On Thu, Apr 23, 2015 at 6:46 AM, Warren Kumari <[email protected]> wrote:
> On Wed, Apr 22, 2015 at 8:43 PM, Watson Ladd <[email protected]> wrote:
>> I agree that DNSCurve is the best solution.
>
> ... which a: was not one of the options, b: is recursive to auth and
> c: has not been written up in a draft and brought to the WG.
>
> Please see the threads from October 2014 on DNSCurve (and DNSCrypt),
> which included a bunch of information and also:
> "Ok, so how about we say that it isn't a solution that we are currently
> considering, but if folk fully document dnscurve / dnscrypt in a
> draft, and bring it to the IETF / DPRIVE we will might consider it
> (and, obviously if it becomes a WG item, change it to meet the our
> requirements...)
> "

Looking through the archives I see that in October the DNSCrypt people
promised to write and upload a draft. That never happened,
unfortunately.

>
>
>>  Many of the proponents of
>> TLS based solution haven't adequately considered how this affects
>> anycast, DOS resistance, etc. Confidential-DNS can only be fixed by
>> essentially becoming DNSCurve.
>>
>> It's clear that deployed, working solutions need to be adopted. When
>> people say it's easy to implement DNS-over-TCP/TLS, and haven't, I
>> think that's a warning sign.
>
> Ok, so I'll assume you haven't been following along then.
>
> Please see:
> From the Dallas meeting: 'draft-hzhwm-dprive-start-tls-for-dns: DNS
> over TLS" - https://www.ietf.org/proceedings/92/slides/slides-92-dprive-0.pdf
> - Slide 6
>
> From the Hawaii meeting: 'DNS over TCP and TLS
> draft-hzhwm-dprive-start-tls-for-dns-00' -
> http://www.ietf.org/proceedings/91/slides/slides-91-dprive-5.pdf -
> Basically all of it, but mainly slides 21, 22, 23. Page 27 had a demo
> of drill talking to unbound...

I agree: we do have an implementation. Thanks for correcting me.

>
> Also, see http://www.isi.edu/ant/software/tdns/index.html, including digit.
>
> $ tar -xvzf digit-1.4.tar.gz && cd digit && autoreconf --install &&
> ./configure --without-gnutls && make
> $ echo -e 'www.google.com\nwww.yahoo.com\nwww.ietf.org' > names.txt
> # StartTLS
> $ ./digit -V  -r 185.49.141.38  -f ./names.txt -p 53 -t ssl
> # Direct over SSL.
> $ ./digit -V  -r 185.49.141.38  -f ./names.txt -p 1021 -t ssl -e
>
> The 185.49.141.38 (and  2a04:b900:0:100::38) test / demo server is
> running on the getdns jail, and is run by the getdns team; the getdns
> developers can all access and manage it (it was originally set up for
> the Bits-n-Bites demo).  NLnet Labs is a collaborator
> in the getdns project, and very kindly provides hosting.
>
> The digit code was written as a test / research project, and is useful
> for testing -- there are other things like the tdns_cli_proxy (also on
> that page) to e.g try using this like a user.
> There is also the Sinodun work
> (https://portal.sinodun.com/wiki/display/TDNS/DNS-over-TLS+Project+Homepage)
> , including a page helpfully called "Try DNS-over-TLS"...
>
>
> W
>
>
>
>
>
>>
>> Sincerely,
>> Watson Ladd
>>
>> On Wed, Apr 22, 2015 at 7:19 AM, Simon Josefsson <[email protected]> wrote:
>>> I support adopting 3) draft-hzhwm-dprive-start-tls-for-dns.  It may not
>>> be in shipping shape, but I think it is a worthy path to pursue and can
>>> be tweaked further along.  In particular, I think it would be a good
>>> idea to consider pushing documents for DNS-over-TCP/TLS and
>>> DNS-over-UDP/DTLS at the same time, with harmonized authentication
>>> language.
>>>
>>> Re 1 and 2, I don't understand what advantage they would give compared
>>> to DNS-over-TLS, DNS-over-DTLS, or DNSCurve.  I believe a strong
>>> justification is necessary to motivate adoption.  Frankly, I would
>>> prefer DNSCurve over both 1 and 2.
>>>
>>> /Simon
>>>
>>> Warren Kumari <[email protected]> writes:
>>>
>>>> Hi all,
>>>>
>>>> So, the big day has finally arrived -- we are initiating calls for
>>>> adoption on the three documents. < http://i.imgur.com/SKX3P8J.gif >
>>>>
>>>> For *each* of the below documents, please **clearly** state if you
>>>> would like DPRIVE to adopt it, or if you think that it will be a
>>>> distraction / not helpful.
>>>>
>>>> 1) Confidential DNS:
>>>> https://datatracker.ietf.org/doc/draft-wijngaards-dnsop-confidentialdns/
>>>>
>>>> 2) Private-DNS: 
>>>> https://datatracker.ietf.org/doc/draft-hallambaker-privatedns/
>>>>
>>>> 3) TLS for DNS: Initiation and Performance Considerations:
>>>> http://datatracker.ietf.org/doc/draft-hzhwm-dprive-start-tls-for-dns/
>>>>
>>>>
>>>> This call for adoption will be wrapping up on April 30th.
>>>> At that point we will decide on one or multiple of the documents....
>>>>
>>>> W
>>>
>>> _______________________________________________
>>> dns-privacy mailing list
>>> [email protected]
>>> https://www.ietf.org/mailman/listinfo/dns-privacy
>>>
>>
>>
>>
>> --
>> "Man is born free, but everywhere he is in chains".
>> --Rousseau.
>
>
>
> --
> I don't think the execution is relevant when it was obviously a bad
> idea in the first place.
> This is like putting rabid weasels in your pants, and later expressing
> regret at having chosen those particular rabid weasels and that pair
> of pants.
>    ---maf



-- 
"Man is born free, but everywhere he is in chains".
--Rousseau.

_______________________________________________
dns-privacy mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dns-privacy

Reply via email to