Link Chrome and Safari bugs for relative rgb()

[foolip]

Notes are originally from #21937.

[foolip]

@chrisdavidmills @romainmenke this removes notes added in #21937, can you review?

[romainmenke]

That is not the right test for those notes 🤔

[romainmenke]

This is a better test : https://codepen.io/romainmenke/pen/bGJJmZx

rgb(from red calc(r - 1) g b)

It tests both that r is of type number and the numeric value of r.

The result is expected to be almost red (254) and not black or white.

But we must not remove those notes, otherwise we make it seem as it was always implemented correctly, which it obviously wasn't :)

Safari is still broken.
Chrome was fixed in 122.

[romainmenke]

I've only now read the linked issue where you say that you think these notes are incorrect.
If they were it would make sense to remove them.

But the notes are accurate :)

[foolip]

Thanks @romainmenke! I couldn't find a Chrome bug for this, so I tested the simplest thing I could infer from the note, and that worked. Do you know if there was a Chrome bug for this?

[romainmenke]

Do you know if there was a Chrome bug for this?

I know there was one, but not filed by me for this specifically, so I would also have to do a search for "relative color" and crawl through bugs :)

[foolip]

I couldn't find the bug by search, but using https://codepen.io/romainmenke/pen/bGJJmZx I was able to bisect my way to https://crbug.com/41490327 which was indeed fixed in Chrome 122:
https://chromiumdash.appspot.com/commit/5deed17c9f1446cf0a061d7566e552e8e9bddfaf

[foolip]

And https://webkit.org/b/267647 works as a WebKit bug for this I think, even though it's not specifically about RGB.

[romainmenke]

I am also assuming that WebKit will fix a large range of issues at once, given how far they have fallen behind WPT at this point.

That bug seems relevant enough with that in mind :)