Re: [Doh] assorted things

[Daniel Stenberg <[e-mail]>]

On Tue, 5 Jun 2018, bert hubert wrote:

>   Some of these non-successful HTTP responses (e.g., redirects or
>   authentication failures) could mean that clients need to make new
>   requests to satisfy the original question.
>
> I would recommending teeth to this. Naive implementations may decide to not 
> follow 3xx codes. I would recommend making this explicit with a MUST. Also, 
> it may be worth it to explicitly say if authentication is in our out of 
> scope.

Isn't all of this basically just how HTTP works and offers? RFC 7231 section 
6.4 explains redirects just fine I think and HTTP authentication is certainly 
a part of HTTP and is explained in RFC7235.

> Here it is a bit uncertain if you'd ever prompt a user for a 
> username/password for doing DNS resolution.

To me that sounds more like a UI/program decision and not something we need to 
specify.

> In 6.3 "Server Push":
>   For HTTP server push ([RFC7540] Section 8.2) extra care must be taken to
>   ensure that the pushed URI is one that the client would have directed the
>   same query to if the client had initiated the request.
>
> This means an API server is free to send responses to DNS queries we haven't
> seen yet?  And should a client do something with that?

Yes, basically. See RFC 75450 section 8.2.

> Or can it ignore the pushed records?

A HTTP/2 client can always opt to either not ever accept pushed resources or 
deny them when offered.

-- 

  / daniel.haxx.se