Auth48: H3 is like H2 but uses QUIC, we get it

[MikeBishop]

RFC Editor says:

Please review the number of times the document mentions
that HTTP/3 uses QUIC and is similar to HTTP/2. This seems somewhat
redundant by Section 2.

🤣🤣🤣

But fair enough -- are we overdoing it?

[MikeBishop]

Having looked at the references, they all seem reasonable with the possible exception of PUSH_PROMISE noting it behaves "as in HTTP/2." Most references to HTTP/2 occur either in the first two sections of the document or in the Appendix which specifically discusses the relationship between HTTP/2 and HTTP/3.

#4978 removes that one extraneous reference. Other than than, I plan to make no further changes to address this feedback unless someone wants to provide a more concrete list of places where the comparison is unnecessary.

[LPardue]

Marginal: Section 4.1.1 and 4.1.1.1 seems to have some egregious use of phrasing similar to "Like HTTP/2, ..." or "as in HTTP/2".

You could possibly condense these to a sentence such as "Like HTTP/2, HTTP/3 has additional considerations related to use of characters in field names, the Connect header field, and pseudo-header fields". That could be inserted into to thefirst para of 4.1.1, or the second para (thus breaking the para so the next one starts with "Characters in field names").

[MikeBishop]

Fixed in #4978.