[Gendispatch] Re: Proposal: Reform of the AUTH48 process via Dual-Timer-Model

[Eliot Lear <[e-mail]>]

Hi

This proposal is inappropriate for GENDISPATCH.  Your proper avenue is 
the RSWG, since you are tinkering not with IETF processes but with RFC 
Editor processes.  Should you take your proposal to the RSWG, please 
include a more detailed account of what happened, and if you have it, it 
would be good to understand the RPC's position on this.

But not here.

Eliot

On 08.05.2026 00:47, Timo Gerke wrote:
> Hello GENDISPATCH Community,
>
> In this discussion, I would like to focus on a procedural reform of the
> AUTH48 process. The proposed Dual-Timer-Model is designed to reduce the
> administrative workload for Area Directors and Chairs while providing
> much-needed predictability for authors.
>
> The Problem:
> Current practice allows the AUTH48 phase to be extended until an Area
> Director or Working Group Chair intervenes and spends much time
> personally following up with contributors. This reactive approach leads
> to significant "Head-of-Line Blocking" in publication clusters.
>
> A current example is Cluster C430: 24 documents are in "PUB" or
> "AUTH48-DONE" status but have been stalled for over five months because
> a single core document is undergoing late-stage technical revisions.
> Crucially, these 24 documents have no normative dependencies on the
> blocker, yet they are held hostage by the current thematic clustering
> and lack of an exit strategy.
>
> The Proposal: The Dual-Timer-Model
> To transform AUTH48 into a predictable finalization step and provide a
> "neutral bad cop" for leadership, I propose two automated timers:
>
>   - Timer 1 (Fixed Duration): A 90-day limit from the start of AUTH48.
>     After 90 days, the document is published in its current state unless
>     a formal appeal/exception is granted.
>
>   - Timer 2 (Action-Triggered): A 30-day limit triggered as soon as an
>     author reports "No action required by the RPC."
>
> Benefits:
>
>   - Leadership Relief: The timers automate the pressure. ADs and Chairs
>     are relieved from manual follow-ups; the process enforces the
>     deadline.
>
>   - Logical Clustering: To be effective, clusters should be formed
>     strictly along normative dependencies (e.g., a "TLS-Cluster" vs. an
>     "IPsec-Cluster"). This ensures that independent work remains
>     unblocked.
>
> This is a simple procedural reform to ensure our publication pipeline
> remains agile. I would appreciate feedback, especially from ADs and
> Chairs, on whether this automation would provide the desired workload
> relief.
>
> Best regards,
>
> Timo Gerke
>
>