Re: [GROW] Call for GROW WG adoption of grow-overlapping-routes

[William Herrin <[e-mail]>]

On Thu, Aug 30, 2012 at 10:03 PM, Christopher Morrow
<[e-mail]> wrote:
> On Thu, Aug 30, 2012 at 9:59 PM, William Herrin <[e-mail]> wrote:
>> On Thu, Aug 23, 2012 at 1:44 AM, Peter Schoenmaker <[e-mail]> wrote:
>>> At IETF in vancouver we had a presentation from Russ White for a new working
>>> group item draft-white-grow-overlapping-routes-00.  I would like to gauge
>>> the working group interest in adopting this draft as a working group item.
>>
>> What am I missing? The last few times this was discussed here or on
>
> you're probably not missing anything in particular, except that you
> didn't state whether or not the doc (or the ideas that the doc
> contains/explains) should show up as a working-group work item.

Hi Christopher,

Well, unless one of the authors wants to step up and say, "Bill, we
anticipated the problem you posed and this is how we cleverly got
around everything like it," I'd have to say thumbs down on adopting it
for discussion and work. Recommend the authors pore through the RRG
archives for a history lesson in all the myriad approaches to RIB
aggregation that are known to not work. If aggregating overlaps isn't
the most popular non-working approach it's surely the second most
popular.

That, by the way, would make a great draft: a canonical list of the
approaches considered for RIB aggregation and the key faults
identified with each. So that the next guy doesn't have to pore
through the RRG archives and, when somebody does find a novel idea, he
can explicitly point out that, "This is why it doesn't suffer the
fault identified in RFC X section 8."

Regards,
Bill Herrin


-- 
William D. Herrin ................ [e-mail]  [e-mail]
3005 Crane Dr. ...................... Web: <http://bill.herrin.us/>
Falls Church, VA 22042-3004