silverwind's comments

On: Update on "Co-authored-by: Copilot" in commit messages

[silverwind]

It's definitely helpful to know whether a PR was AI-assisted or not and the git attribution line is a simple and effective way of communicating that.

I also recommend specifying model name and version so the maintainer knows upfront the level of slop they are dealing with.

On: Should I run plain Docker Compose in production in 2026?

[silverwind]

I'm going even simpler, just bare docker with a idempotent bash wrapper.

On: I am worried about Bun

[silverwind]

I hope Node eventually gets a WebSocket server like Bun has.

On: I am worried about Bun

[silverwind]

Problem with Go is the type system is rudimentary, so you can't "restrict" AIs as well as you could in Typescript.

On: I am worried about Bun

[silverwind]

https://www.typescriptlang.org/tsconfig/#erasableSyntaxOnly covers them all, I strongly recommend running with that option enabled to be future-proof.

On: I am worried about Bun

[silverwind]

Modern Typescript does not need runtime features.

[moritzwarhier → silverwind]

Your comment might lack explanation, but indeed the TS team has mentioned multiple times that they don't want to add any more features that require transpilation (as opposed to "dumb" type stripping and being a strict superset of JS).

IIRC they "almost" recommend against using them (the last part, I haven't researched again now).

But the usage of many features has reached a sort of point of no return, so I hope Node will go the route of making the experimental transpilation the default for TS files at some point.

Goes to show how strong the appeal of syntax is, especially enums.

To people coming from languages with enum support, it just looks so much more organized to use them, compared to union types, despite all of the (many) drawbacks.

On: Shai-Hulud Themed Malware Found in the PyTorch Lightning AI Training Library

[silverwind]

Maybe now people can stop blaming npm and realize none of these unreviewed package ecosystem are safe.

On: Regression: malware reminder on every read still causes subagent refusals

[silverwind]

Could that be the explanation for the recently increased token use?

On: VibeVoice: Open-source frontier voice AI

[silverwind]

Slopilot

On: GitHub Copilot code review will start consuming GitHub Actions minutes

[silverwind]

This seems nonsensical. Why would non-actions activity consume actions budget?

[semiquaver → silverwind]

  > Last month, we shared how GitHub Copilot code review runs on […] GitHub Actions using GitHub-hosted runners.

They say that they’re now billing against their actual costs

[WorldMaker → silverwind]

Copilot Code Reviews are Actions workflows. Just privileged ones you can't edit the YAML for. They even litter your Actions tab list and Deployment environments.

[horsawlarway → silverwind]

My guess is that they're moving to a spot where they can pitch an LLM "doing something" as an action, and copilot is their first move. I don't see it as crazy to think of a "copilot code review" in a similar way to other build actions.

But also - enterprise accounts already have budget assigned to github actions, and this allows them to start billing right away without having to actually get (or allow) businesses to evaluate the return of having copilot do code reviews.

So seems like it's a mix of immediate incentives and long term architecture. I don't like it, though. If I were an enterprise my first response would be to turn it off.

[swiftcoder → horsawlarway]

> enterprise accounts already have budget assigned to github actions, and this allows them to start billing right away without having to actually get (or allow) businesses to evaluate the return of having copilot do code reviews

Hang on, I read this as copilot reviews with bill both actions minutes and AI credits. Did I miss something?

[zdragnar → swiftcoder]

I'm assuming the running of the model is consuming the tokens, and the client coordinating and orchestrating the calls to the model to perform the review is happening in an action runner, thus using action minutes.

[freedomben → silverwind]

Agreed, especially weird since they just rolled out usage-based billing for Co-pilot. It would make a lot more sense to just re-use that usage instead IMHO

[semiquaver → freedomben]

It does, read the article. This feature now consumes actions credits and AI credits.

[freedomben → semiquaver]

Yeah my mistake, I wasn't very clear in my comment.

Though actually the more I think about it, I think this change actually does make more sense. In the case of the AI running on GitHub side, that does feel pretty equivalent to CI minutes. I would hope that the number of minutes they bill for is pretty minimal though, since the vast majority of that will be I/O waiting on the agent to return

[whalesalad → silverwind]

Code review ostensibly takes place inside a container runtime just like tests or other actions would. It makes sense to me.

[PunchyHamster → whalesalad]

but consumes vastly more resources than most app's build process.

Done that way it obfuscates cost of the code review and I think that's on purpose

[whalesalad → PunchyHamster]

The cost of running a container (to github) is someone else not being able to run a container.

[throwatdem12311 → silverwind]

“F*ck you. Pay me.”

That’s why.