Another field in the DNS

[John C Klensin]

Does this conversation belong in DSNOPS, or some similar place?
It seems  to me that you are discussing bugs that have only an
indirect impact on IDNs.

    john


--On Tuesday, March 03, 2009 08:55 +1100 Mark Andrews
<Mark_Andrews at isc.org> wrote:

> 
> In message <20090302212316.GN1700 at shinkuro.com>, Andrew
> Sullivan writes:
>> On Tue, Mar 03, 2009 at 08:10:06AM +1100, Mark Andrews wrote:
>> > 
>> > 	You should get AAA.com back in the question section.
>> > 
>> > 	The answer section should return the case the data was
>> > 	entered with in a case preserving system.
>> 
>> Well, that's my reading of the protocol, too, but that
>> doesn't seem to be how someone is reading it.  I gave it a
>> whirl, and here's what I got:
> 
> I keep thinking about turning on case sensitive compression for
> responses to regular queries.  It is on for axfr/ixfr
> responses. The only reason I haven't done it is I have heard
> reports about middleware that assumes that the answer section
> will start with 0xc0 0x0c (a compression pointer to the start
> of the question section).  Such middleware is broken.
> 
> That reminds we we need to check that this is covered by
> draft-ietf-dnsext-dnsproxy.
> 
> Mark