Consider adding windows-936-2000 as a label for GBK about encoding

[hsivonen]

I'm not against this.

However, if we're adding more labels, I think there's probably more opportunity for benefit from supporting Java 1 encoding names. See https://searchfox.org/comm-central/source/mailnews/intl/charsetalias.properties and the java.io column at https://docs.oracle.com/javase/8/docs/technotes/guides/intl/encoding.doc.html

Particular categories of labels that are relevant to consider:

• ISO-8859 series labels of the form iso8859_N where N is the part number, e.g. iso8859_1.
• Code pages shared by Windows and DOS with the ms prefix. Of these, ms932 was already added as an ad hoc special case.
• Code pages shared by Windows and DOS with the cp prefix.
• Underscore versions of Unix CJK encodings: euc_cn, big5_hkscs, etc. My vague recollection is that Presto-era Opera found EUC_JP not to be Web-compatible due to Shift_JIS content labeled as EUC_JP. However, it's been long enough that my memory may be generating false recollections of Web compat problems.

[annevk]

https://annevankesteren.nl/2009/09/web-character-encodings agrees with your memory. The fact that no user agent ships those today makes me a bit more hesitant as it would probably require a lot more research than adding this single label to Gecko. Having said that, I'm not opposed if someone is willing to drive that effort.