Damien Neil

This seems reasonable. Actually, I'm surprised debug_redact was added without Go support--it seems like an oversight. I'd say that: - prototext.MarshalOptions should have a field controlling whether redaction is applied....

> Extending interfaces is notoriously not backwards-compatible. Instead, if one wants to expand an interface, it’s best to use a separate interface that embeds the original interface (like fs.ReadDirFile, fs.ReadDirFS,...

Hard to reproduce this. Pretty sure the root cause is as follows: When sending a request, we send a Content-Length header if we know the request body's size, and send...

FYI, it won't help you right now but https://go.dev/issue/67002 may be of interest. This is a proposal (which I'm hoping to land in Go 1.24, although no promises) to add...

I'm having difficulty reproducing this (not aided by my main Windows machine being a gomote which is only accessible via ssh). Any tips on reproducing would be helpful, especially ones...

Okay, I wrote that and naturally managed to finally reproduce the problem immediately afterwards. The issue seems to be that while the unprivileged user can delete a file in `c:\windows\temp`,...

Except in this case, the problem is that the user can't open the *parent* directory of the file. The user can remove `c:\windows\temp\foo`, but they can't open `c:\windows\temp`. I'm not...

+1 For historical context (assuming my memory holds): protoc passes code generator parameters in the CodeGeneratorRequest message. It used to take parameters from the `--_out` flag, splitting the flag on...

> ```Missing output directives.``` I'm not certain, but I'm going to guess that your search-and-replace removed the `--go_out` directive entirely. What you'd want to change is (for example) `--go_out=module=google.golang.org/protobuf,annotate_code:/path/to/output` into...

The assignment rules already cover the case where the same value (or part of a value) appears multiple times in an assignment. > The assignment proceeds in two phases. First,...