Hi Michael,

See inline.

> On Dec 4, 2025, at 6:21 PM, Michael Richardson <[email protected]> wrote:
> 
> 
> Mahesh Jethanandani <[email protected]> wrote:
>> Thanks for posting an update. You beat me to the update. I was hoping
>> for the discussion to settle down and see an updated version that
>> captured all the discussion. Well …
> 
> Okay. I wish you'd said this earlier :-)
> please see: 
> https://github.com/IETF-OPSAWG-WG/draft-ietf-opsawg-pcap/pull/195/files
> I can post -17 if this satisfies you.

I did take a quick look at the diffs. Thank you for sharing.

What did we decide for grandfathered entries that do not have a reference? Did 
we not decide to refer to the document itself? E.g., LinkType Value 4, 5, etc.

> 
>> Besides these editorial changes, I am hoping for a resolution on the 
>> following points.
> 
>> 1. Consistency
> 
>> - The description of the LinkType Registry talks about 4 items per entry. 
>> They are:
>> - LinkType Name
>> - LinkType Value
>> - Description
>> - Reference
> 
>> The initial version of the table in Section 2.2.1 shows:
> 
>> - Name
>> - Number
>> - Description
>> - Reference
> 
> okay, I've renamed the initgial version file "stanzas"
> (note it's presented this way because a table did not fit, and IANA suggested 
> thisA)
> We changed the order of the table from Name/Value -> Value/Name.
> I have not changed the order in initial value stanzas, as there are many of 
> them, and
> we've departed from the .csv file and scripts at this point...

Is the order now Value/Name? Hmm. Maybe then I do not understand how the 
linktypes.xml file is rendered, and the order is what you say it is. I can wait 
for you to publish the version.

> 
>> Can these be made consistent?
> 
>> 2. I would have deferred the definition of Change Controller to Section
>> 2.3 of RFC8126. As is, the definition is incomplete.
> 
> So you want me to say:
> * Change Controller: as per {{RFC8126, Section 2.3}}

Yes, thank you.

> 
>> 3. In reading Section 2.3 of RFC8126, there seems to be some
>> flexibility on who is the Change Controller for entries that are being
>> grandfathered vs new entries. I see that the document now calls out
>> [email protected] for entries that are grandfathered. Thanks for
>> adding that.
> 
> okay.
> 
>> 4. Guy had suggestion for registration that do not have a reference. In
>> particular, if there is no document, can the email address for the
>> person who requested the LinkType be added?
> 
> That's the intention for all new entries.
> 
>> I believe that is captured
>> in Section 2.2.2 of the document which is somewhat misplaced as the
>> Section Title is “Guidance for Designated Experts”, says "The minimal
>> requirement is to provide a contact person for that link type.” I think
>> what was intended was “contact information for a person”, not just
>> “contact person”, or better still “email address of the person”.
> 
> I have changed it to say:
>  The minimal requirement is to provide a contact for that link type.
> 
> 1. email addresses for the person can go stale.
>   In fact, the person who registered the 23 JUNIPER ones (over a period of
>   time), no longer works at Juniper.  {A person who does has initiated an
>   internal query}
>   So while we have some additional information for some legacy allocations
>   recorded in emails and in the libpcap/pcap.c file, I would be hesistant to
>   put that into an RFC.
> 
> 2. If we got a request from, say, the NSA, we would say yes.  We might not
>   be able to list more than "NSA, Fort Mead, Drone to Aircraft Division"
>   (To make something up. Likely, it would be some contractor asking...)
> 
> 3. Like, phone numbers are not yet obsolete :-)

How about “contact information for a person, which should include the name as 
the minimum, and any additional information such as email address, phone 
number.”?

Thanks.
> 
> --
> ]               Never tell me the odds!                 | ipv6 mesh networks [
> ]   Michael Richardson, Sandelman Software Works        |    IoT architect   [
> ]     [email protected]  http://www.sandelman.ca/        |   ruby on rails    
> [
> 
> 
> 
> 
> --
> Michael Richardson <[email protected]>   . o O ( IPv6 IøT consulting )
>           Sandelman Software Works Inc, Ottawa and Worldwide
> 
> 
> 
> 


Mahesh Jethanandani
[email protected]






_______________________________________________
OPSAWG mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to