Re: last call Feedback for Opportunistic Security for HTTP (Experimental)

Good point, I'll try to clarify.

Cheers,


> On 29 Sep. 2016, at 2:49 pm, Kari hurtta <hurtta-ietf@elmme-mailer.org> wrote:
> 
>> Early cut here, feedback welcome:
>>  http://httpwg.org/http-extensions/opsec.html
> 
> 
> ( very quick read )
> 
> http://httpwg.org/http-extensions/opsec.html#rfc.section.2.1
> 
> | However, it is possible that the server might become confused about whether requests’ URLs have 
> | a HTTP or HTTPS scheme, for various reasons; see Section 5.4. To assure that the alternative 
> | service has opted into serving HTTP URLs over TLS, clients MUST check the “http-opportunistic” 
> | well-known URI defined in Section 3 before directing HTTP requests to it.
> 
> and then
> 
> 5.4 Confusion Regarding Request Scheme
> http://httpwg.org/http-extensions/opsec.html#confuse
> 
> and
> 
> 3. The “http-opportunistic” well-known URI
> http://httpwg.org/http-extensions/opsec.html#well-known
> 
> 
> and either
> 
> 2.2 Interaction with “https” URIs
> http://httpwg.org/http-extensions/opsec.html#rfc.section.2.2
> 
> | Because of the risk of server confusion about individual requests’ schemes (see Section 5.4), 
> | clients MUST NOT mix “https” and “http” requests on the same connection unless the 
> | http-opportunistic response’s origin object Section 3 has a “mixed-scheme” member whose value 
> | is “true”.
> 
> 
> does not make clear that this “mixed-scheme” check must be done for http-opportunistic resource
> returned by chosen alternative service.
> 
> 
> After all it is chosen alternative service which can be confused about scheme.
> 
> 
> 2.1 Alternative Server Opt-In
> http://httpwg.org/http-extensions/opsec.html#rfc.section.2.1
> 
> there is
> 
> | The chosen alternative service returns the same representation as the origin did for the 
> | http-opportunistic resource.
> 
> but that only apply when determining  “reasonable assurances”.
> 
> 
> Or have I missed something?
> 
> / Kari Hurtta
> 

--
Mark Nottingham   https://www.mnot.net/

Received on Friday, 30 September 2016 07:15:34 UTC