search

bmeck / I-D

Internet Drafts
4 stars 4 forks source link

Move [HTML] and [TC39-MIME-ISSUE] from Normative to Informative #6

Closed linuxwolf closed 7 years ago

linuxwolf commented 7 years ago

Fixes #5

annevk commented 7 years ago

This is wrong. HTML needs to be normative for the encoding bits.

linuxwolf commented 7 years ago

@annevk

This document is carrying forward the charset encoding instructions from RFC 4329. Those look correct to me, and don't seem contradict the WHATWG document. Can you elaborate why it should be changed?

annevk commented 7 years ago

Looking at https://tools.ietf.org/html/rfc4329#section-4.2 step 2 and 1 are reversed these days and step 3 is bogus as far as HTML embedding goes (it defaults to the encoding of the document or an attribute on the script element if there's nothing else).

linuxwolf commented 7 years ago

I'm looking at https://html.spec.whatwg.org/multipage/scripting.html#concept-script-type, which is what [HTML] points to right now. This doesn't talk about the charset parameter on the media type, but the charset on the \<script /> tag. As such, it still reads as compatible with RFC 4329 § 4.2 to me, so I'm clearly missing something:

The [HTML] reference is turning into a bigger, arguably separate, issue.

annevk commented 7 years ago

https://html.spec.whatwg.org/multipage/scripting.html#prepare-a-script (step 21) https://html.spec.whatwg.org/multipage/webappapis.html#fetch-a-classic-script (step 6,7) https://encoding.spec.whatwg.org/#decode (step 4)

linuxwolf commented 7 years ago

@annevk thanks for the pointers. I'll create an issue to track this concern.