This code doesn't exist anywhere besides my repo. So I posted it to stack overflow for others to reference. This was then downvoted, with no comments as to why. So I said, fuck that site and deleted my answer. Terrible website, terrible community, guess the AI will have to train on my repo instead.

As you type into the slot, it remembers the previous valid markup and uses that as a fallback until you finish typing and your input for the slot is valid HTML, then the component being demo'd updates instantly.

The slot content was using v-html, but that requires a wrapping HTML element. To get around this, I created a new component to pass the HTML into that uses a custom recursive render function I wrote that ensures all the attributes and child nodes render, but also doesn't break things while typing.

GNOME:
1. Every 7 seconds an application is in focus, give it 1 point
2. If it has <3,214 pts & hasn't been opened in a week, it's removed from the list automatically
3. If any app reaches 25,714 pts, all apps have their points reduced by half
4. There is no way to manually remove apps from the list

Windows:
1. When you open an app, increment a count by 1
2. Order all apps in the start menu by their count total so the apps you open most are at the top
3. If you right-click an app to remove it from the list, it's count is reset to zero

Sorry for spreading the misinformation. It does sound like Vitest shares a lot of common roots though. How much of Vitest shares the same ancestral code as Jest? From when it started vs now. Just curious on the ballpark. I was under the impression that improvements to either project helped the other

There's a reason converting from Jest to Vitest "just works" by changing `jest.*` to `vi.*`. Jest has done a good job of breaking up individual features to their own packages, and Vitest uses most of them: github.com/jestjs/jest/...
I don't know the actual percent, "80%" is just ballpark.