[ih] ROAD [was: This Review is for Everyone]

[Brian E Carpenter]

On 23-Mar-26 15:31, Craig Partridge via Internet-history wrote:
> On Sun, Mar 22, 2026 at 6:25 PM Dave Crocker via Internet-history <
> internet-history at elists.isoc.org> wrote:
> 
>> On 3/22/2026 5:20 PM, Noel Chiappa via Internet-history wrote:
>>> but the person in charge of the nascent IETF needed a rare mix of skills
>>
>>
>> It is indeed a rare mix.  And I think many of us worked very hard -- and
>> very well -- over the decades, demonstrating just how rare it is.
>>
> 
> Expanding here (and recognizing that Dave and Noel were both on the initial
> IESG with me), figuring out how to get the IETF to work was a tortuous
> effort (and we can argue about definitions of "work").  If a working group
> (WG), especially one working on a critical problem, failed to make progress
> between successive meetings (which were roughly every 3.5 months) that was
> a red flag, and failure to make progress over three meetings was an
> emergency.  The IETF leadership tried different things, including:
> 
>     - snap WG meetings, scheduled on short-notice, where in-person
>     participation was required.  This was used by SNMP and MIB WGs to deal with
>     efforts to slow down the WG to favor impending non-SNMP products (it worked
>     but I think in retrospect feels expedient and anti-openness);
>     - changing WG chairs -- notably worked well for 8-bit SMTP (John Klensin
>     stepped in and averted a disaster [with a LOT of technical help from
>     others]);
>     - splitting WGs and/or allowing competing WGs -- e.g. IS-IS and OSPF
>     (generally worked well and continues as a practice to this day);
>     - advertising that certain decisions would happen at the WG's next
>     meeting e.g. for IP over ATM (worked well because it shut off debate -- a
>     form of a closure vote);
>     - closed WGs -- most notably the ROAD group, which sought to find a
>     solution to IPv4 address exhaustion (mixed reviews on whether this worked
>     or caused problems and I don't think has been done again)

Was ROAD actually a WG, or just an ad hoc committee formed by the IESG?

Either way, I think it *did* work, and (without going back to check the documents),
I believe it identified three separate work items that all had to be tackled:

1. Classful addressing was harmful
2. BGP3 was on the edge of disaster
3. IPv4 would run out of addresses

and then the IESG did stimulate the necessary work on those three topics.

(I only got roped in to help with #3 some time after Kobe.)

    Brian

> 
> Part of what allowed things to come out OK in the end as we experimented
> with different management solutions was that the vast majority of IETF
> attendees wanted the Internet to work -- the cats were all trying to get to
> the same destination -- the challenge was getting them to move briskly
> together in that direction (which is less bad than the usual cat herding
> problem).
> 
> Craig