ENG-6268: unify HTTP stack on http-1.x (reqwest 0.12, httptest 0.16) + dep dedup - #95
Open
imp wants to merge 2 commits into
Open
ENG-6268: unify HTTP stack on http-1.x (reqwest 0.12, httptest 0.16) + dep dedup#95imp wants to merge 2 commits into
imp wants to merge 2 commits into
Conversation
…per 0.14) reqwest 0.11/httptest 0.15 still pulled in the old hyper-0.14/http-0.2 stack alongside axum 0.8's hyper-1.x/http-1.x, duplicating the HTTP stack in the dependency tree. reqwest 0.12 and httptest 0.16 are both http-1.x based, so the workspace now resolves to a single unified stack (confirmed via `cargo tree -i http@0.2` / `cargo tree -i hyper@0.14`, both now empty). reqwest's "trust-dns" feature was renamed to "hickory-dns" upstream (trust-dns-resolver renamed to hickory-dns) - updated accordingly. No other code changes needed; both bumps were source-compatible.
- tower-http: 0.7 -> 0.6, matching reqwest 0.12's hard-pinned "0.6.5" requirement (non-optional, unconditional dep of reqwest itself) - 0.6 already covers everything our own TraceLayer/classify usage needs, so this fully collapses the duplicate the reqwest 0.12 bump introduced - toml: 0.7 -> 0.8 in server (only consumer, single trivial to_string_pretty call site), matching what `config` already needs - thiserror: transport and client now use the workspace's "2.0" instead of pinning "1" directly Checked the full `cargo tree --duplicates` list for anything else easily alignable: everything else remaining (base64, bitflags, hashbrown, rand 0.8/0.9, rustls, syn, webpki-roots, etc.) is forced by third-party deps (arrow, polars, sqlx, hickory-dns, config's pest dep) that don't share a common version we could converge on - not fixable from our side without those upstreams moving.
imp
force-pushed
the
imp/plateau-reqwest012
branch
from
July 20, 2026 15:36
244d2de to
2e38590
Compare
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.
Summary
Stacked on #94 (axum 0.8 migration).
trust-dnsreqwest feature renamed tohickory-dns(upstream rename). Confirmed viacargo tree -i http@0.2/cargo tree -i hyper@0.14- both now empty, single unified stack.rustls-platform-verifier(real behavior change for containerized deployments),query/formfeatures now opt-in, and it's only ~2 months old. 0.12 already achieves the actual goal (single http-1.x stack) without those open questions.tower-http = "0.6.5"(non-optional), so bumping reqwest reintroduced a duplicate. 0.6 covers everything our ownTraceLayer/classifyusage in server needs, so downgrading collapses it back to one version.to_string_prettycall site), matching whatconfigalready needs.2.0instead of pinning1directly.Checked the full
cargo tree --duplicateslist for anything else easily alignable - everything else remaining (base64, bitflags, hashbrown, rand 0.8/0.9, rustls, syn, webpki-roots, etc.) is forced by third-party deps (arrow, polars, sqlx, hickory-dns,config'spestdep) that don't share a common version we could converge on - not fixable from our side without those upstreams moving.Test plan
cargo check --workspacecargo clippy --workspace --all-targets(clean except 2 pre-existing warnings in catalog, unrelated)cargo fmt --all -- --checkcargo test --workspace(all green except the same pre-existing machine-local port-3030 flake noted in ENG-6268: migrate plateau to axum 0.8 #94)