Repository navigation
Reduce table prefetch CPU while preserving intent-based navigation - #28
Merged
Merged
Conversation
droplister
marked this pull request as ready for review
September 19, 2026 01:35
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Dense record tables start server renders for address and transaction links as soon as they enter the viewport. A 25-second production sample showed that prefetches accounted for 85% of address requests and 63% of transaction requests.
Preload those record links on hover, keyboard focus, or touch start, once per chosen href. Keep Next link navigation, fresh API reads, polling, and the existing visual layout. Changed destinations after pagination can preload independently.
The same local production-build Sends-table scroll exercise went from 1,184 automatic address/transaction prefetch requests to zero. Total automatic prefetch requests fell from 1,226 to 68; the chosen address still preloaded on hover. This is a controlled browser measurement, not a claimed reduction in the Cloudflare bill. One click-to-heading observation was 373 ms before and 378 ms after; production latency needs post-deploy measurement.
Validation: root checks; 20 existing production-browser E2E/accessibility tests; 4 new production-only prefetch/navigation tests covering hover, keyboard, mobile touch/tap, pagination, and API-error navigation; full OpenNext build; Wrangler dry run; desktop screenshot comparison and mobile overflow check. CI now runs the prefetch tests after its production build. Measurement details and rollout checks are in
docs/frontend-prefetch-budget.md.Production deployment remains pending owner approval. Current rollback version:
f8c44226-a805-4e1d-9db2-cc82e4a7cbfe.Release scope: the web Worker was last deployed September 7 at 18:16 UTC. Main also contains the later, already-merged quantity-precision fix
52e60a2, which is not live: both fixture-driven fairmint display checks still show rounded quantities on xcp.io and pass in this candidate. Deploying the main-based candidate includes that existing fix as well as this PR; no new quantity-format changes are part of this PR.