[coc-tf] Draft Minutes – CoC TF Seventeenth Call

Dear TF members,

Here are the draft minutes from last week's call.

Cheers

Ant


***

*Draft Minutes – CoC TF Seventeenth Call*

22 October 15:30 (UTC+2)

*Attendees:* Denesh Bhabuta, Athina Fragkouli, Antony Gollan, Leo 
Vegoda, Salam Yamout

*Scribe:* Kjerstin Burdiek*
*

*Summary: *The TF discussed the relevance of the Python community’s CoC 
and what they could borrow from this. They discussed what might 
constitute a conflict of interest for a member of the CoC Team and 
agreed that it might be wise to have someone from the RIPE NCC’s Legal 
team available if an impartial voice was needed. The TF also decided 
that having a few investigators rather than a single individual might be 
the best approach for quick decision-making. The TF discussed the role 
of WG chairs in cases where violations took place on a mailing list. The 
TF concluded that input from the WG chairs could inform an investigation 
but the CoC Team itself must make the decision. The issue of acceptable 
evidence in a case was also brought up, and Athina advised that publicly 
available records were permissible as well as private records. Finally, 
the TF considered whether an escalating process of consequences would be 
useful and decided that it would be best to have a list of consequences 
that did not have to occur in sequence. Leo said he would draft a 
proposal for the CoC Team process that he would share with the rest of 
the TF for discussion.

*1. Minutes*

The minutes from last meeting were approved.

*2. Agenda*

There were no changes to the agenda.

*3. Actions*
*
**- (Ongoing) TF to think about how the CoC Team and the WG Chairs would 
work together to manage discussions on mailing lists. Including tools, 
e.g. Mattermost channel**
**
**- (Ongoing) TF to think about the appropriate level of process to 
allow further improvements to the CoC.**
**
**- TF to review RIPE NCC Arbitration Panel, ICANN Ombuds, and other 
relevant comparative process documents*

Leo asked for thoughts on the process documents that were shared at the 
last meeting.

Salam said she was in favour of the Python community document.

Leo said he found that was more practical than some of the others. The 
book Brian had recommended was also useful. He marked this action as 
complete.

*4. Process*

Leo said the goal of the meeting was to get a feeling for possible steps 
in the process and timeframes. As an example, he referenced the 
multi-step process in the Python document.

Denesh asked if the RIPE NCC’s legal team needed to be involved when 
evaluating conflicts of interest.

Athina said that parties with conflicts of interests would typically 
highlight this in advance so it could be evaluated.

Salam said she agreed, but she thought conflicts of interests had to be 
investigated on a case-by-case basis. This was essentially a situation 
where a person on the CoC Team felt they had a conflict, such as a 
family member being involved, and this was then reviewed.

Denesh agreed that conflicts had to be evaluated on a case-by-case 
basis. However, he suggested that the RIPE NCC’s Legal team might need 
to be involved, to evaluate potential conflicts of interest for 
candidates during the CoC Team selection process.

Leo said that they needed a fit and proper person for that.

Salam asked Denesh to clarify, as he seemed to be saying that anyone on 
a the team would need to be evaluated for a conflict of interest. She 
asked what would constitute a conflict of interest that made someone 
unfit and if they were considering ethical judgments as part of that 
evaluation.

Denesh said that he meant a legal risk management assessment should be 
done when determining who could be on the CoC Team to evaluate the 
potential legal consequences of conflicts of interest.

Salam asked if he meant that there should be a de facto legal 
professional on the CoC Team.

Denesh said there should be a legal professional either on the team or 
as a liaison. He asked what Athina thought.

Athina said this was a complicated topic. Sometimes there were clear 
rules regulating conflicts of interest, such as when selecting an 
association’s Executive Board members, but that in this case, there were 
no such regulations. Instead, they should focus on what would make the 
CoC Team strong and trusted. In the arbitration process, it was up to 
the arbiters to evaluate whether there was a conflict of interest, and 
the same was true of the RIPE NCC Executive Board. A lawyer’s input 
would provide more certainty but was not necessary.

Leo said they were discussing two different things here. One was what 
the requirements for serving on the CoC Team should be. The other was 
whether a person on the team should be involved in a particular report. 
Salam was asking about criminal background checks being conducted on 
candidates for the CoC Team, and she had asked what would happen in the 
case of a family member being a potential conflict of interest. The 
Python process covered this by pointing out that although someone might 
not have named all of their family members from the beginning, if such a 
situation were to arise, they must acknowledge that they could not be 
impartial. He asked if there should be an iterative conflict of interest 
process, where the first person to receive the report was the ‘triage’ 
who either handled the case themselves or passed it on if they had a 
conflict of interest. Or should reports go to the whole team, and people 
with conflicts of interest then step back for that investigation?

Antony said a conflict of interest would probably have more to do with 
who was involved in a report rather than the details of the case.

Leo said this was a good point.

Athina said this was connected to the question of whether this should be 
done by a committee or by one person. In the arbitration process, there 
was a panel of arbiters, but only the arbiter at the top of the list of 
panellists’ names would get the case and issue the ruling. They could 
not apply a similar method if there was a group of people making the ruling.

Antony said they might want at least one other person involved for 
oversight.

Salam said they should decide how many people would be making the 
decision and how those people were chosen.

Leo said that at the last meeting they decided it was necessary to have 
a diverse group of people on the team to maximise (linguistic and 
cultural) understanding. This requirement might not work with Athina’s 
suggestion to imitate the arbitration process.

Salam said whatever they decided was fine, as Athina could adjust it as 
needed. She agreed with the need for diversity and suggested having 
seven CoC Team members. She asked if the team member on a case would be 
chosen round-robin or if there would be a subcommittee of two or three 
people. They had agreed that they would have a pool of diverse and 
ethical individuals to choose from, but they had not agreed yet who of 
these people would work on a case.

Antony said the RIPE NCC arbitration process was not time-critical and 
so could take as much time as needed to evaluate conflicts of interest. 
They could not do the same at meetings. He asked what they would do if 
there was only one CoC Team member present and this person had a 
conflict of interest that prevented them from handling a report at the 
meeting.

Salam agreed that the urgency issue was a good point to consider.

Leo said, given that, it might be important to have a professional from 
the RIPE NCC, such as a person from Legal or HR, who was more likely to 
be impartial in such a situation.

Salam asked if he meant as part of the team or as the investigator.

Leo said he meant someone from the RIPE NCC should be available. It was 
not necessary to decide their exact status right now.

Salam asked how many people should perform the investigation. She 
suggested three.

Leo said he preferred an odd number of people for effective decision-making.

Athina said they could have a secretariat to solve these issues and get 
the advice needed from the RIPE NCC (on any matter, not just conflicts 
of interest). The number of investigators might be relevant to the 
number of people available at a meeting. An odd number was a good 
decision, and furthermore it was important to require a certain number 
of these people to be present at a meeting.

Leo said they needed to consider two things. One, an issue might not 
happen at a meeting but at an in-between meeting or on the mailing 
lists. Two, the WG chair might need to act in such a situation if it 
happened on their WG’s list. The CoC Team would then have more people to 
draw on to assist, but the RIPE NCC might therefore need to provide some 
training and guidance to WG chairs for these situations. However, the WG 
chairs might not have anticipated getting involved in these kinds of 
disputes.

Salam asked if there was someone in charge of every mailing list.

Leo said WG chairs were responsible for their mailing lists. There was 
now documented guidance on how to manage those lists. However, non-WG 
mailing lists were managed by the RIPE Chair.

Salam said there was also someone in charge of the venue at a meeting. 
The group investigating a case should therefore be made up of three 
people: the secretariat, the person investigating, who could be a WG 
chair if the issue took place on their list, and a person from the group 
of people trained to be CoC Team volunteers.

Athina said if the ruling were to come from the CoC Team, then all three 
people should be from the Team. Any extra advice from other 
professionals could be taken into consideration but not be part of the 
ruling itself.

Salam agreed that this made sense.

Leo said this element was crucial. He said they might want WG chairs to 
remind people about following the CoC but should not ask the chairs to 
be involved in investigating a report. If someone thought a serious 
breach had occurred, then it would go to the CoC Team.

Salam said the investigator(s) from the CoC Team could ask the WG chair 
for their opinion, but the WG chair would not make the decision.

Leo said this sounded good. If a breach occurred on a mailing list, 
there would be a record. The same was true for a video at a RIPE Meeting 
or a Zoom intersessional meeting. He asked what the rules were for 
accessing that record and any other related information.

Athina said that if it happened in public, the records were publicly 
available. This was true for the cases Leo mentioned. If something 
happened in a private setting, but the reporter had evidence of it, they 
could still submit that evidence for the case. This would include 
personal information about the person who committed the violation even 
if that person were to argue this information was intended to be 
private. She and Antony had been looking at the issue regarding the data 
protection implications for someone accused of a CoC violation. The 
solution could be having waivers in which the waiver signee gives 
permission to share such private information should they be accused of a 
violation. This could be given to any attendee agreeing to the CoC and 
the process around it.

Salam asked for a recap of the items they had covered in the Python 
community process.

Leo said they had covered the first four items of the Python CoC 
document and were now on the step relating to who would be involved in 
the CoC process. He put an action on himself to write up a brief text 
and share it with the mailing list for discussion.

Salam said they should decide if they wanted to model themselves after 
the Python community’s CoC document. Then they could make a parallel 
process, and Athina could share everything at once on the CoC mailing list.

Leo asked if she was saying that Athina should write a proposal for who 
should be part of the team that conducts an investigation. He asked how 
Athina felt about this.

Athina said she thought Salam was suggesting someone take this structure 
and write up a similar process for the CoC Team. She asked Salam if this 
was correct.

Salam said it was. She asked if they should decide now whether they 
should create a similar process and adapt it as needed. Someone would 
need to propose such a process and then share it for discussion among 
the TF and on the mailing list.

Leo said he would create a draft and send it to Athina and Antony for 
some adjustment. They would then send it to the mailing list for 
discussion. The Python community document was a sound structure but 
needed to be slightly adapted.

Salam said the structure was acceptable but there was no information 
about who could access or publish a report. They would need to add this. 
She and Leo had had a previous conversation about it.

Leo said they had discussed not publishing individual reports but 
instead reporting about their themes and statistics. They had also 
spoken about the RIPE NCC needing to maintain records for implementing 
decisions, such as when someone is temporarily banned from posting on a 
mailing list.

Salam said they needed to add a 12th step for archiving decisions as 
necessary, and then Athina could add simple explanations to adapt it for 
their case.

*5. Ladder*

Leo said the ladder of consequences was interesting because it focused 
on a software development community and had restricting tool access as 
one of its consequences. The tool access aspect was not relevant for 
them, but for them it could be mic access, for example. He asked if they 
also wanted to implement an escalating process that began with revoking 
permissions and culminated in banning, for instance, in the case of 
violations.

Athina said the escalation made sense, but they needed to consider the 
different platforms where the violation could happen, such as on the 
mailing list or at an event. In the mailing list, they could restrict 
people from sending messages for a certain period of time and escalate 
to banning them if needed. Regarding events, they already had an option 
to ban people from the meeting but no restriction of permissions during 
the meeting. The justification for the ban came from the terms and 
conditions, but it would be possible to have an escalation process 
before that. They needed to discuss how to actually implement those 
other steps before a ban.

Salam agreed that escalation was important to make clear this was a 
process. Several small violations were as bad as one big one. They 
should think critically about restricting access to tools because 
attendees were mostly from democratic countries and also paid for the 
meetings. She asked Athina how they should address a violation if it 
were to happen somewhere private in the venue outside of the meeting.

Athina said they could count this as part of the meeting.

Antony asked if it would be better to have a ladder with levels or to 
list what different consequences would entail. The aspects of the ladder 
that might be relevant for them were verbal warning, restriction from 
the list, and banning from meetings.

Leo said they needed it to make it clear that anything in the escalation 
process was just an example, and the people involved needed to use their 
judgment to decide an appropriate solution. For example, if someone 
acted badly at a social event, they could be banned from social events, 
but maybe not from the whole meeting.

Athina said she was confused about what they meant by escalation. She 
said that although there were different levels of punishment, one did 
not need to pass through all of them and could go straight to a more 
serious punishment.

Antony agreed that you could skip levels.

Athina said, in that case, banning people from mailing lists may be as 
bad as or worse than banning them from a meeting. They needed to 
consider the forum as well. Maybe they should not present these as 
different levels of punishment but just as different punishments.

Denesh said this was true. Banning people from the mailing lists was as 
severe as banning from them meetings, as the lists were where most 
discussion took place.

Salam asked if they agreed there should be a list of consequences, but 
not necessarily levels.

Leo said there should be examples of consequences and that punishments 
did not have to go from the first example to the second. Novel 
consequences would be fine if they were justifiable.

*6. AOB*

Salam asked who would write the levels of consequences.

Leo said they should make that into one or two action items. He would 
draft some text to send to Antony and Athina. Then they would have a 
draft proposal to discuss, but not necessarily to stick to. Regarding 
the last meeting’s minutes, Mirjam Kühne had responded to them by 
suggesting they look at the IETF process. He would do this when putting 
together the draft proposal. He had also asked Athina to come up with 
ideas for how CoC members could be legally protected from retaliation, 
as inspired by the arbitration process. He would put this on the next 
agenda.

*7. Actions*

*- Leo to draft a proposal for the process and share it with Antony and 
Athina before sending it to the mailing list*

-------------- next part --------------
An HTML attachment was scrubbed...
URL: </ripe/mail/archives/coc-tf/attachments/20211104/e294bbd8/attachment-0001.html>