167 wires and counting

$ follow Go

Keep up with Go in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-10-28
stories 20

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

Raw text rendering fix in golang/net, and faster darwin builders

By RepoJournal · Filed · About Go · Composed from the cited sources · methodology

The day's one consumer-facing change is an html rendering fix in golang/net, alongside infrastructure work that cuts roughly ten minutes off every darwin race build.

html: check all ancestor namespace transitions when rendering raw text golang/net

by Ian Alexander

In childTextNodesAreLiteral, the walk up the tree stopped at the nearest foreign ancestor once an HTML integration point was found, so a subtree foster-parented into a foreign namespace could be re-parsed in the wrong namespace. The renderer now checks all ancestor namespace transitions; treat this as a breaking change for any code depending on the previous raw-text output.

cmd/go: slice ~10 minutes off darwin race builders by adjusting GORACE=atexit_sleep_ms golang/go

by thepudds

cmd/go script tests now run with GORACE=atexit_sleep_ms=100, cutting end-to-end darwin race builder time from about 1800s to about 1100s and cmd/go test time from about 1000s to about 300s. The change exists because the race detector's 1-second default exit sleep was being paid on every test binary.

cmd/cgo/internal/testplugin: remove hard-coded timeout in TestIssue18676 to help slow builders golang/go

by thepudds

The hard-coded 10-second timeout in TestIssue18676 is gone, a test that accounts for roughly half of the amd64 darwin builder failures and about 350 entries in the watchflakes issue. It follows the earlier conversion of other hard-coded test command timeouts to the same pattern.

cmd/dist, make.bash: clear GOWASM when using bootstrap toolchain golang/go

by Carlos Amedee

make.bash, make.bat, make.rc, and cmd/dist's bootstrapBuildTools now clear GOWASM while the bootstrap go command runs and restore it afterward. Previously a newer value like GOWASM=stackswitch was rejected by the GOROOT_BOOTSTRAP toolchain at buildcfg init, before toolchain1 could be built.

internal/worker: skip redis cache invalidation when inserting modules golang/pkgsite

by Hyang-Ah Hana Kim

And in the long tail: pkgsite restored CL 852148, skipping Redis cache invalidation on module insert, after reverting it proved unnecessary, plus podman dependencies reaching the trixie Dockerfile and a new forky Dockerfile in golang/build, an invariant validation pass added to the go/types tests, a loong64 math fix that lets 0.0 be computed from F9-F9 instead of F7-F7, and 2 first party vulner...

Quick answers

What shipped in Go on October 28, 2026?
The day's one consumer-facing change is an html rendering fix in golang/net, alongside infrastructure work that cuts roughly ten minutes off every darwin race build. In total, 20 commits landed.
Who contributed to Go on October 28, 2026?
9 developers shipped this update, including thepudds, Carlos Amedee, Peter Weinberger, XiaolinZhao, Gopher Robot, Michael Matloob, Hyang-Ah Hana Kim, and Ethan Lee, and 1 more.
What were the notable Go updates?
html: check all ancestor namespace transitions when rendering raw text, cmd/go: slice ~10 minutes off darwin race builders by adjusting GORACE=atexit_sleep_ms, and cmd/cgo/internal/testplugin: remove hard-coded timeout in TestIssue18676 to help slow builders.