Activity
Yesterday
dmitshur commented on github.com/go-gl/gl: Jerky refresh with high update rate7h
Does the same behavior reproduce if you try the example programs at [go-gl/example](https://github.com/go-gl/example)? If not, you can try bisecting between the two programs to narrow in on the cause…dmitshur reviewed +2 on cmd/go/internal/bug: sync go bug headings with GitHub form13h
Thanks.dmitshur commented on golang.org/x/build/...: apply omitzero fix14h
(1 comment)
It's actually a minimal client for the k8s API, added back in 2015 (CL 14390). At some point I think it was used more heavily for k8s-based builders. By now I think our only remaining us…dmitshur commented on cmd/fix: fix can print "applied 8 of 10 fixes; 2 files updated" and fail without actually applying any fixes15h
In case it helps, noting here that I found breaking out a single analyzer run to happen first in the dashboard package is a viable workaround:
```sh
go fix -stringsseq ./dashboard # workaround for …dmitshur commented on golang.org/x/build/...: replace slices.Contains-likes with slices.Contains15h
This change is ready for review.dmitshur opened a change golang.org/x/build: -- dividing line after reaching go fix zero diff --15h
At this point, 'go fix ./...' produces zero diff and exit code 0.
The following CLs are follow ups to some of the earlier changes
that were generated. Their scope was intentionally to run a fix
as i…dmitshur opened a change golang.org/x/build/...: apply fmtappendf fix15h
For golang/go#77519.
[git-generate]
go fix -fmtappendf ./...dmitshur opened an issue golang.org/x/build: modernize for Go 1.24+ (via `go fix`)16h
The x/build repo has a good number of opportunities to modernize Go code, some older and some newer. In Go 1.26, [`go fix`](https://go.dev/doc/go1.26#go-command) is improved to make it much easier to…dmitshur opened a change golang.org/dl/internal/version: add explanatory comments to runGo19h
Explain some of the subtle reasons for why runGo behaves as it does.
Do this to reduce chances of something accidentally breaking in the
future.
Based on suggestions Russ made in
https://go-review.g…dmitshur opened a change golang.org/x/build/internal/relui: leave out GOROOT_FINAL when building distpacks19h
As of Go 1.23, the GOROOT_FINAL environment variable is no more¹,
and so it's no longer necessary to set it here. Start gradually
removing it.
Also, delete some code that became dead as of Go 1.24.…dmitshur opened a change golang.org/x/build/cmd/watchflakes: include all failures in "log" field19h
The "log" field is documented as "the entire failed build log" at
https://go.dev/wiki/Watchflakes#records. In practice it was only
the top-level log text, and not the additional failures. Their logs
…dmitshur commented on cmd/go: use AppendPath(OrigEnv) as env for tool executable20h
Filed issue #77509 to document this behavior.dmitshur commented on cmd/go: place GOROOT/bin at the beginning of PATH in 'go run'20h
Filed issue #77509 to document this behavior.dmitshur opened an issue cmd/go: document GOROOT/bin/go PATH entry for `go run` and `go tool`20h
This is a continuation of issue #57050, to add missing documentation for a cmd/go behavior that's already implemented for a few more of its commands. ([CL 461683](https://go.dev/cl/461683) previously…This Week
dmitshur commented on github.com/shurcooL/githubv4: Maintenance status and future plans1d
Thanks for offering to help.
I wouldn't describe its maintenance levels as very active, but I do still use it myself and ocassionally update it to include latest GitHub GraphQL schema changes. That'…dmitshur pushed to
main in github.com/shurcooL/graphql1ddmitshur closed a change github.com/shurcooL/githubv4: regenerate for schema changes by 2022-10-241d
Thanks. These schema changes are already merged via PR https://github.com/shurcooL/githubv4/pull/110.dmitshur commented on github.com/shurcooL/githubv4: regenerate for schema changes by 2022-10-241d
Thanks. These schema changes are already merged via PR https://github.com/shurcooL/githubv4/pull/110.dmitshur commented on github.com/shurcooL/githubv4: Regenerate for schema changes by 2025-11-211d
Thanks for the PR. I prefer to update go.mod and go.sum files separately.
I pushed commit 2402fdf4a9edcf51ed3f2eab4dc57c6867b10eea that includes schema changes by 2026-02-08. It doesn't yet includ…dmitshur pushed to
main in github.com/shurcooL/githubv41ddmitshur commented on golang.org/x/example/gotypes: update for go/types.Alias 1d
[CL 620075](https://go.dev/cl/620075) updated the gotypes tutorial.dmitshur closed an issue runtime: TestRuntimeLockMetricsAndProfile/runtime.lock/sample-1 failures1d
dmitshur closed an issue cmd/go/internal/bug: sync issue body format with GitHub template1d
Last Week
dmitshur commented on golang.org/x/build: add LUCI linux-mipsle builder3d
Adding a note here that the exec format error on cipd_cache/bin/cipd is likely the same problem that was encountered in https://github.com/golang/go/issues/67306#issuecomment-3633590314. If so, you m…dmitshur commented on net/http: fix WaitGroup race in TestTransportNoReuseAfterEarlyResponse3d
Thanks.dmitshur opened a change golang.org/x/build/cmd/gopherbot: close reopened cherry-pick issues3d
Maintain the property that gopherbot doesn't enter a non-stop fighting
match with a human trying to reopen a cherry-pick issue, but still let
the submission of newer backport CLs re-close said issues…dmitshur opened an issue cmd/go/internal/bug: sync issue body format with GitHub template (as of February 2026)3d
[CL 553155](https://go.dev/cl/553155) reordered headings in the GitHub issue templates to be "got, want" (for issue #64890), but the issue template used by `go bug` still has the old inverse "want, g…dmitshur commented on cmd/go: add `go env -bug-report`3d
I agree that `go env -changed` cannot fully replace `go env`, but it can be a useful addition in many cases. Go's CI system prints all 3 of `go env`, `go env -changed`, and even `cmd/internal/metadat…dmitshur reviewed +2 on golang.org/x/website/_content/doc/go1.26: minor adjustments3d
Thanks.dmitshur reviewed +2 on golang.org/x/build/go.mod: upgrade github.com/fflewddur/ltbsky3d
Thanks.
Needs a go mod tidy, I suspect.
I really can't tell which is less expensive: leaving this unresolved comment that gets in the way of this auto-submitting, or leaving it for a future CL at t…dmitshur commented on crypto/tls: [release-branch.go1.26] avoid data race when canceling a QUICConn's Context3d
Adding TryBot-Bypas since only go1.26-linux-amd64-longtest had a seemingly unrelated TestTraceGCSTW failure, and after being restarted it passed.dmitshur reviewed +2 on go.dev/wiki/GerritAccess: clarify encouragement to vote Auto-Submit+13d
Thanks very much.
Optional: maybe "new exported API" is better than "new API"; it doesn't add much verbosity but I think may help solidify that this isn't talking about internal API changes too.
> …dmitshur commented on golang.org/x/build/internal/relui: fix BaseLink to keep RawPath in sync with Path3d
Thanks. I can redeploy relui with this now.dmitshur opened a change log/slog: [release-branch.go1.26] correct method names in MultiHandler docs3d
For #77428.
Fixes #77429.dmitshur opened an issue cmd/fix: fix can print "applied 8 of 10 fixes; 2 files updated" and fail without actually applying any fixes3d
### Go version
go version go1.26rc3 darwin/arm64
### Output of `go env` in your module/workspace:
```shell
AR='ar'
CC='clang'
CGO_CFLAGS='-O2 -g'
CGO_CPPFLAGS=''
CGO_CXXFLAGS='-O2 -g'
CGO_ENABLED=…dmitshur commented on access: approvers4d
Hi @HeliC829, you recently sent a GitHub request to join the [golang/riscv64](https://github.com/orgs/golang/teams/riscv64) maintainer team. As suggested at https://go.dev/wiki/PortingPolicy#requirem…dmitshur reviewed +1 on github.com/fflewddur/ltbsky: Avoid logging arbitrary length byte slice4d
Thanks.dmitshur fixed an issue golang.org/x/vulndb: TestLintReports/data/reports/GO-2026-4336.yaml failures4d
Fixed in [go.dev/cl/741780](https://go.dev/cl/741780).dmitshur fixed an issue golang.org/x/vulndb: TestLintReports/data/reports/GO-2026-4336.yaml failures4d
Fixed in [go.dev/cl/741780](https://go.dev/cl/741780).dmitshur fixed an issue golang.org/x/vulndb: TestLintReports/data/reports/GO-2026-4335.yaml failures4d
dmitshur fixed an issue golang.org/x/vulndb: TestLintReports/data/reports/GO-2026-4335.yaml failures4d
Fixed in [go.dev/cl/741780](https://go.dev/cl/741780).dmitshur commented on golang.org/x/vulndb: TestLintReports/data/reports/GO-2026-4335.yaml failures4d
Fixed in [go.dev/cl/741780](https://go.dev/cl/741780).dmitshur fixed an issue golang.org/x/vulndb: TestLintReports/data/reports/GO-2026-4334.yaml failures4d
Fixed in [go.dev/cl/741780](https://go.dev/cl/741780).dmitshur commented on golang.org/x/vulndb: TestLintReports/data/reports/GO-2026-4334.yaml failures4d
Fixed in [go.dev/cl/741780](https://go.dev/cl/741780).dmitshur commented on golang.org/x/build/cmd/watchflakes: factor log.Fatal out of LUCIClient.GetBuildResult4d
Thanks.dmitshur commented on golang.org/x/build/cmd/watchflakes: update ResultDB host for January 2026 change4d
Thanks.dmitshur opened a change golang.org/x/build/cmd/watchflakes: factor log.Fatal out of LUCIClient.GetBuildResult4d
Make some progress on removing log.Fatal calls from inner functions,
in favor of returning an error to the caller and doing log.Fatal in
main. This will help eventually retry more retryable errors wi…dmitshur opened a change golang.org/x/build/internal/relui: fix BaseLink to keep RawPath in sync with Path4d
The custom functionality to prepend an optional base URL to a target URL
implemented in BaseLink involves modifying the Path field. To avoid
disrupting what the EscapedPath method returns, it's neces…dmitshur commented on golang.org/x/build/internal/task: use authentication for Gitiles in RealGerritClient4d
Thanks.dmitshur commented on golang.org/x/build/cmd/relui: automate upstreaming private-track security fixes on release day4d
Thanks.
So far we had a chance to try out the "nothing to upstream" path and that worked without uncovering anything new. We'll try the "1+ CL to upstream" path when it comes up next and adjust as n…dmitshur commented on golang.org/x/pkgsite-metrics/...: add codereview.cfg5d
Thanks.dmitshur commented on golang.org/x/vuln/...: add codereview.cfg5d
Thanks.dmitshur commented on golang.org/x/pkgsite/...: add codereview.cfg5d
Thanks.dmitshur commented on golang.org/x/website/_content/blog: add Go 1.26 release post5d
(2 comments)
The Go 1.25 blog post didn't link to https://go.dev/doc/go1.25#language because there weren't notable language changes, but I agree it's worth at least including a link to https://tip.g…dmitshur commented on rsc.io/rf/...: allow for custom go executable5d
Thanks for considering this.
As mentioned in https://github.com/rsc/rf/issues/16#issuecomment-812623872, I think it's generally expected that tools working on packages in GOROOT need to use the co…dmitshur commented on golang.org/x/build: add LUCI aix-ppc64 builder5d
As I understand that PR applied the fix to the v4 major version of badger, but cas currently uses v3. That means either the badger fix needs to also be backported to its v3, and cas updated to that, …dmitshur commented on rsc.io/rf/...: allow for custom go executable5d
I wanted to ask what's the motivation to add a flag for this? There have been proposals in the past in the main Go project to add alternative ways to specify the go command to use, such as golang/go#…dmitshur reviewed +1 on github.com/fflewddur/ltbsky: Extend error handling5d
Looks good, one inline comment.
"%+v" on `img` as is might create excessively verbose errors due to its [`Bytes []byte`](https://github.com/fflewddur/ltbsky/blob/9a65bc6aa513f18bb638c208c9187f8da33c…dmitshur commented on rsc.io/rf: rf fails with spurious type errors when refactoring projects with complex dependencies5d
Thanks for filing this issue and sending a fix @jitsu-net!
I tried doing a brief refactor in x/build again with the latest rf version, and can confirm it worked without errors this time. So I think …dmitshur fixed an issue rsc.io/rf: mv broken5d
dmitshur commented on golang.org/x/website/_content/blog: add Go 1.26 release post5d
(1 comment)
As far as I can tell both [crypto/mlkem/mlkemtest](https://pkg.go.dev/crypto/mlkem/mlkemtest?tab=versions) and [testing/cryptotest](https://pkg.go.dev/testing/cryptotest?tab=versions) pa…dmitshur reopened an issue golang.org/x/build/cmd/relui: simplify HTTP route handling code by using net/http's pattern matching5d
dmitshur reopened an issue golang.org/x/build/cmd/relui: simplify HTTP route handling code by using net/http's pattern matching5d
dmitshur commented on golang.org/x/build/cmd/relui: simplify HTTP route handling code by using net/http's pattern matching5d
> Also noting here that some of the route registration uses task names, which we've recently realized may contain "/" characters, and those will need some thought to handle.
This was intended to be …dmitshur commented on github: switch "what did you expect to see" and "what did you see instead"5d
CC @adonovan.dmitshur commented on time: document that Ticker.Reset can be called after Stop6d
(1 comment)
The fact the channel is reused seems more like an internal implementation detail that is secondary to the fact that calling Reset after Stop is permitted. Perhaps it'd be better not to m…dmitshur reviewed +2 on time: mention Reset in Ticker.Stop doc6d
Thanks.dmitshur commented on golang.org/x/website/_content/doc/go1.26: fix typo in slog.MultiHandler doc6d
(1 comment)
```suggestion
Its `Handle`, `WithAttrs` and `WithGroup` methods call the corresponding method
```dmitshur reviewed +2 on golang.org/x/website/_content/doc/go1.26: fix typo in slog.MultiHandler doc6d
Thanks.dmitshur reviewed +2 on log/slog: correct method name in MultiHandler docs6d
Thanks.
Let's also fix WithAttrs since it's nearby.
```suggestion
// Its Handle, WithAttrs and WithGroup methods call the corresponding method on each of the enabled handlers.
```dmitshur reviewed +2 on golang.org/x/website/internal/history: document Go 1.25.7 and 1.24.136d
Thanks.dmitshur reviewed +2 on golang.org/x/build/env/freebsd-amd64: switch to ufs image6d
Well, it was worth a try. Thanks.dmitshur commented on golang.org/x/build/main.star: remove linux-amd64-spectre known issue6d
Just checking, is it expected this builder would not have caught the issue reported in #77420, but rather it's intended to check for easier-to-catch problems that may come up in the future?
Asking b…dmitshur reviewed +2 on golang.org/x/build/main.star: finish upgrading freebsd-amd64 to FreeBSD 14.36d
Thanks.
A single passing commit across all repos seems pretty good signal for a 14.2 -> 14.3 change, so I suspect it's fine to move ahead without much waiting.dmitshur reviewed +2 on golang.org/x/build/main.star: add freebsd-amd64_15.0 builder6d
Thanks.dmitshur reviewed +2 on golang.org/x/build/main.star: add spectre builder6d
Thanks. It'll be interesting to see the 'go env -changed' output on the builder.dmitshur reviewed +1 on internal/runtime: fix assembly for spectre retpoline instrumentation6d
(1 comment)
Thanks for catching that and improving their name. Including it in this CL seems fine.dmitshur reviewed +2 on golang.org/x/build/env/freebsd-amd64: add 15.06d
dmitshur reviewed +2 on golang.org/x/build/env/freebsd-amd64: add 15.06d
Thanks.
Not a big deal to leave these on older version. That said, if you want, consider updating them:
```suggestion
./make.bash 15.0
After it completes, it creates a file freebsd-amd64-150.t…dmitshur reviewed +2 on golang.org/x/build/env/freebsd-amd64: don't restart buildlet on exit6d
Thanks.dmitshur commented on crypto/tls: revalidate whole chain on resumption on Windows and macOS6d
(1 comment)
I might've overlooked something, but to check here, can this condition be true? Since i is ranging over chain[1:], at [first glance](https://go.dev/play/p/5ZViTX7tvUY) it seems it might …dmitshur commented on golang.org/x/build: bot linux-s390x-ibm--02 reported as broken6d
Thanks for looking into this.
From looking at the list of bots connected to LUCI at https://chromium-swarm.appspot.com/botlist?f=cipd_platform%3Alinux-s390x&f=pool%3Aluci.golang.shared-workers, it d…dmitshur commented on golang.org/x/build: add LUCI linux-mipsle builder6d
Thanks for requesting to add a builder to test the linux/mipsle port (CC @golang/mips). There's also issue #67304 tracking this. Is this meant to be an additional builder?
CC @golang/release, @pratt…dmitshur commented on os: File.Close hangs with concurrent I/O on Windows using go1.26rc2 [1.26 backport]6d
CC @prattmic.dmitshur reviewed +2 on golang.org/x/mobile/bind: support go/types.Alias1w
Looks good to me, though the go/types API and best practices for code that needs to handle aliases isn't something I've looked into closely. Someone more familiar (@adonovan@google.com maybe) might b…dmitshur reviewed +2 on golang.org/x/build/env/freebsd-amd64: add 14.31w
Thanks.dmitshur reviewed +2 on golang.org/x/build/env/freebsd-amd64: add doas configuration1w
Thanks.dmitshur reviewed +2 on golang.org/x/build/main.star: add freebsd-amd64_14.3 builder1w
Thanks.dmitshur reviewed +2 on golang.org/x/build/main.star: remove freebsd-amd64_14.1 builders1w
Thanks.
I think I tried to use the 14.1 builders to iterate a bit on that issue with the builders not cleaning up after themselves on termination requests or so. Does that work fine now?
If not, ma…dmitshur commented on golang.org/x/website/internal/web: do not treat Markdown files as templates by default1w
Missed this thread, so marking it as resolved too. Thanks again.dmitshur reviewed +2 on golang.org/x/website/internal/web: do not treat Markdown files as templates by default1w
Thanks.
Thanks. For reference, in case it comes in handy later, note that there's a bit more information about tests for x/website/cmd/golangorg in its [README](https://cs.opensource.google/go/x/web…dmitshur commented on internal/poll: readWriteUnlock should destroy fd when no remaining references1w
Minor comments on the test; otherwise leaving this to Damien to review. Thanks.
https://pkg.go.dev/testing#T.Fatal's "FailNow must be called from the goroutine running the test or benchmark function…dmitshur commented on cmd/go: add 'go test -cachelink' boolean flag1w
@bradfitz@golang.org Please note that the Run-TryBot+1 by now is meant only when looking to trigger legacy slowbots. Otherwise there aren't any trybots that run in its default set. So in most CLs lik…dmitshur commented on net: don't wait 5 seconds to re-read /etc/resolv.conf1w
There weren't any legacy slowbots requested, so Run-TryBot+1 isn't doing anything except blocking submission. Dropping it.Earlier
dmitshur commented on cmd/go: repository URL resolution behavior doesn't match documentation1w
CC @golang/command-line.dmitshur commented on golang.org/x/tools/go/analysis/passes: fix buf.WriteString(x + y) to the more efficient buf.WriteString(x); buf.WriteString(y)1w
CC @adonovan.dmitshur commented on cmd/go: `go tool` and `go get -tool` could be documented better1w
CC @golang/command-line.dmitshur commented on golang.org/x/website: code in README.md is excessively HTML escaped1w
@qrasick I should be able to review it by tomorrow. From an early look, it looked good. Thanks for preparing it.dmitshur commented on doc: write release notes for Go 1.261w
The Go 1.26 release notes draft seems ready. It's okay for this issue to stay open to track any further refinement and improvement leading up to the release, but removing the release-blocker label si…