NIST released an initial public draft of SP 800-185 Revision 1 on October 8. It adds streaming interfaces for SHA-3-derived extendable-output functions, allowing incremental input and output without knowing their total lengths in advance. Comments are due December 7.
The draft preserves equivalence with the previous single-call functions. It also identifies a distinction worth reviewing: each TupleHash ABSORB call processes a complete tuple element, whereas cSHAKE, KMAC and ParallelHash can concatenate successive absorbed inputs. NIST specifically asks whether that difference could confuse implementers. This remains a draft, not a final revision.
NAOAT analysis
The concrete implementation issue is whether a software interface distinguishes a piece of data from a complete logical field. Imagine an application receiving a customer identifier and a transaction identifier in several network chunks. A transport boundary is simply where delivery paused; a field boundary conveys structure. An interface that lets a developer confuse those meanings creates an avoidable integration hazard.
A library review could therefore examine the names and types exposed to callers, not just the cryptographic calculation underneath. One useful design question is whether passing a partial field should be impossible, explicitly marked or handled through a separate wrapper. That decision belongs in the API contract, where the application developer will encounter it.
Compatibility work should include examples that present the same intended input using different delivery chunks, plus examples that deliberately change the logical grouping. The expected relationships should be written down before implementation so the test suite captures intent rather than copying the code’s behavior. These are proposed review cases, not a report of flaws in existing products. The comment period gives maintainers an opportunity to identify interface ambiguity while the specification can still change. For organizations using these primitives through dependencies, the immediate action is to follow the relevant library’s review and adoption plan rather than assume that publication of a draft demands an emergency migration.