Part V · 3 — The alternatives (and the closing)
Most of what you want from a blockchain — verifiable immutability, provenance, timestamping — can be obtained without decentralized consensus, at a fraction of the cost. It is what Koder uses. This section closes the compendium.
3.1 Verifiability without blockchain
A blockchain couples two things: verifiability (proving that a record has not changed) and decentralization of consensus (agreeing without a root of trust). Most cases want only the first — and it is cheap without the second:
| Tool | What it provides | Without needing |
|---|---|---|
| Append-only log with a hash chain | each record pins the hash of the previous one — tampering is detectable (the idea from Part I, without a network) | consensus, tokens |
| Verifiable Credentials (W3C) | signed credentials the holder proves | an identity chain |
| Sigstore / Rekor | a transparent, append-only log of artifact signatures | blockchain |
| OpenTimestamps | proof of existence on a date, anchored to Bitcoin only as a notary, at no cost per use | running a node / paying gas |
The key piece is the hash chain (also called a Merkle log / append-only ledger): the same cryptography (Cryptography compendium) that gives a blockchain its immutability, applied to a centralized and fast log. You get tamper-evidence without paying the price of global consensus.
The deciding question: do you need third parties who do not trust you to verify — or do you only need to detect tampering in a record of yours? If it is the latter (the common case), a hash chain solves it, and a database serves as the base.
3.2 The posture, distilled
Koder treats blockchain like any tool in the Stack: by the problem, not by the hype.
- Default — a database (kdb-next/Postgres) + classic cryptography
(
koder_kit). Solves the overwhelming majority of cases. - Internal verifiability — an
append-onlylog with a hash chain.Tamper-evidence without consensus.
- Pointwise public verifiability — Sigstore, OpenTimestamps, Verifiable
Credentials.
- Blockchain — only when the six filters pass:
multiple parties without a root of trust and public verifiability and the performance/regulatory budget adds up.
It is the same principle as self-hosted-first and the discipline of koder_kit in Cryptography: the simplest solution that satisfies the real requirement wins.
3.3 The compendium's closing
From Part I (what it is and the "ownerless" problem) to consensus, to the chains and layers, to what gets built on top, and finally to Koder's posture: the blockchain is a brilliant and expensive answer to a specific problem — agreement without trust. Understanding blockchain deeply includes knowing when not to use it — and that is why this compendium ends not with hype, but with the decision yardstick.
For the dense detail of any point — protocols, L1s, DeFi, incidents, regulation — Part VIII — Reference is the almanac.
End of the narrated arc of the Blockchain Compendium. The five parts go from the why to the how to the posture. With Cryptography (I–V) and Blockchain (I–V) complete, the three compendiums of the genre — AI, Cryptography, Blockchain — have their narrated+illustrated layer live, on top of the reference almanac (stack#196).