[go-nuts] GC hint for collection timing

I recently recalled that someone in the Go Team—I forget who—said that the Go compiler slowed down a lot after converting from C to Go because the Go GC was freeing a lot of memory that the C code wasn't. I wonder if an approach like that of memory regions (https://github.com/golang/go/discussi

[go-nuts] Re: Go Module Mirror

Looks like the package is still in the proxy, and sadly is used by one known person. It would be useful if the proxy site had a tamper warning at the top of a package’s page when the code hash for the version has changed. Perhaps it would be useful to list all the tampered packages in a master

[go-nuts] Is there a way to browse the stdlib packages in pkg.go.dev?

To be clear, I mean see a listing of all the packages in the stdlib, or at least the root package directories. Thanks, Will -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send a

Re: [go-nuts] Looking for Go module path (contains version) suggestion for github.com/spdx/spdx-go-model

Meng, It's unclear to me what you're asking. Are you asking whether the path should be changed to be more idiomatic? I would have expected it to be github.com/spdx/spdx-go-model if the major version is 0 or 1; and github.com/spdx/spdx-go-model/v3 if the major version is 3. Will On Thursday, J

Re: [go-nuts] "go-import" is great, we need "go-delete"

Jeffrey, If modules worked like you suggested, a domain owner could be compromised, and the cryptographic module hashes for the domain could be changed to disrupt or infect dependencies. Version immutability prevents that: a module version is code + current import path ownership at the time the

Re: [go-nuts] Golang ORM Performances

https://github.com/ent/ent seemed to perform well, and was quite flexible. Will On Thursday, December 19, 2024 at 7:27:46 AM UTC-8 Robert Engels wrote: > I go back and forth on ORMs. I think a lot depends on the complexity of > the project. > > Still, I wouldn’t expect the overhead of an ORM t

[go-nuts] Why doesn't Go use Windows's time zone data for time.LoadLocation?

I read in a recent Go issue that Go uses time zone data in GOROOT for Windows and Plan 9. I looked through the Go proposal documents, but couldn't find one for the time package that had the original reasons, but I assume the reasons are because Windows didn't used to have IANA-compatible time

[go-nuts] Why isn't the memory model part of the spec?

It seems required to understand how Go programs work. If the memory model was never written, or is omitted from an implementation, would the concurrency features still be reliable and useful? -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To u

[go-nuts] go env doesn't seem to work well with GOTELEMETRY

Despite showing up an an env var, it's not treated consistently as one: ❯ go telemetry off # nothing printed # as expected ❯ go env | grep GOTELEMETRY= GOTELEMETRY='off' # as expected ❯ go env GOTELEMETRY off # as expected ❯ go env -changed # nothing printed # as expected ❯ go env -u GOTELEMET

[go-nuts] go doc broken in official go repo

I cloned the official Go repo locally, went into src/text/template, ran `go doc .`, and got this error: go: downloading go1.23.0 (darwin/arm64) go: download go1.23.0 for darwin/arm64: toolchain not available It looks like src/go.mod declares 1.23. Shouldn't it be the latest self-host version, w

Re: [go-nuts] Go import redirect docs are unclear

>Most web servers automatically serve an `index.html` for a request to a directory. The intent is to use something like that. Though note that you can also use a fuller path: That is, if your git-repo is at `https://code.org/r/exproj` and has import path `example.com/exproj`, you can host a sin

[go-nuts] go clean -h isn't helpful

❯ go clean -h usage: go clean [clean flags] [build flags] [packages] Run 'go help clean' for details. This just tells me to invoke another help command. The flags package has the opinion that command help should print the doc for flags. Shouldn't we do that for go clean -h too? -- You received

[go-nuts] Unused local constants not a compiler error

Was this always the case? Is this a bug? Seems like it should be a compiler error, like unused local variables. -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to g

Re: [go-nuts] comment.Printer.DocLinkURL isn't being called

ations entirely, so `[MyType]` doc syntax is rendered in Markdown as just `MyType` instead of `\[MyType]`. Is there a way to do that? On Sunday, July 9, 2023 at 5:55:02 PM UTC-7 will@gmail.com wrote: > Thanks, that was it! > > On Sunday, July 9, 2023 at 5:10:08 PM UTC-7 Sean Liao wrote:

Re: [go-nuts] comment.Printer.DocLinkURL isn't being called

Thanks, that was it! On Sunday, July 9, 2023 at 5:10:08 PM UTC-7 Sean Liao wrote: > your parsing setup is missing the lookup to identify which things are > actually linkable symbols > > https://go.dev/play/p/sDtGA_JJZdz > > - sean > > On Sun, Jul 9, 2023, 23:45 w

[go-nuts] comment.Printer.DocLinkURL isn't being called

I'm trying to print the Markdown for package documentation using the new go/doc/comment package. It seems like I have everything set up correctly to customize the doc link URLs, but the comment.Printer.DocLinkURL callback isn't being called: https://go.dev/play/p/7irKc6dTRRw The comment.Printe

[go-nuts] Re: Package structure & testing

>Collapse all three. This works but I really don't like having Fake code in the same package as real code. This is the answer, in my opinion. Don't split code across packages unless there's a reason to do so. A fake implementation is a feature. If it hurts that much, just call it something more

[go-nuts] Show your support for GitHub adding Gerrit features

https://github.com/github-community/community/discussions/18879 It would be nice if the Go project could do reviews on GitHub too. -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it,

Re: [go-nuts] Add comparisons to all types

Checking if they are equivalent raises difficult issues >>> around recursion, slices that point to themselves, and other problems >>> that prevent a clean, efficient solution. >>> >>> Believe me, if equality for these types was e

[go-nuts] Add comparisons to all types

All types should have unrestricted comparisons (`==`, `!=`), but a few pre-declared types don't. Adding them would bridge a semantic gap between pre-declared and user-declared types, enabling all types to be used as map keys, and otherwise make reasoning about them more consistent and intuitive.

[go-nuts] Slice and map type sets for range operations

Do slices and maps have compatible type sets regarding the range operator? I was trying to iterate through either, where slice keys are the indexes, such that this would work: type KV[K comparable, V any] interface { ~[]V | map[K]V } func f[KV2 KV[K, V], K comparable, V any](kv KV2) { f

[go-nuts] What is the Go build cache for fuzz testing?

`go help clean` says: *The -fuzzcache flag causes clean to remove files stored in the Go buildcache for fuzz testing. The fuzzing engine caches files that expandcode coverage, so removing them may make fuzzing less effective untilnew inputs are found that provide the same coverage. These f

[go-nuts] Alternative generics design with some interesting aspects

Hey all, Congrats to the Go Team for shipping generics! I was experimenting with a design for generics for a long time. I thought I'd share it anyway for fun. I think it has some interesting (IMHO) aspects and trade-offs compared to the design that shipped: - Type variable declarations are