@misc filed Fedora-Council/tickets ticket #393. Discuss here and record votes and decisions in the ticket.
I agree that they should, but in practice thatâs very hard to do, particularly since people can create new accounts to evade the ban. The right answer is probably âuse services that authenticate against Fedora Accounts so that we can enforce bansâ.
2 Likes
Having 2 personal free accounts on Github is a ToS violation, you have to pay for 1 of them ( point 3 of the Terms of Services). So it shouldnât be too hard to convince Github to ban someone platform wide if repeated bans evading happen (and thatâs a consequence not everybody is ready to face, especially if the main account get banned as well).
Also, the same apply to FAS, people can create fake accounts to evade bans.
I think we should encourage official Fedora repositories and etc., to be linked to services where we can use Fedora accounts for authentication.
1 Like
In theory, but it assumes that we can prove the case to GitHubâs satisfaction. I doubt theyâd just take our word for it.
Sure, but thereâs less chasing down third-party services to enforce ban evasion.
Basically, what it comes down to is that trying to enforce bans on third-party platforms where Fedora has no official relationship adds a lot of burden to enforcement that could be avoided.
That do not give me great confidence if we canât enforce the CoC, because it would adds a lot of burden.
So counter proposal, we should make it clear in the CoC and say it does not apply to those platforms, cause thatâs not the impression I get from reading the text.
If there is spaces where we are not going to act and where we know we are not going to act because âit adds a lot of burdenâ, it should be documented. Or it should be clear they are not official (eg, not use the logo or the name), because this is misleading.
The Code of Conduct does apply, but our ability to enforce it is limited. This is additionally complicated because most cases are âtimeoutsâ where people are only temporarily suspended, so theyâd need to be unbanned after a period of time.
If you have a way that we can enforce these bans that doesnât involve someone having to manually go to each third-party service provider (or moderators of individual channels on those platforms) to notify them of changes, then Iâm all ears. But otherwise, I donât see a practical solution here.
For github, which is the one I had in mind when I opened the ticket (and I already listed the lists of orgs), there is a API:
There is a need for some org wide account to act on the banning/unbanning, but that would be a one time setup.
For the others, I guess we need to start to have a list of those platforms.
Some automation can be created for Reddit, given there is something to associate your account with your Fedora account ( https://verify.valkyrja.app/ ).
For Telegram, this can be a bot (like zodbot can silence IRC wide), and ask people to register on the bot like for Reddit.
I think itâs pretty futile to try and automate this. Thereâs lots of sites on the internet, and lots of different parts of fedora have presences there.
Even on github all those orgs are controlled by different parts of fedora. You would need to setup api calls and credentials for each one after talking to each of those teams to get access. This would get used⌠once every 5 years?
Iâd say if you see someone banned in fas interacting with a fedora community on another site, just bring that fact to the attention of whatever part of fedora that is to act on.
That wouldnât work, because Fedora seems to have a informal policy of not publicly saying someone is banned. So unless you are one of the few peoples who is aware of a ban, you canât really see much.
This also rely on the good faith of whoever is in charge of managing that space, who may disagree with the bans and simply not enact it (for example, this was the root of the Fram controversy on the English Wikipedia, and I suspect it might have happened elsewhere), or start to be noisy about it.
1 Like
well, I suppose it gets down to what you consider the âfedora communityâ for this. If thats just services/things we directly control or everywhere some fedorans are interacting.
I think itâs pretty impractical to tie every external space fedorans interact into this.
Just look at the github projects/orgs you mentioned, they are all run by different subsets of fedora folks and banned users may be using different usernames/whatever on each place to add to the complexity.
Anyhow, Iâll stop now and let others chime in.
I think this is a valid point to bring up, and I believe there is probably some middle ground that we can find as a resolution. Some thoughts:
- if person X has received a permanent ban, I think it makes sense that we might remove them from select Fedora spaces outside of things authenticated by FAS. For example, our Fedora repos on github and gitlab kinda make sense to me as there is a chance person X has some permissions which they could exploit to create further issues for Fedora. Maybe that guess is wrong, but it might make sense to come up with a list of which services outside of FAS in which there is some risk associated with their access. We could organize a way for one or two people to have the permissions in all of the correct spaces to be able to enact the ban as confidentially and quickly as possible.
- telling every single Fedora moderation team(10+ telegram channels, reddit, etc) to ban person X does not seem like something we want to do. It would require disclosing the identity of person X to at least 30 people or more based on my very rough estimate of who is doing Fedora moderation. Although person X may not be welcome in the Fedora community any longer, imo, we have no desire to make the banning action public knowledge. Once you have moved beyond notifying a small number of people, I believe it would become common knowledge. It could lead to shaming, people taking sides without having all the knowledge, etc etc etc, which is all very messy and has the potential to lead to more CoC incidents
- to the end of making our âsatelliteâ Fedora spaces safe and up to our CoC standards, I have started work on a set of Moderation Guidelines for Fedora [0]. I believe most of these spaces do have some moderation guidelines, but we want it to be cohesive across the board. My point is, if person X is in one of those spaces and causing issues, then the moderators should have the tools to remove them. If person X continues to create problems in those spaces and the moderators have no options left, then it should move to the CoC repo. If person X is in those spaces and causing no problems, then I think it becomes mostly a non-issue. If someone were to notice a banned person was active in a Fedora space, they could open a CoC ticket to see what can be done.
- last but not least, I think it is important to document what we will and wonât be doing when someone is banned. I will be incorporating that into the CoC Supplemental Documentation [1] that is in progress and will be reviewed and approved by the Council. It will also be open for community comment for some time during the process.
[0] Issue #392: Fedora Moderation Guidelines - tickets - Pagure.io
[1] Issue #391: Fedora Code of Conduct Supplemental Documentation - CoC - tickets - Pagure.io
1 Like
So I do not know if anyone ever been banned from Fedora, but if that did happen
I like that attitude. Thinking about banning people for good without any precedence. Looks like someone has a plan. 
In any case I think it would be helpful to implement a protocol like https://lumendatabase.org/
IBM just disabled my account few weeks ago, because I am in Belarus and have no funds to leave. 
In a month all my projects will be deleted there. I can not login to backup them, and nobody will know about it.
If you will have to ban me from Fedora, because it is a Red Hat sponsored projects, and Red Hat is owned by IBM, I would like at least have an official memorial.
abitrollybanned from all Fedora services with account erased. Reason: failed to leave a sanctioned country of origin.
2 Likes