[DNSOP] "DNS resolver should not use 'ANY' to get cached records for TTL" (bugzilla)

[Paul Vixie <[e-mail]>]

this made the news tonight:

> Tracking Flags:    
> tracking-firefox36:    +
> status-firefox36:    fixed
> tracking-firefox37:    +
> status-firefox37:    fixed
> tracking-firefox38:    +
> status-firefox38:    fixed
> tracking-firefox39:    +
> status-firefox39:    fixed
> relnote-firefox:    36+ 


(https://bugzilla.mozilla.org/show_bug.cgi?id=1093983)

notes--

1. i support the adoption and progression of olafur's draft, regardless
of mozilla's retraction of the bad logic in tonight's firefox release.

2. it's not going to change the amplification/reflection problem and
should not mention that problem at all.

3. it should be renamed to "restricting DNS meta-data queries".

4. it should be expanded to include RD=0 against recursion-only servers,
IXFR/AXFR, and anything else in the DNS protocol that would be useful
for both diagnostics and surveillance.

5. non-response to these queries should be designed to avoid re-query
storms from common initiators.

i agree to review and contribute.

vixie