Linus and tools

Speaking of changes of heart: you might remember Linus (yes, that one) writing that you can't trust a C++ compiler?

Here's a recent email from him:

From: Linus Torvalds

Date: Tue, 13 Oct 2026, 13:45:09 -0700

Subject: Re: [PATCH] docs: remove Assisted-by mentions

On Tuesday, 13 Oct 2026 at 13:16, Laurent Pinchart laurent.pinchart@ideasonboard.com wrote:

> This is about setting our expectations, communicating what we want and what we don't want.

No. You're talking about something completely different. This has nothing to do with "expectations", and everything to do with your own insecurities.

I don't understand this whole idiotic concept of "don't use AI for X".

Here's a new rule just for you: don't use compilers for any of your kernel patches.

Because if you can't show that you really understand what the machine does by writing assembly, I don't want to see your code contributions. You're obviously incompetent if you can't write it that way.

And by the way, you'd better not use an editor to write your commit message either.

Because I need you to show that you can write a good commit message without tools that let you easily erase text and hide the mistakes you made while typing.

Got it?

Otherwise, those tool-assisted commits clearly don't show that you understand what you're doing.

Can't you see how STUPID those rules would be?

So no. We're not going to introduce stupid rules like "you can't use tools for X".

My expectation is that people do their best, and I expect them to understand what they're doing.

But that understanding has very little to do with which tools they use.

And frankly, in my experience, the people I respect and consider smart don't shy away from new and interesting tools.

Linus

Twenty years on, the guy's gotten a bit wiser, though he still talks like an asshole.

Shall we vibe code some kernel patches? Oh, right, Russians aren't allowed in anymore. Still an asshole.