Re: [DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Dick Franks]

Some of this lack of precision is spread about by much-loved DNS tools.


 ; <<>> DiG 9.11.4-RedHat-9.11.4-4.fc28 <<>> @b.root-servers.net . -t NS
 ; (2 servers found)
 ;; global options: +cmd
 ;; Got
answer:
<<<
 ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 37483
 ;; flags: qr aa rd; QUERY: 1, ANSWER: 13, AUTHORITY: 0, ADDITIONAL: 27
 ;; WARNING: recursion requested but not available

 ;; OPT PSEUDOSECTION:
 ; EDNS: version: 0, flags:; udp: 4096
 ;; QUESTION SECTION:
 ;.INNS

 ;; ANSWER SECTION:
 .518400INNSj.root-servers.net.
 .518400INNSh.root-servers.net.
 ...


And before anyone complains, similar infractions in perl Net::DNS
documentation will be fixed in next release.




Dick

___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Re: [DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Joe Abley]

Hi Richard,

I'd suggest that the unqualified nouns "question" and "answer" are sufficiently 
ambiguous as to be dead to us at this point.

I agree with the idea that a "response" is a DNS message generated in response 
to a query. Both "query" and "response" can be defined as DNS messages with 
particular values of QR observed in their respective headers.

I think "question" and "answer" can only usefully be used in other definitions 
as "question section" and "answer section". The document can certainly provide 
suggestions as to the set of meanings that other documents have previously 
ascribed to the bare nouns, but I think there ought to be a recommendation that 
they not be used in that way.


Joe

> On Aug 10, 2018, at 12:36, Richard Gibson  wrote:
> 
> On 08/10/2018 11:51 AM, Giovane C. M. Moura wrote:
>>> The response is the entire DNS message that is sent in reply to a
>>> question.
>> Great. Do you plan to define it in the document as well?
> Analogous to a "response" DNS message containing an "answer" (presumably 
> consisting of zero or more records), I'd be equally precise about a "query" 
> DNS message containing a "question" (consisting of zero or more entries, but 
> with any count other than 1 being essentially meaningless) and avoid 
> conflating those terms.
> 
> https://tools.ietf.org/html/rfc1035#section-4.1.2 :
> > The question section is used to carry the "question" in most queries, i.e., 
> > the parameters that define what is being asked.
> 
> ___
> DNSOP mailing list
> DNSOP@ietf.org
> https://www.ietf.org/mailman/listinfo/dnsop

___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Re: [DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Richard Gibson]

On 08/10/2018 11:51 AM, Giovane C. M. Moura wrote:

The response is the entire DNS message that is sent in reply to a
question.

Great. Do you plan to define it in the document as well?
Analogous to a "response" DNS message containing an "answer" (presumably 
consisting of zero or more records), I'd be equally precise about a 
"query" DNS message containing a "question" (consisting of zero or more 
entries, but with any count other than 1 being essentially meaningless) 
and avoid conflating those terms.


https://tools.ietf.org/html/rfc1035#section-4.1.2 :
> The question section is used to carry the "question" in most queries, 
i.e., the parameters that define what is being asked.


___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Re: [DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Giovane C. M. Moura]

Thanks Paul,

> First: we were probably sloppy in the use of the word "answer". In
> many cases, that should be "response".

Thanks for clarifying it, and don't get me wrong,  it's not only this
draft that had this --  many academic papers do the same (including
mine) -- but since this is a terminology doc, it's better to get it
right here.  So the next docs follow it.

> The response is the entire DNS message that is sent in reply to a
> question.
Great. Do you plan to define it in the document as well?

> However, in your GitHub comment, you suggest that an "answer" is
> both the contents  of the Answer section *and* the contents of the
> Authority section. Why would the contents of the Authority section be
> considered an answer?

That was just a clarifying question; I was in doubt myself.
So in that case the _response_ then does not contain an answer -- just
the authority.

Maybe this is clear for DNSOP fellows, but for  many researchers this
may not be.

> Also, in searching for "answer" in the document, I find that most of > the
> uses are as a verb, which is unambiguous. >It's only ambiguous when > the
> word "answer" is used as a noun.

Yes, you are right.

 >Unfortunately, there are some places
> where "answer" as a noun is used in material we quote from other RFCs,
> and we cannot change those.

True. Which supports the need for a terminology rfc :)

thanks,

giovane

___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Re: [DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Paul Hoffman]

Also, in searching for "answer" in the document, I find that most of the 
uses are as a verb, which is unambiguous. It's only ambiguous when the 
word "answer" is used as a noun. Unfortunately, there are some places 
where "answer" as a noun is used in material we quote from other RFCs, 
and we cannot change those.


--Paul Hoffman

___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Re: [DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Paul Hoffman]

On 10 Aug 2018, at 5:53, Giovane C. M. Moura wrote:


Maybe I am missing something,  but it seems to be that the term
'answers' and 'response' are currently being used interchangeably in 
the

draft.

I opened a issue on Github for that, with an example.

https://github.com/DNSOP/draft-ietf-dnsop-terminology-bis/issues/70


It would be good to have the conversation here on the list instead.

First: we were probably sloppy in the use of the word "answer". In many 
cases, that should be "response".


The response is the entire DNS message that is sent in reply to a 
question.


However, in your GitHub comment, you suggest that an "answer" is both 
the contents  of the Answer section *and* the contents of the Authority 
section. Why would the contents of the Authority section be considered 
an answer?


--Paul Hoffman

___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Re: [DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Giovane C. M. Moura]

Maybe I am missing something,  but it seems to be that the term
'answers' and 'response' are currently being used interchangeably in the
draft.

I opened a issue on Github for that, with an example.

https://github.com/DNSOP/draft-ietf-dnsop-terminology-bis/issues/70

/giovane

___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

[DNSOP] Genart last call review of draft-ietf-dnsop-terminology-bis-11

[Roni Even]

Reviewer: Roni Even
Review result: Ready with Nits

I am the assigned Gen-ART reviewer for this draft. The General Area
Review Team (Gen-ART) reviews all IETF documents being processed
by the IESG for the IETF Chair.  Please treat these comments just
like any other last call comments.

For more information, please see the FAQ at

.

Document: draft-ietf-dnsop-terminology-bis-??
Reviewer: Roni Even
Review Date: 2018-08-08
IETF LC End Date: 2018-08-13
IESG Telechat date: Not scheduled for a telechat

Summary: The document is ready for publication as a BCP with nits

Major issues:

Minor issues:

Nits/editorial comments:

1.  In section 2 the term DNAME is mentioned and while CNAME is specified DNAME
is not (maybe reference RFC6672?)

2. In section 5 "Most resorece record " typo.

3. This is more a comment and since I did not follow the progress of the
document I am not sure the motivation here. Reading the text I noticed that in
the definition of referrals in section 4 the text include also what looks to me
like logic starting from the third paragraph. I was wondering why is it here
and not in one of the standard track documents and referenced here. I saw that
this is a big change from RFC7719.


___
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop