Block Timestamps vs Event Ordering
Order blockchain records using execution positions and stable tie-breakers, rather than assuming timestamps uniquely identify or sequence every event.
Overview
When translating these concepts into an integration, keep the scope described by the wallet history api with the result. The examples in wallet transaction history and smart contract event logs show how the same distinction affects neighbouring data questions.
Many blockchain events share one block timestamp. Sorting by time alone cannot reliably establish their execution order or provide a stable pagination boundary.
Use chain position for deterministic order
Within an EVM chain, accepted block position, transaction index and log index can define event ordering under the selected evidence model. Network identity remains essential. Across different chains, timestamps do not create a universal causal order. Avoid deriving a strict cross-chain sequence from clocks that have different rules and precision.
Worked example: equal timestamps
Two transfers occur in one block with the same timestamp. A timestamp-only sort can reverse them between requests. Adding transaction and event positions produces a stable order for that accepted block. A date-filtered report can still use timestamps, but its internal row order should retain deterministic tie-breakers.
Test pagination and causality
Include several events in one transaction and a reorganisation that replaces a block. Preserve the accepted block hash alongside order fields. A display sequence is not proof that one independent transaction caused another. Where cross-chain workflow causality matters, use protocol message relationships instead of inferring it from whichever timestamp is smaller.
A practical acceptance test
Give three events identical timestamps but different transaction and log positions. Reverse their input arrival order and assert the same accepted-chain display order. Add a second chain with an earlier timestamp but no causal relationship. The interface may sort it visually, but must not infer that it caused or preceded another workflow stage logically.
Questions about the topic
Frequently asked questions
Can a timestamp uniquely identify an event?
No. Many events share timestamps.
Do earlier timestamps prove cross-chain causality?
No. Use explicit protocol relationships and evidence.
Standards & technical documentation