@@ -4,23 +4,18 @@ pub(crate) const OPTION_TOMBSTONE_MARKER: [u8; 2] = [0xFF, 0xFE];
44/// # Namespaced Prefixes for Storage Features
55///
66/// These prefixes are **not** used to differentiate values themselves;
7- /// the storage engine already handles type and structure differentiation.
7+ /// SIMD R Drive already handles type and structure differentiation.
88///
99/// Instead, these prefixes are used to distinguish **storage features** such as:
1010/// - **Option handling** (explicit tombstones for `None` values)
1111/// - **TTL-based auto-eviction** (keys prefixed with expiration timestamps)
1212///
1313/// By applying **feature-based** prefixes, we ensure that:
14- /// - Different storage mechanisms do not conflict.
15- /// - Read /write operations apply the correct logic.
14+ /// - Different feature extensions do not naturally conflict.
15+ /// - Per-extensions read /write operations apply the correct logic.
1616/// - Keys remain distinct even if their raw values are identical.
1717///
18- /// ## Binary Prefix Design:
19- /// - **Start Boundary (`0xF7`)** → High UTF-8 non-standard range, unlikely in normal data.
20- /// - **Feature Name** → Encodes the purpose of the key (e.g., `"option"`, `"ttl"`).
21- /// - **End Boundary (`0xFD`)** → Another rare high-range byte for closure.
22- ///
23- /// This ensures **safe, efficient, and collision-free feature separation**
18+ /// This ensures relatively safe, efficient, and collision-free feature separation
2419/// without interfering with the actual stored values.
2520macro_rules! namespace_prefix {
2621 ( $name: expr) => { {
0 commit comments