Re: [Idr] Ben Campbell's Discuss on draft-ietf-idr-as-migration-06: (with DISCUSS)

["Ben Campbell" <[e-mail]>]

On 18 Sep 2015, at 9:16, George, Wes wrote:

> Though I do not want to see this document delayed any more, I would
> actually prefer not to change the introduction unless WG consensus
> changes. This was DISCUSSed (and discussed) during evaluation of -03, 
> and
> this was my response to Pete at the time:
>
> "This one is a little bit of a grey area. It doesn't strictly require
> interoperability, because its functionality is local to a given BGP
> speaker. It doesn't make changes to the BGP Protocol per se but it 
> does
> describe the current behavior of something implemented in multiple
> vendor-specific methods so that desired behavior is documented for 
> future
> implementers. While it is optional, it does modify BGP's behavior from 
> the
> strict interpretation of the relevant RFCs, and that could technically 
> be
> considered a protocol change. Since it was a question as to which 
> status
> (PS vs info) was appropriate, we solicited WG consensus, and PS was 
> the
> recommendation.
> Some of the language used is to reflect the intent to codify existing
> behavior so that as future changes to BGP are contemplated (most 
> notably,
> BGPSec), this is considered part of the protocol so that it doesn't 
> get
> "broken"."
>
> The abstract and intro attempt to acknowledge this fact. I'm open to
> suggestions on wording changes to make it clearer, but I do think that 
> the
> text that is there is useful and should not be removed altogether.

I don't necessarily object to a PS that, for example, doesn't impact 
interoperability. It's just that the first paragraph as written feels 
like an explanation of why something _wouldn't_ be a PS. So even adding 
few sentences explaining why the working group wants this to be 
standards track in spite of that would be helpful, even if you otherwise 
keep the existing text.

Part of my concern is that lots of readers ignore headings and 
boilerplate. So if the body text seems to lead down a path that is 
different from what is said in the headings and boilerplate, at least 
some readers will misunderstand.

Thanks!

Ben.