Re: [Drip] Nit comments on draft-moskowitz-drip-uas-rid-06.txt

[Robert Moskowitz <[e-mail]>]

On 9/9/20 4:55 PM, Alexandre Petrescu wrote:
>
>
> Le 09/09/2020 à 21:47, Robert Moskowitz a écrit :
>>
>>
>> On 9/9/20 3:26 PM, Alexandre Petrescu wrote:
>>>
>>>
>>> Le 09/09/2020 à 16:58, Robert Moskowitz a écrit :
>>>> Most addressed.  See inline.
>>>>
>>>> On 9/9/20 9:01 AM, Alexandre Petrescu wrote:
>>>>>
>>>>> I have read the document draft-moskowitz-drip-uas-rid-06.txt "UAS 
>>>>> Remote ID".
>>>>>
>>>>> Here I describe only the nit comments.  I save the higher
>>>>> level comments for later, after I gain more knowledge of this 
>>>>> Identity-centered environment.
>>>>>
>>>>> The terms are very numerous, and sometimes colliding, or some are 
>>>>> not enough well defined.  In an alphabetical order (alpha before 
>>>>> beta) the 'HID' term should appear before HIP term; similarly, HDA 
>>>>> before HI.
>>>>>
>>>>> It might be that it is not necessary to use the term 'Hierarchical 
>>>>> HIT Assigning Authority' too; maybe using the equivalent term 
>>>>> 'Registered Assigning Authority (RAA)' is enough.
>>>>>
>>>>> It  might be that it could be good to use a single set of terms, 
>>>>> by combining the terms from [drip-requirements] and the terms 
>>>>> listed in this document.
>>>>>
>>>>> Page 5 last paragraph: use an 'r' in 'through': "bind all content 
>>>>> in the ORCHID through the hashing function".
>>>>>
>>>>> Section 3: may I dare to suggest ASTM a new TYPE-5 to be "An
>>>>> IP address".
>>>>>
>>>>> Where it says " 100.50.hhit.uas.areo   IN PTR foo.uss.areo.", I
>>>>> think it should say 'aero'?
>>>>>
>>>>> Where it says "2001:14:28:14:a3ad:1952:ad0:a69e" I think it should 
>>>>> say "2001:db8:28:14:a3ad:1952:ad0:a69e" because
>>>>> 2001:db8 is the reserved prefix for documentation.
>>>>
>>>> No.  The HHIT prefix proposed is:  2001:10::/28, or some such, then 
>>>> you have the HIP HIT_Suite.
>>>>
>>>> 2001:10::/28 is what is used in HIPv1, deprecated. 2001:20::/28 is 
>>>> HIPv2.
>>>
>>> Bob, With all due respect. When I see such strong statement I
>>> think I mght be so wrong.
>>>
>>> But let me ask this: why was 2001:db8 invented?
>>
>> RFC 3849 provides this prefix for documentation that needs an IPv6 
>> address but does not have one to use. And said address will follow 
>> common routable address formats as in 64/64.
>>
>> RFC 3849 is a /32 and for HHITs we are requesting a /28 as used in 
>> RFC 5201 and 7401 and the whole 128 bit layout takes this into 
>> account. Thus using 2001:db8/32 would cause more problems than it 
>> would address.
>>
>> BTW, this is not the first comment to use 3849.  I have thought
>> about it, but followed more general RFC 2928.
>>
>> Please check out:
>>
>> https://www.iana.org/assignments/ipv6-address-space/ipv6-address-space.xhtml 
>>
>>
>>
>>
>>
>> and see footnotes 7 - 14.
>
> I looked.
>
> I do not want to sound provokative, and certainly I do not want to sound
> as if I wanted to re-track the discussion, but this does not answer my
> question.
>
> My question is why where 2001:db8 addresses invented?
>
> My answer is because we must make sure we dont put true IPv6 addresses
> in documents, because someone (probably the first implementers) will
> certainly copy-paste addresses from drafts in implementations, which is
> not good.  Another reason is because what one puts now in a draft (e.g.
> "2001:14:28:14:a3ad:1952:ad0:a69e") will one day be somebody's true
> address on a true system.  S/he will not be happy to find it then in a
> document available publicly, presumably the RFCXXXX of this draft.
>
> I do not know why does one think the 2001:db8 addresses were invented?
>
> Finally, if I struggle so hard in discussion to persuade with such a
> small nit - I frankly consider this 2001:db8 to be a nit - I wonder what
> might happen if I wanted to propose something else than HIP and ORCHID
> for identifying flying objects for safety purposes... but I have already
> seen that in my proposal about Bluetooth not being appropriate for
> distant drones, or in my proposal to use IP Router Advertisements sent
> by routers carried by drones.
>
> So, that is about that, I think.

If 2001:db8 were a /28 prefix, it would be usable here.  But part of the 
HHIT proposal is using a /28 prefix allocation.  It cannot be documented 
with a /32 prefix.  That is why it does not work here.

Those 4 bits (28 - 32) are the HIT_Suite_ID in the HHIT format.

I do see your point.  But it does not fit into the format of HHIT IPv6 
addresses.

Please read Appendix B for the HHIT format.


>
> Alex
>
>>
>>>
>>> Alex
>>>
>>>>
>>>>>
>>>>> Where it says:
>>>>>> Proof of UA transmission comes when the Authentication Message 
>>>>>> includes proofs for the Location/Vector Message and the observer 
>>>>>> can see the UA or that information is validated by ground 
>>>>>> multilateration
>>>>>
>>>>> What is the Authentication Message?
>>>>
>>>> I have added ASTM Msg Type to all first use of an ASTM F3411-19 
>>>> message.
>>>>
>>>>>
>>>>> What is 'UA'?
>>>>>
>>>>> Alex
>>>>>
>>
>