Post by @benpate@mastodon.social

DanSup is @ FediCon

@dansup@mastodon.social 3 wk. ago

@benpate I am very interested in this, I can see this being very useful for Loops and Pixelfed!

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

@dansup Awesome.. just one more thing to hash out over beers 🍻

Evan Prodromou 🇨🇦🇺🇸🇬🇷🇵🇸

@evan@cosocial.ca 3 wk. ago

@benpate @scottjenson @DePemig @thisismissem @johannab @oli @laurenshof @wjmaggos @smallcircles @GavinChait @iftas @tchambers I added an issue here:

https://github.com/EmissarySocial/fep-endorsements/issues/6

It's good to have a collection of activities, also, so you can `Undo` them.

GitHub

Access to Endorse activities · Issue #6 · EmissarySocial/fep-endorsements

For the ActivityPub API, it would be good to have a collection of Endorse activities that someone has sent or received. These can be used for Undo activities.

Evan Prodromou 🇨🇦🇺🇸🇬🇷🇵🇸

@evan@cosocial.ca 3 wk. ago

I added a few more questions. I think this is an interesting piece of work!

julian

@julian@activitypub.space 3 wk. ago

@benpate@mastodon.social I'm envisioning a sort of "DNS-like" endorsements flow where I as a server want to know whether this "Ben Pate" fellow is trustworthy, so:

My software asks the only person it trusts, activitypub.social, who responds with "no clue, but I trust fosstodon.org, so ask there"
So it asks fosstodon.org, who says "no clue, but I trust mastodon.social and someothersite.social, so ask them"
So it asks mastodon.social and it says "yes I know Ben Pate, he's a pretty cool fellow"

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

@julian This is a fantastic visualization.

I haven't detailed it well in the early draft, but this is where I'm heading. You might implicitly endorse your server's account. Servers might endorse each other (after some vetting process).

The scenario you described would make me your 3rd-level endorsement (to temporarily use LinkedIn vocabulary).

Then after we go out for beers at FOSSY, we might endorse each other as 1st-level contacts. Others could use that as a stronger signal of who to trust.

julian

@julian@activitypub.space 3 wk. ago

@benpate@mastodon.social I guess there are two parts... the backend establishing of trust — perhaps the trust score goes up the more content is received... although a case could be made that this ought to be left as an exercise for the implementor...

And the proactive query "do you trust xyz"... might be two FEPs

I'm wary of tossing around Endorse activities, because of all the sync issues endemic to ActivityPub in general, which is why my mind immediately went to some queryable API...

However that's not so much AP

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

@julian

All of you are so far ahead of me on this already -- working out how to make this real before I've even made a full FEP. It's awesome.

There MIGHT be a way to query it. We could lean on the proposals for querying collections, and get that done with a minimum of paiin.

But the more data we're talking about, the more likely that I might want some way to sync up ALL of the endorsements across an entire server, to do some kind of big data across them.

What do you think?

Golemwire

@golemwire@social.golemwire.com 3 wk. ago

@benpate@mastodon.social @scottjenson@social.coop @DePemig@social.coop @thisismissem@hachyderm.io @johannab@cosocial.ca @oli@olifant.social @laurenshof@indieweb.social @wjmaggos@liberal.city @smallcircles@social.coop @GavinChait@wandering.shop @evan@cosocial.ca @iftas@mastodon.iftas.org @tchambers@indieweb.social

This reminds me of human.json. We need more things like this to help us tell humans from hidden bots!

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

Yes. I didnt know about human.json, but its very similar in spirit. Just translate from IndieWeb -> ActivityPub

Do you know of any prior art that helps quantify the NATURE of a relationship? (In-person, online, coworkers, i admire their music, etc) Thats the next big hurdle I think we’ll need to cross.

@golemwire

Tim Chambers

@tchambers@indieweb.social 3 wk. ago

@benpate @scottjenson @DePemig @thisismissem @johannab @oli @laurenshof @wjmaggos @smallcircles @GavinChait @evan @iftas

Love this, seems like a structural part of in essence the beginnings of a data source for a descentralized trust score of sorts that users are legit and follow-worthy... Did I get that right?

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

Yes. And thank you for reading this 🙏🏻

One thing about a *decentralized* trust score is that it is personal - you would get different results than me because you might trust different people (in different domains) than I do.

The draft hints at this, but doesn’t really address use cases yet. I’ll try to make updates, but most implementation will be left up to devs themselves.

@tchambers @scottjenson @DePemig @thisismissem @johannab @oli @laurenshof @smallcircles @GavinChait @evan @iftas

Ben Ramsey

@ramsey@phpc.social 3 wk. ago

@benpate @tchambers This reminds me a little bit of PGP’s “web of trust” concept.

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

Yes. It is exactly that, with extra metadata.

What do you think we could do to: a) improve on this idea, and b) foster adoption around the Fediverse?

@ramsey @tchambers

1+ more replies

william ⁂ maggos

@wjmaggos@liberal.city 3 wk. ago

@benpate @scottjenson @DePemig @thisismissem @johannab @oli @laurenshof @smallcircles @GavinChait @evan @iftas @tchambers

I'm glad you're trying but I'm generally not a fan of adding complexity, esp when most people already can't handle choosing a server.

The fedi I want looks like the web. We'd be able to follow "horrible" people and not get cut off from also following friends and the "best" people on other servers. the problem is norms not tech.

I'm not a bad guy...

https://fba.ryona.agency/?domain=liberal.city

fedi-block-api liberal.city

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

Yeah, it gets complicated really fast :)

According to one of the listings from your link, they're blocking liberal.city because: "crypto-fascists" 🤣

@wjmaggos

william ⁂ maggos

@wjmaggos@liberal.city 3 wk. ago

@benpate

yea most of those don't matter. tiny servers blocking for whatever reasons. having a long list here isn't a big deal when you factor that in. fine.

it's the bigger servers like newsmast and hachyderm and then when I find out a friend is on a smaller server that limits/blocks mine.

and why? what am I doing that impacts their users? a server is an obvious problem for all of fedi when it lets their users spam other servers' users with shit, but outside of that, I do not get it.

Gavin Chait;

@GavinChait@wandering.shop 3 wk. ago

@benpate I think this is useful, but needs to be calibrated against the trustworthiness of the instance the endorsement is coming from. A well-moderated instance being more trustworthy than one which "isn't". But that then creates a layer of complexity between individual endorsements. However, at the instance level, perhaps a method to tune what endorsement messages will be accepted (i.e. an accept-list, rather than a block-list), and go from there?

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

@GavinChait

To your point, Gavin, @ozoned recommended that servers be able to "Endorse" other people or other servers themselves.

If our model assumes an implicit endorsement from me to my server, then *its* endorsements would automatically count as my own 2nd-level endorsements.

So yes, this is going to make some interesting math, and we'll have to recalibrate how to navigate this once real datagets published.

Gavin Chait;

@GavinChait@wandering.shop 3 wk. ago

@benpate @ozoned You mentioned bad actors, & we have to be real mindful of this. The old spy novels (Forsythe, Le Carre) used to talk about spycraft of setting up wholly fictitious, but perfectly "real", identities which could be picked up when needed. Entire bureaucracies dedicated to moving paper around to give the impression of a real life. Of course, if you have social media with networks of endorsements ... well, you can automate this.

Gavin Chait;

@GavinChait@wandering.shop 3 wk. ago

@benpate @ozoned You mentioned bad actors, & we have to be real mindful of this. The old spy novels (Forsythe, Le Carre) used to talk about spycraft of setting up wholly fictitious, but perfectly "real", identities which could be picked up when needed. Entire bureaucracies dedicated to moving paper around to give the impression of a real life. Of course, if you have social media with networks of endorsements ... well, you can automate this.

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

@GavinChait @ozoned

Agree wholeheartedly. @scottjenson had some great initial ideas on this. Here's some of the controls he suggested as a starting point:

- limit the total # of endorsements one person could make
- rate limit the speed that actors could make endorsements

Enforcement is up to the listener. Looking at who and how you've endorsed, do I believe you? With this, a botnet with 1000 self-endorsing sock-puppets could be identified automatically and silenced.

1+ more replies

Gavin Chait;

@GavinChait@wandering.shop 3 wk. ago

@benpate @ozoned For hopsauna, I've been thinking about exactly this. Before release, every product must be endorsed by allocated "editors" who review & authorise it. I was going to produce my own namespace term for this, but happy to use yours. However, my implementation was going to be different. An instance "recognises" an editor (an Actor), and the editor is responsible for endorsing a product (also an Actor).

2+ more replies

Gilles DePemig :TwinPines:

@DePemig@social.coop 3 wk. ago

@GavinChait @benpate I do not know if I understand this correctly, but I think an app/client would be able to summarize trust for the "trustee" based on how "trusted" their "trusters" are, cascade-wise starting from the observer (you). Something like "This person is trusted by 2 of your first level trustees and by 13 second level trustees."

Such would naturally reflect the trustworthiness of any server or person, from your own perspective.

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

Theoretically, yes. Setting aside what could become an enormous mathematical challenge, the center point of each person's "web of trust" is themselves, and it branches out in a shape that is unique to them.

This is all public information (and consensually opted-in) so you could (theoretically) get a pretty good sense of a person from their endorsements.

Your online reputation, therefore, becomes what you make it, based on the affects you have on those around you.

@DePemig @GavinChait

Gilles DePemig :TwinPines:

@DePemig@social.coop 3 wk. ago

@benpate @GavinChait Almost like life!

ozoned

@ozoned 3 wk. ago

@benpate@mastodon.social reading these FEPs and seeing definition of terms always reminds me of reading laws and how things have to be clearly defined. Because communication is so inefficient, especially in text, interpretation is so easy to get incorrect compared to original intent. But it makes new appreciate all the work even more. I absolutely love this idea and greatly appreciate all the work you and others do. Thank you.

Ben Pate 🤘🏻

@benpate@mastodon.social 3 wk. ago

@ozoned Hey, I’m glad you found this - and sorry I left you off the original list of tags. It wasn’t intentional.

Many of th ActivityPub specs are intentionally vague. There’s power in that, too, but it’s a design decision that has made interop… frustrating.

A lot of the recent additions, and all of the proposals I’m writing, are taking this other approach of spelling things out in very clear terms. Hopefully it helps implementors 🤓

ozoned

@ozoned 3 wk. ago

@benpate@mastodon.social lol dude, I'm not bothered at all by it. 😁 I figured those folks actively contributed to the spec. I take no offense at all. I'm just happy I get to be a fly in the room man. 😊