Stream: git-wasmtime

Topic: wasmtime / PR #11323 wasi-http: Add default-send-request ...

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 13:30):

brian14708 opened PR #11323 from brian14708:feature/wasi-http-default-send-request to bytecodealliance:main:

This PR adds a default-send-request feature flag to the wasmtime-wasi-http crate that allows disabling the default HTTP request handler implementation and its associated TLS dependencies.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 13:30):

brian14708 requested wasmtime-wasi-reviewers for a review on PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 13:30):

brian14708 updated PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 13:34):

brian14708 updated PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 15:37):

dicej submitted PR review:

Looks great, thanks; just one suggestion inline.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 15:37):

dicej created PR review comment:

Would you mind prefixing each of these with dep: (e.g. "dep:tokio-rustls" per https://doc.rust-lang.org/cargo/reference/features.html#optional-dependencies to avoid possible ambiguity?

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 18:31):

alexcrichton submitted PR review:

(if you don't mind me doing some review too)

Thanks for this! In addition to the comments below can you add a check on CI that the crate builds without default features? For example somewhere around here

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 18:31):

alexcrichton created PR review comment:

FWIW we generally do not test all possible combinations of features, so I think it's reasonable to leave this out and have it be a compile error if the feature is disabled and tested. My thinking is along the lines of my above comment where I'd prefer to avoid reachable always-return-error functions if possible

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2025 at 18:31):

alexcrichton created PR review comment:

Personally I don't feel this is the best way to model the behavior here. Could the crate feature be used to control whether this is is a provided function or not? For example without the crate feature this would be an empty function meaning that callers would be required to implement it. With the crate feature though it would be a provided function meaning it has a default implementation.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 00:11):

brian14708 requested alexcrichton for a review on PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 00:11):

brian14708 requested wasmtime-default-reviewers for a review on PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 00:11):

brian14708 updated PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 00:16):

brian14708 updated PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 00:30):

brian14708 updated PR #11323.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 00:35):

brian14708 submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 00:35):

brian14708 created PR review comment:

Thanks, missed a no-default-features in wasmtime-cli otherwise.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 26 2025 at 03:39):

alexcrichton merged PR #11323.

Last updated: Dec 06 2025 at 07:03 UTC