Talk:Requests for comment/wikimedia-l-post-limits

Identification

For proposal 4, there is some quite understandable concern regarding an identification process (and privacy policy), especially any 'verification' of identities, which again is understandable given I used the word 'verify'. I hadnt intended on it being a sleuthing activity, but didnt put much effort into describing what I had in mind. This oversight on my part was largely there are very few email list participants that it would pertain to anyway. So I appreciate being pulled up on it; thanks MZMcBride for your colorful approach ;-) Allow me to clarify, and suggest some mechanisms which might avoid the problems that could eventuate if a verification process was implemented badly, but also please suggest better ones.

It is important to re-iterate that the proposal is to require either a real world identity or a Wikimedia identity, and only when a person is reaching/exceeding five posts to the mailing list. It is a very small cohort, especially after the first few months when we would be processing the existing regulars who have thus far avoided linking their email account to an identity that our community knows them as elsewhere.

If someone emails using an identifying email address, such as a journalist emailing from their work email account, or any other email address which is publicly linked to a real world identity, identification is automatic (and the mailman list subscription system does the verification for us).

Likewise for anyone using a Wikimedia username in their email address or signature or similar, identification is effectively implicit; if they are doing a w:joe job, the community will fairly quickly detect it as the identity being claimed is part of their public email posts. The subject of a joe job will typically hear about it very quickly via the grape vine, and the list admins would be notified promptly, and problem solved. But if the grape vine doesnt work, anyone can check by sending an email using special:emailuser to ask the Wikimedia username to confirm they are sending the emails. That is the process the English Arbitration committee uses(used?) to avoid w:joe jobs. This is the process that I was thinking of when I wrote 'verify'.

The remaining cohort is people using non-identifying email address, such as a gmail address not part of a public profile, and does not include identifying information in their signature, the proposal is that we would be asking them to provide us with either

1. their Wikimedia account, or,
2. a real world identity

If they failed to do either, no problems, they are moderated after 5 posts per month. This approach was chosen so list admins dont need to be sleuths, and we're not banning potentially constructive posters (and avoiding putting them on permanent moderation, which tends to silence a person unless everyone is moderated, they feel targetted and rejected). This proposal is intended to be explicitly stating that undisclosed identities are not to be marginalised and subject to sleuthing or moderation/bans; instead they are expected to make their case effectively, within 5 posts.

I can appreciate the concern that list admins would be asking these questions, or doing some verification, and definitely didnt think about that concern in advance. Some options:

1. A very easy adjustment, and I think this is my preferred approach, is that list admins ask the correspondent to add either their Wikimedia account or real world identity to their next post to the mailing list, thereby allowing all list readers to sanity check the identity being claimed.

2. Another approach is to pass the problem over to stewards, and/or meta checkusers, if there is any concerns about the validity of an identity being claimed. We'd need either group to be willing to do that, which would likely mean we need to describe the probably scenarios in more detail to them.

3. Remove any reference to 'verify' from that proposal and explicitly state that undisclosed alternative identities are restricted to five posts until list admins have been told by stewards or an appropriate arbcom that the undisclosed alternative identity has been approved within their local community, effectively deferring to people closer to the situation.

More ideas welcome. John Vandenberg (talk) 10:54, 24 August 2017 (UTC)

Multiple people have now suggested to me that these proposals are being put forward primarily to address the "Rogol Domedonfors" poster. Would it make more sense to put Rogol on moderation instead? It seems like a much faster and more straightforward solution.

When a user is put on moderation, or prohibited from posting altogether, does that get noted anywhere publicly? --MZMcBride (talk) 04:39, 25 August 2017 (UTC)

Do you have anything to say to 'the "Rogol Domedonfors" poster' who is a (currently almost inactive) participant on this very site? Rogol Domedonfors (talk) 15:38, 26 August 2017 (UTC)

At least in the past, if a frequent poster was put on moderation, the list admins sent a message to the list. I think this is a healthy practice and should be continued.--Ymblanter (talk) 05:49, 25 August 2017 (UTC)

Yes, we are committed to publicly noting admin-initiated moderation. (There have been cases of self-requested moderation, which we accommodated without public comment.) Ijon (talk) 06:16, 25 August 2017 (UTC)

Just to be clear regarding "and the mailman list subscription system does the verification for us" - it's most definitely possible to spoof email addresses in mailman. Legoktm (talk) 10:39, 26 August 2017 (UTC)

Hi user:legoktm, do you mean it is possible for someone to send a fake email to the mailing list that looks like it originated from an email address that is already subscribed to the list?

Or it is possible to subscribe to the list with an email address and forge a confirmation of that email address without having access to the email which mailman sends to the subscribed email address?

Or something else? ;-) The former is a known problem with a phabricator ticket about it. I hope the second option isnt possible; if it is, I'd like to know more about it. John Vandenberg (talk) 11:16, 26 August 2017 (UTC)

The former. I'm not aware of the latter being possible. Legoktm (talk) 18:09, 28 August 2017 (UTC)

Outcome of this RfC

I'm sorry I may have missed this in the last few months, but what was the outcome of the discussion? Ckoerner (talk) 21:42, 6 October 2017 (UTC)

@Austin, Ijon, Esh77, and Jayvdb: any update here? I don't recall one on the list either, but may have missed it. Effeietsanders (talk) 05:53, 5 November 2017 (UTC)