The delete-in-child-of-head.tentative.html tests proposes a few cases where certain metadata-content should be protected against merges / removal when they are part of a contenteditablearea.
The have been added by Firefox in bug as tentative, so it's understandable that even Firefox don't pass all of them. However, I have some doubts about when it's allowed and when not to consider style elements as editable, hence subject to editing commands, like delete or merge. For instance:
Case 1
<!DOCTYPE html>
<pre contenteditable style="border: solid">
foo
<style contenteditable="false">pre { background: cyan }</style>
bar
</pre>
The 3 major browsers (Firefox, Safari and Chrome) allow the styleelement to be deleted. This would allow the user to remove the format of the editable area, which seems like a legit use case.
Case2
<!DOCTYPE html>
<head contenteditable style="display: block; border: solid"></head>
<script>
document.head.textContent = "foo";
</script>
Only Firefox prevents the script to be removed.
So the question is, is the intention of the tests to imply that any metadata-content element should be considered non-editable in any circumstance ?
The delete-in-child-of-head.tentative.html tests proposes a few cases where certain metadata-content should be protected against merges / removal when they are part of a
contenteditablearea.The have been added by Firefox in bug as tentative, so it's understandable that even Firefox don't pass all of them. However, I have some doubts about when it's allowed and when not to consider
styleelements as editable, hence subject to editing commands, like delete or merge. For instance:Case 1
The 3 major browsers (Firefox, Safari and Chrome) allow the
styleelement to be deleted. This would allow the user to remove the format of the editable area, which seems like a legit use case.Case2
Only Firefox prevents the
scriptto be removed.So the question is, is the intention of the tests to imply that any metadata-content element should be considered non-editable in any circumstance ?