[dns-at-ietf] Re: Discussion: big groups vs small, focused groups

[Peter Thomassen <[e-mail]>]

Hi Wes,

Thank you for moderating this discussion.

I have some thoughts, and am collating below the inspiration I took from various threads on this list.

Core to most of my conclusions is that overlap between DNSOP and any split-off group (unless it's very focused like DCONN) will draw a large fraction of the same people, just like we have big overlap with DNS OARC. With this in mind:

On 11/6/25 05:40, Ondřej Surý wrote:
> dnsop is overwhelmed with new ideas and I think it also suffers from burnout. It is hard to progress on drafts without feedback and the amount of work is not proportional to the number of people actually reviewing and commenting on these documents.

The amount of work is not reduced when distributed across a number of groups with largely the same people attending.

I am really skeptical whether expectations w.r.t. individual load will be fulfilled.

Work load is reduced by not following or reviewing certain drafts. Maybe we should remind people of that. Maybe we should make it easier to not follow drafts, by asking folks to prepend subjects with draft names, so it's easy to spot what's interesting. Maybe we should create sub-mailinglists for adopted drafts *where all DNSOP participants are automatically subscribed, but you can unsubscribe if you don't care about that draft*. Maybe ...

On 11/5/25 14:58, Shumon Huque wrote:
> I do think that splitting the core DNS work into operations and protocol development is probably a good idea, but I do not think we should make any more.

I agree that more than one (permanent) spin/off is probably not a good trade-off w.r.t. admin overhead, scheduling conflicts, information transfer (as raised by Wes) etc.

> Maybe for "new work"  we should require the work to be floated on the mailing list first and generate some level of discussion before granting them meeting agenda time.

I agree; this may be a useful policy (with exceptions at the chairs' discretion, such as for urgent discussions on a recently discovered vulnerability).

On 10/26/25 20:41, Eric Vyncke (evyncke) wrote:
>  9. Focused WG can sometime deliver I-D faster than larger ones, i.e., bypassing ‘head of line blocking’

I believe that much progress could be made here by

- having more volunteers for shepherding documents (often, authors wait for shepherds);
- by allowing more documents to advance to various WG states simultaneously.

The shepherd issue is not an issue that stems from having a large WG. Many smaller WGs with the same total work would need the same amount of shepherding.

With participants properly excluding drafts they don't care about from their attention, the simultaneous progression of drafts also seems to be only indirectly related to how many WGs there are. However, their is a chairs capacity/latency issue. Perhaps we can think about how to improve that by splitting work more between them, so that less (bi)weekly alignment is needed (which often dictates the cadence of events).

In sum, I'd like to caution that spin-offs might not lead to the desired result, mainly because the net work and the people pool remain the same.

Best,
Peter