Small note about :has() performance

[ghost]

refined-github/refined-github#7331 (comment)

Is there an issue with the selector? Can that be improved too?

I wasn't entirely sure myself. After chatting with Mozillians about it, they said that anchoring :has() to an element (or to all elements) slows down all DOM mutations inside that subtree (with more performance impact the more complex the selector inside :has() is). So stuff like body:has() or :root:has() or *:has() are recipes for bad performance.

For example, the quick-mention feature uses body:has(). It would likely be better to use a more specific anchor than body here. It's perhaps even better to avoid using has() if a simpler selector would suffice.

---

And refined-github/refined-github#7378 (comment)

[gorhill]

Related bugzilla issue: https://bugzilla.mozilla.org/show_bug.cgi?id=1896380

[stephenhawk8054]

So, do filters like ##div:has() have the same issue as the above CSS selectors? We are having ~30 filters like that in uAssets and EL.

[Yuki2718]

If I understand correctly, the problem is counts on subtree (and thus body:has() is bad) but in general won't be good.