Notices by Prizmo (unsafelyhotboots@sharkey.world)

@evan@cosocial.ca @alice@lgbtqia.space I am glad to see you are thinking through trust and safety - the posts detailing that you use IFTAS as a source for truth is a little reassuring. With that said, I am gonna push back on this because this is a clunky answer that absolutely will not scale if the Fedi takes off at all.

There are moral pieces that other people are airing in a much clearer way, so I will not touch on those beyond acknowledging they are more impactful and important to work through than my argument.

Instead, I want to aim at your costs and saving your future self a lot of work. Solving the opt-in issue now so every part of the service is opt-in solves your consent problem you faced today - which your current problem does solve - and solves the consent problem at scale. As well as somewhat resolving the T&S issues people are currently airing, and saves you on infrastructure costs.

@evan@cosocial.ca I am glad to see that Alice raising their point about the hashtags they cooked for tracking their post has helped others.

This answer is sidestepping my question about how you would mitigate or prevent harm from harassment/dogpiling/other forms of abuse that use hashtags and puts the onus on the person being harassed to notify you as the person who runs the service.

There isn't a clear answer here - I acknowledge that. This harassment or abuse problem isn't hypothetical though - it happens on most other social platforms, so it really is a matter of when, not if. Asking how to handle this abuse now means that when it happens, you aren't blindsided.

@evan@cosocial.ca @alice@lgbtqia.space Before I dig in, I want to say thank you for the work you have done on ActivityPub, and I understand the propogation issue with hashtags is a big deal.

To your answer directly, it hasn't happened that you know of, and the fix in your eyes is a retract message / announce that the receiving server can ignore. If they ignore it, the harm continues to the person whose nudes are floating around the Fedi, amplified (even briefly) because of your service. As someone looking at running a Bonfire server, I look at that as a pretty big liability.

The point everyone here makes is that setting up servers as opt-in solves the above issue as any server that opts in, you can take a few minutes to check them out and see what content is circulating on that server (most bad actors are just obviously bad). That solves Alice's issue (as her server likely wouldn't opt-in), as well as giving you a second to catch legally problematic content before it crosses your server.

Obviously you and I are not lawyers so my point about potential legal issues is a hypothetical. I am curious about how do you see the trust and safety side with your service? Do you have a plan for how to mitigate harm if your service is used for harassment or other awful troll behaviors?

@evan@cosocial.ca @alice@lgbtqia.space man, the fact that she had to make a big stink to get stuff pulled from the service is in and of itself a big problem. How do you handle it when your service propogates non-consensual nudes or worse?

@ryanc@infosec.exchange DHCP configuration?

I am so stoked to see @bonfire@indieweb.social finally releasing Bonfire Social. A long time ago I said that the Fedi client that prioritized trust and safety over handy design things (quote posts, markdown support, etc) will be the one that causes the Big Tech social platforms to completely collapse and looking through Bonfire's moderation tools they advertise (and who they worked with to understand what tools are needed) this seems to tick all the boxes.

Hats off to the Bonfire team, and I am excited for my next payday when I can sign up for a hosting class!