search

larseggert / moderation

This is the working area for the individual Internet-Draft, "IETF Community Moderation".
https://larseggert.github.io/moderation/
Other
1 stars 2 forks source link

Email suggestions from @ekr #26

Closed larseggert closed 3 months ago

larseggert commented 1 year ago

Email suggestions from @ekr:

I think this is a good idea and this draft is a good start. I have
made a few comments below, but I want to make one general suggestion
for enhancement.

In my experience one of the things that makes moderation difficult
is when each moderation decision is itself subject to extensive
debate in the same venue as it is happening. This constitutes a
DoS on the venue as well as discouraging strong moderation. I
think it would be good to establish a separate venue for debating
moderation discussions (e.g., an issue tracker or something)
that is distinct from the lists being moderated. This will allow
for public debate where necessary but not subject everyone else to it.

S 4.1.
   The moderator team SHALL operate according to a consistent and
   uniform set of criteria, processes, and actions.  The moderator team
   SHALL independently define and execute their role.  They SHALL
   maintain a public set of moderation criteria, processes, actions, and
   other material that allows the community to understand and comment on
   the role of the team.  The moderator team SHOULD consider adopting
   moderation criteria, processes, and actions that other technical
   communities have found suitable.  The moderator team's criteria and
   processes SHALL be developed with community input, but they are not
   expected to be documented in the RFC series.

So these do not require approval of IAB or IESG?

   The moderator team MAY initiate moderation actions by itself;
   individual participants SHOULD also alert the team to disruptive
   behavior they observe.  Participants should be able to contact the
   moderator team in ways that are, ideally, integrated into the various
   participation channels the IETF offers.

I think you should clarify that this is confidential.

Maybe I'm missing it, but it's not clear to me what the *powers*
of moderators are. For instance, can they ban someoen entirely
from a list? From an IETF list? It just seems to say "moderation
actions". FWIW my view is they should have quite broad powers.

S 4.2.

   Because the IESG and IAB are in the appeals chain for moderator team
   decisions (see Section 4.3), the IETF Chair SHOULD NOT appoint a
   moderator who is serving on the IESG or IAB.

I think this should probably be a MUST and I would change it to
forbid someone becoming a moderator and then joining the I*.

   Because the IESG and IAB are in the appeals chain for moderator team
   decisions (see Section 4.3), serving members of the IESG and IAB
   MUST NOT serve as moderators.

I don't see a clause that allows replacing moderators. That seems
like it needs to exist.

S 4.4.
   Due to the global nature of the IETF, the membership of this team
   SHOULD reflect a diversity of time zones and other participant
   characteristics that lets it operate effectively around the clock and
   throughout the year.  Team diversity is also important to ensure any
   participant observing problematic behavior can identify a moderator
   they feel comfortable contacting.

This seems to implicitly assume a very high level of responsiveness
(24x7). I'm not sure that's necessary but if you think it is, then
you should say so.
larseggert commented 1 year ago

And @mnot responded with

Piggybacking additional feedback on EKR's because I agree with much^H^H^H^Hall of it.

> On 7 Jul 2023, at 6:04 am, Eric Rescorla <ekr@rtfm.com> wrote:
[...]
> In my experience one of the things that makes moderation difficult
> is when each moderation decision is itself subject to extensive
> debate in the same venue as it is happening. This constitutes a
> DoS on the venue as well as discouraging strong moderation. I
> think it would be good to establish a separate venue for debating
> moderation discussions (e.g., an issue tracker or something)
> that is distinct from the lists being moderated. This will allow
> for public debate where necessary but not subject everyone else to it.

This is a very good suggestion, provided that we very firmly move that discussion there. I'm not sure folks have the discipline.

> S 4.1.
>    The moderator team SHALL operate according to a consistent and
>    uniform set of criteria, processes, and actions.  The moderator team
>    SHALL independently define and execute their role.  They SHALL
>    maintain a public set of moderation criteria, processes, actions, and
>    other material that allows the community to understand and comment on
>    the role of the team.  The moderator team SHOULD consider adopting
>    moderation criteria, processes, and actions that other technical
>    communities have found suitable.  The moderator team's criteria and
>    processes SHALL be developed with community input, but they are not
>    expected to be documented in the RFC series.
> 
> So these do not require approval of IAB or IESG?

This is perhaps the most important point. In the past, there have been very strong disagreements about moderation, and questioning of legitimacy of particular actions within the IETF process. Allowing the moderators to 'define [...] their role' and 'adopt moderation criteria' outside of the IETF consensus process opens up a number of questions about how that works, whether the output is legitimate in the eyes of the community, etc. 

> Maybe I'm missing it, but it's not clear to me what the *powers*
> of moderators are. For instance, can they ban someoen entirely
> from a list? From an IETF list? It just seems to say "moderation
> actions". FWIW my view is they should have quite broad powers.

Yes. It'd also be helpful if this draft listed the RFCs the authors believe will be updated or obsoleted if it is published.

> S 4.4.
>    Due to the global nature of the IETF, the membership of this team
>    SHOULD reflect a diversity of time zones and other participant
>    characteristics that lets it operate effectively around the clock and
>    throughout the year.  Team diversity is also important to ensure any
>    participant observing problematic behavior can identify a moderator
>    they feel comfortable contacting.
> 
> This seems to implicitly assume a very high level of responsiveness
> (24x7). I'm not sure that's necessary but if you think it is, then
> you should say so.

If WG chairs lose the ability to moderate (which I know is an open question), handling an issue through a moderator team is going to inevitably add latency. So I think it is necessary -- IME it's much more effective if bad behaviour is caught and stopped early.

To that question of whether WG chairs should still be able to moderate -- I'm not sure, but I observe very different behaviour in different WGs. What passes in some WGs as normal debate would be flagged by chairs as disruptive. Will moderators appreciate these differences, or set a new level for some WGs? And what will that level be?

Whether or not WG chairs still have the ability to moderate, it'd be good to understand what the expected interactions between chairs and moderators would be. E.g., will moderators check in with chairs on a proposed resolution, to get context and assure that it will work in that community? If chairs retain the ability to moderate, do they need to inform moderators when they act? Etc.

Finally, an editorial nit: I found Section very difficult to read, in part due to the liberal sprinkling of references throughout the text. It might help to move references to the end of the applicable sentence, where possible.