What happened

Hydra PR #2825 consolidates onto existing libraries and fixes several bugs. Most notably, it corrects a minimum-fee estimate that omitted Conway’s reference-script charge from a published cost table.

The record gives a close-transaction estimate of 0.6453 ADA after the correction, compared with 0.4532 ADA previously. The upstream record characterizes that gap as an under-reporting of about 30%.

Why it matters

Developers who used Hydra’s earlier benchmark table to model Layer-2 transaction costs may need to re-derive those estimates with the Conway reference-script charge included. The work also reuses the Prometheus registry, removes modern-uri and mono-traversable, and closes a monitoring lookup path that could silently do nothing when metric names were mistyped.

This is maintenance activity on Hydra’s development track rather than a new user-facing feature. It follows another substantive Hydra pull request in the same window, alongside otherwise minor development commits.

What to watch next

Watch for the corrected benchmark figures to appear in the relevant Hydra materials, and for any follow-on development or release work that makes the change easier for operators and builders to verify. Any cost model based on the earlier table should be checked against the corrected estimate.

What to watch

Watch for updated Hydra benchmark material or release follow-through that reflects the corrected 0.6453 ADA close-transaction estimate.

Sources and limits

Upstream references and independent checks

Digest dated 2026-08-22 · upstream model claude-sonnet-4-6. Direct links are matched to all 1 upstream source IDs.

  1. 1
    Some library re-use fixes (#2825)Direct upstream source · b1d5b229846e5673a301b48d6e6e6f743976c14c

This Research brief was generated by Terra from a dated upstream research digest. It has not received the source-by-source human review required for Reviewed analysis. Material limit: This is a single approved source from a development-repository commit, not a release or independent confirmation; the roughly 30% correction is limited to Hydra’s published benchmark table and does not establish a change to live mainnet fees.