Are all HTML character entities valid in unicode?

[chimerical]

Are all HTML character entities valid in unicode? Are they ASCII only?

Additionally, how would they fare with encodings like these:

<meta http-equiv="content-type" content="text/html; charset=utf-8" />
<meta http-equiv="content-type" content="text/html; charset=iso-8859-1" />

[trolly]

Yes.
UTF-8 is a unicode encoding. ISO-8859-1 not so it can only represent a small subset of unicode except using HTML entities. HTML entities are an ascii encoding of unicode characters.

[chimerical]

When you say ISO-8859-1... "except" using HTML entities, do you mean that ISO-8859-1 represents HTML entities + subset of unicode? And using HTML entities is acceptable practice in both cases?

[trolly]

HTML entities are special codes like &amp; for & .There are lots (better most) of unicode characters which can only be written as entities in ISO. In UTF-8 you can (theoretically) write them directly without using HTML entities. ISO has a range of 256 characters while unicode has a range of (theoretically) 4 billion characters.

[daniel219]

HTML codes are valid only in unicode they dont have any relationship with ASCII

[peter.reisio]

Just to clarify, since daniel219's post is misleading...

Many are defined in HTML.

You don't need any if you use UTF-8 and send your code as UTF-8 (meta elements are a truly gimpy alternative), though for certain situations you may well save time by typing &copy; rather than knowing how to input ©.

[Pim]

I don't think daniel219's post is misleading at all in relationship with the original question.

Are all HTML character entities valid in unicode? Are they ASCII only?

HTML codes are valid only in unicode they dont have any relationship with ASCII

simply saying that you can use entities in any HTML file (no matter if the file's charset is USASCII or otherwise), and that the results are always Unicode, even with numerical entities that have the same value as a value in the charset.
For instance, even if the charset is x-ebcdic, &#80; should always yield 'p' rather than '&'.
And with 'valid in Unicode', he means that a code like &#55555; is not a valid character, even if the charset is, ehm, well, some encoding where 55555 is a valid character code.

[peter.reisio]

By misleading I meant wrong.

[chimerical]

I see. Good to know, guys! Yeah, I was mainly asking because I wanted to know if something like &raquo; would display nothing if I declare a random encoding in the meta tag, but it seems like they would all understand this and display it properly.

[peter.reisio]

Yes.

[dtobias]

Some old broken browsers may have treated entities and numeric references differently based on the character encoding, but that's not what the standards call for, and not what current browsers such as Firefox do.

[Franadora]

by "old broken browsers" do you mean IE6 - I only ask because it's still around and is sometimes not acting like modern browsers.

[Pim]

I don't have IE6 here, so someone correct me if I'm wrong, but I'm fairly sure it treated *most* entities correctly. Exceptions are the ones between &#x80; and &#x9F; it will act as if those are in the current charset. (e.g. &#x9F; will display a 'Ÿ' when the current charset is windows-1252).

[dtobias]

You might have to go back to really ancient browsers like Netscape 3.0 to find really broken behavior in this area, but there was a time when it was still a problem.