No idea why this comment is getting downvoted so hard. This was exactly what I thought of too, and it provides a concrete answer to the question.

There’s valid concern with these types of laws and scope creep. But there’s also precedent which shows they can work and be applied reasonably.

Further to this, the quality problem is affecting the entire industry, not just FOSS. Anyone working on a large enough team has already seen some contributors pushing slop.

And while banning AI outright is certainly an option at a private company, it also feels like throwing out the baby with the bath water. So we’re all searching for a solution together, I think.

There was a time (decades ago) when projects didn’t need to use pull requests. As the pool of contributors grew, new tools were discovered and applied and made FOSS (and private dev) a better experience overall. This feels like a similar situation.

Projects back then did not use pull requests because that is a concept created by GitHub.

Apologies, I forgot to use the more general term.

FOSS projects didn’t always have a standard process for Fagan inspections.

https://en.wikipedia.org/wiki/Fagan_inspection

the original pull request was an email to the upstream developer to run "git pull your-repo your-branch".

the concept is at least as old as git itself, possibly even older. github just created a webinterface for it.

Scanning a few. Some are definitely written by AI but most seem genuinely human (or at least, not claude).

Anecdata: I read five and only found one was AI. Your sampling may vary.

Isn’t this a form of what he labels the “human coefficient”?

Some businesses prefer tools built by other businesses for some tasks. The author advocates pretty plainly to identify and target those opportunities if that’s your strength.

I think his point is to recognize that’s moving toward a niche rather than the norm (on the spectrum of all software to be built).

Sounds a bit like man pages. I think you’re onto something.

It’s moving fast. Just today I noticed Claude Code now ends plans with a reference to the entire prior conversation (as a .jsonl file on disk) with instructions to check that for more details.

Not sure how well it’s working though (my agents haven’t used it yet)

I believe they’re talking about Claude Code’s built-in agents feature which works fine with a Max subscription.

https://code.claude.com/docs/en/sub-agents

Are you talking about the same thing or something else like having Claude start new shell sessions?

Okay...and continue to work up the levels? Why do you think OAuth might be limiting? Why do you think they started building subagents first? What is the difference between subagents and products like Aider?

If they were able to wrap the API directly, this is relatively easy to implement but they have to do this within Claude Code which is based on giving a prompt/hiding API access. This is obvious if you think carefully about what Claude Code is, what requests it is sending to the API, etc.

What does OAuth have to do with any of this? I think you are deeply confused.

Someone said it best after one of those AWS outages from a fat-fingered config change:

> Automation doesn't just allow you to create/fix things faster. It also allows you to break things faster.

https://news.ycombinator.com/item?id=13775966

Edit: found the original comment from NikolaeVarius

I do like the idea of crowd-sourced collections of resources like skills.

It might be more useful if it was an index of skills managed in GitHub. Sort of like GitHub actions which can be browsed in the marketplace[1] but are ultimately just normal git repos.

[1] https://github.com/marketplace?type=actions

i thought of that but i didn't want to build a job to migrate that to the db. maybe we'll go that route.

It’s telling that Russia stands out in this list. “One of these is not like the others”

Georgians are literally light-skinned Caucasians.