Re: [Ietf-and-github] Tracking drafts

[Rob Sayre <[e-mail]>]

On Fri, Feb 14, 2020 at 5:32 AM Eric Rescorla <[e-mail]> wrote:

> I can understand the motivation for this, but I think it just confuses
> things as-is.
>
> In general, we've long had a tradition that editors have a lot of
> discretion in how they prepare documents,
>

For most documents, it probably doesn't matter. There's a spectrum from:

Editors discuss among themselves and others, then submit drafts.
Design Team with closed membership
Design Team with some open aspects, etc.
Mailing List / Github with active design discussion in public

When a WG document has several implementations, and has been long-running,
it seems that the WG should make this call. Otherwise, implementations both
public and private can fall behind or diverge if the editors aren't
transparent about their progress (or lack thereof). This is particularly
true if an early draft is widely deployed (there might be implementations
the editors are not aware of).

For example, I'd expect HTTP WG edits to be pretty open and regular for the
core documents, given the number of implementations.

Note also the IESG statement on this:
https://www.ietf.org/about/groups/iesg/statements/design-teams/

I prefer the text we have, but the edit below is also true.

thanks,
Rob



>
> Work done in GitHub -- even when there is extensive discussion -- has no
> special status. That is, the output of any activity using GitHub needs to
> be taken into the Working Group
> and subject to approval, rejection or modification by the Working Group as
> with any other input.
>
> Note that I have removed the "mailing list" text as well, because 2418 is
> actually somewhat vague on exactly how work done offline in drafts is to be
> brought to the WG. The key point is that this is like any such input.
>
> -Ekr
>
>
> On Thu, Feb 13, 2020 at 2:03 PM Martin Thomson <[e-mail]> wrote:
>
>> On Fri, Feb 14, 2020, at 08:45, Martin Thomson wrote:
>> > +A choice to use GitHub is similar to the formation of a design
>> > +team (see Section 6.5 of {{!RFC2418}}) provided that the work uses a
>> public
>> > +repository.  That is, the output of any activity using GitHub needs to
>> be taken
>> > +to the Working Group mailing list and subject to approval, rejection,
>> or
>> > +modification by the Working Group as with any other input.
>>
>> Updated as per Rich and Rob.  Thanks for the prompt review.
>>
>> _______________________________________________
>> Ietf-and-github mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/ietf-and-github
>>
> _______________________________________________
> Ietf-and-github mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/ietf-and-github
>