steps, and the compliance project that says blockchain when it means tamper-evident, which is buying consensus to solve a problem consensus does not address.
LoomSeal takes the boring side of this trade on purpose. Its design rules fix SHA-256, HMAC-SHA256, and ed25519, no consensus, no tokens, no network dependency, and put the external fixity in anchors: timestamp authorities, public git, transparency logs. One file, verified offline by a free tool, with the rules public in the spec. The chain does the tamper evidence, the anchors do the part people thought they needed a blockchain for.