Re: [dnsext] draft-mohan-dns-query-xml-00.txt

[Nicholas Weaver <[e-mail]>]

On Oct 1, 2011, at 2:02 PM, Alex Bligh wrote:
>> No, what I'm assuming is SOME minimal-level of nonborkenness.
>> 
>> IF a cache takes url
>> 
>> http://www.example.com/url_a
>> 
>> and instead returns
>> 
>> http://www.example.com/url_b
> 
> This is true, but the GET/POST question is half orthogonal (so that would be 45 degrees?) to the nonce question. IE you could use a POST plus a nonce if you wanted. IIRC POST /should/ do enough alone to avoid the need for a nonce though, as I can't see how any web cache would cache POSTs.

Yeah, but if we're assuming maximum borkennes, I could see truly borken proxies failing to properly return data in the reply of a POST.

Thus if the goal of using POST instead of GET is cache-busting, GET with NONCE should work unless the case is "Broken to the point of uselessness HTTP proxy".


Oh, one BIG warning that might sabotage this that I just realized:  I've seen proxies, eg the Bluecoat proxy, which will intercept all traffic until the BROWSER is logged into the proxy and has a cookie set, so there isn't HTTP access unless through a browser that's already gone through the login process.  :(

Since thats likely to the the same sort of network where you can't bypass the DNS borkenness by TCP or UDP port 53, this could be a problem.



_______________________________________________
dnsext mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/dnsext