The W3C Credentials Community Group

Meeting Transcriptions and Audio Recordings (2014-today)

Go Back

W3C CCG Weekly Teleconference

Transcript for 2024-02-13

<kerri_lemoie> Hey all
Our Robot Overlords are scribing.
Dmitri Zagidulin: Harrison if you're speaking we can't hear you unless that's just me.
<bumblefudge> i also no hear
Kimberly Linson: Now it's me too I just texted him to see if he.
Kimberly Linson: If he wanted me to get started so I I will do that um so that we leave lots of time for you Demetri and I think what will be a very good discussion um.
Kimberly Linson: Hi everyone uh welcome to this week's uh credentials community group meeting uh let me pull up my little.
<harrison_tang> that's weird
Kimberly Linson: Agenda cheat sheet um but I know that the very first thing on our agenda is to uh talk about professional conduct and just make that reminder that we are all uh striving towards the same goals um to make this technology accessible and usable.
Kimberly Linson: And uh if you are interested in reviewing that uh policy in more detail it is uh Linked In the agenda.
Kimberly Linson: If uh we definitely are glad you're here I see lots of folks on the call today um so if you are new to this group um we are very very happy that you're here and uh you are welcome to attend these meetings um and not be a member of the ccg but just uh come and join if you are interested though in um contributing to the work becoming a more active participant then we would ask that you go ahead and uh join.
Kimberly Linson: The W individuals you can so you can check that out the links to that are also in the uh agenda.
Kimberly Linson: We uh.
Kimberly Linson: We keep minutes uh and recording of this of this meeting for uh future future interests and to keep an archive of what we've what we've been talking about um and we managed it uh via we manage the queue and the participation of um of.
Kimberly Linson: Contributions using a queue system so if you're interested in have a question or have a comment on something that someone is saying then I would invite you to put Q Plus uh in the chat field and that will.
Kimberly Linson: Um allow you 1 of the things that Harrison and I will be doing in well if he's on here too um we'll be monitoring that queue and making sure that everyone has the opportunity to contribute.
Kimberly Linson: Let's see if you're done speaking and you don't want to uh you want to take yourself off the queue you can use Q minus.
Kimberly Linson: And uh that brings us I believe to um.
Kimberly Linson: Introductions and reintroductions so if you are new to this community or you haven't been here for a while and you'd like to uh.
Kimberly Linson: Introduce yourself or say hello to the group um go ahead and put yourself on the Queue and I will.
Kimberly Linson: Okay thank you um well no 1 No 1 uh wants to to introduce or reintroduce themselves but I'm hoping that someone uh might have uh an update or uh a community sort of notification that they want to share with us uh so that is our next section.
Kimberly Linson: Updates and reminders.
<dmitri_zagidulin> sign up for the April IIW! :)
Kimberly Linson: Demetri everyone is very excited for you to to have plenty of time.
Kimberly Linson: And yes sign up for the April iiw if you haven't already.
Kimberly Linson: All right any questions or comments about work items.
Kimberly Linson: All right well that brings us to the end of our housekeeping notes and so I will go ahead and turn it over to Dmitri.
Dmitri Zagidulin: Thank you so much uh really excited to be here uh okay so uh.
Dmitri Zagidulin: Oh let me just share my screen as well.
Dmitri Zagidulin: Uh so some welcome everyone uh.
Dmitri Zagidulin: We figured that the Social Web community group another 1 of uh w3c CGS uh has a lot of overlap with uh what we're doing here the ccg so we wanted to come visit.
Dmitri Zagidulin: Uh tell tell everybody a little bit about what we're doing in Social Web CG and invite you to join the conversation uh.
Dmitri Zagidulin: So hold on 1 second let's join.
Dmitri Zagidulin: Or rather let's share screen is it sharing it is not sharing.
Dmitri Zagidulin: Ah 1 second let me reconnect.
Kimberly Linson: Demetri we see your slides but we are not hearing you if you're speaking.
Kimberly Linson: Still nothing I wonder if you want to try going out and coming back in again.
Harrison_Tang: It might be the microphone setting.
Kimberly Linson: And there you are.
Juan Caballero: Should I should I do my best to meet you impersonation.
Juan Caballero: I was.
Juan Caballero: Helping him with.
Juan Caballero: For the meeting.
Juan Caballero: Because I just like.
Juan Caballero: Uh there um and I think maybe she doesn't hear us telling of his uh.
Kimberly Linson: That that we can't are you able to like text him or something and tell him okay.
Juan Caballero: Yeah yeah that's a I'll text them uh you.
Bob Wyman: You might suggest that it try if you can't get the computer working try to come in on the phone for the for the audio.
<harrison_tang> @Dmitri It might be your microphone setting in Jitsi. I had an issue with it earlier.
Juan Caballero: Yeah uh so um the the Social Web working group uh at w3c has a long complicated history and it gets a little um.
Juan Caballero: Uh it requires a bit of familiarity with the w3c nomenclature so it was a working group which you know would be the equivalent of like VC working group or dead working group uh for 4 years and the the sort of scope and intention of that working group was to um aligned or unify a really splintered landscape there were a lot of different communities and things working out there.
Juan Caballero: Um in the wild that wanted to have a more you know the the the wanted to have a shared and open platform across them um uh may maybe Evan would like to speak to this I wasn't there.
Juan Caballero: And and uh the the um sorry I'll try to share the slides or something I don't know what I don't know what Dmitri is doing over there um.
Juan Caballero: The um working group sort of uh shipped the the um.
Juan Caballero: Trcr everyone it's called Uh and went into hibernation I guess it it was the working group was spun down and the CG was spun up and sort of like governance transferred I don't know the exact mechanism or terminology for it but uh the community group became the steward of those CRS uh and.
Juan Caballero: For a few.
Juan Caballero: Mailing list was turned off there was there was sort of like very little activity at w3c and and most of the developer Community was working in like independent channels like Social Hub and codeberg.
Dmitri_Z._(pres.): Quick sound check.
Dmitri_Z._(pres.): Oh yes thank you deep apologies for that no idea why some cut off I just switched to different computer uh thank you so much Juan uh and and Evan for uh for speaking.
Juan Caballero: But you heard all of what we were saying so you know where where we are and how much we covered.
Dmitri_Z._(pres.): Uh I think so remind me so where where did you get to.
Juan Caballero: So so uh yeah the the CG has now become more active again in the last 2 years mainly let's turn back on uh meetings TPAC um and that was the slide we were trying to impersonate you on.
Dmitri_Z._(pres.): I'm deeply appreciate that and I probably did a better job anyways right so yeah so this this call is uh about us wanting to share uh a little bit about what we're working on.
Dmitri_Z._(pres.): Because as uh as we're going to get into the it's now becoming more important than ever.
Dmitri_Z._(pres.): Uh what else um Evan Juan uh what did I miss from the the history and the context.
Dmitri_Z._(pres.): Okay fantastic yeah.
Juan Caballero: Uh I think I think the I think the slide is good to just read you could just leave it up for a second and you would uh anyone who missed a section while we were babbling might catch up um yeah I think it's I don't think you need to run through the Bullet by Bullet um.
Dmitri_Z._(pres.): So let's get it.
Juan Caballero: That accurate factual background okay go ahead.
Dmitri_Z._(pres.): Excellent excellent yeah let's get into the juicy stuff uh specifically so what's the Social Web up to right now what are the challenges right what's hard.
Dmitri_Z._(pres.): And how does this intersect with the credentials community group with all of you here uh because we want you to come join the conversation uh we need as many eyes and and developers and users on this as as we could possibly get.
Dmitri_Z._(pres.): All right so.
Dmitri_Z._(pres.): Let's talk about what's hard and both challenges and opportunities really.
Dmitri_Z._(pres.): So due to various scandals like Cambridge analytica like the Twitter acquisition the renaming to acts uh various scandals with Reddit.
Dmitri_Z._(pres.): Existing social media there is.
Dmitri_Z._(pres.): A rising wider awareness of Need for Alternatives or at least for better interoperability for better oversight better motivation in some cases.
Dmitri_Z._(pres.): Uh to to put it put it another way.
Dmitri_Z._(pres.): we got some.
Dmitri_Z._(pres.): Uh you know New York Times and Wall Street Journal and wired and whatever else uh.
Dmitri_Z._(pres.): You might be reading so that's good that's good we have more more people sort of aware that hey not not everything is well.
Dmitri_Z._(pres.): as well as.
Dmitri_Z._(pres.): Media land and that possibly there are these Alternatives coming up that people are working on possible solutions.
Dmitri_Z._(pres.): Of course as always uh if you're interested in this and you start diving into the work as always and this should be familiar to us in the ccg from dids and verifiable credentials our documentation and onboarding could be better.
<bumblefudge> lol
Dmitri_Z._(pres.): What if you if you're interested in this you dive in you don't know where to get started please reach out to us uh reach out to me reach out to Evan to Juan to the other co-chairs we will be happy to guide you through.
Dmitri_Z._(pres.): Either as a as a user as a uh journalist or commentator as a developer and so on.
Dmitri_Z._(pres.): So since the technical recommendations were published in 2018.
Dmitri_Z._(pres.): Millions of users there's been uh dozens and dozens of different implementations.
Dmitri_Z._(pres.): In in multiple languages we've gotten a lot of feedback from users and Developers.
Dmitri_Z._(pres.): We have a lot more experience.
Dmitri_Z._(pres.): And so just like with the credentials working group The credential working group 1.0 finished up.
<bumblefudge> And also reach out to Bob W and Emelia and Laurens and Matthias and Peter Z who are here!
Dmitri_Z._(pres.): It was already time to uh to start up granches group 2.0 and so on so we're in a in a similar sort of space uh we're currently.
Dmitri_Z._(pres.): In the process of.
Dmitri_Z._(pres.): Ing curating uh updating the very specifications uh that are needed to write social software uh in this in the Social Web CG.
Dmitri_Z._(pres.): What else has happened well there's.
<bumblefudge> the CG is very horizontal, you don't need to talk to the chairs and editors
Dmitri_Z._(pres.): There were some well publicized launches.
<emelia_s.> Thanks @bumblefudge
Dmitri_Z._(pres.): Since then a relatively smaller 1 from Blue Sky uh partly funded by uh Jack Dorsey of Twitter as sort of like a.
Dmitri_Z._(pres.): As as a aspiring Twitter killer if you will although lots of people in the Federation activity Pub Community would definitely argue with that and and point out that hey we we were there before.
Dmitri_Z._(pres.): We had millions of users before Blue Sky came along but it doesn't matter everybody's working towards the same general goal which is we want Alternatives we want things to be open and standardized we want things to be portable.
Dmitri_Z._(pres.): Facebook or or rather the parent company meta.
Dmitri_Z._(pres.): Announced that it's long it's launching its own.
Dmitri_Z._(pres.): Sort of Facebook alternative threads.
Dmitri_Z._(pres.): Or Facebook style Twitter alternative if you will.
Dmitri_Z._(pres.): But most importantly.
Dmitri_Z._(pres.): 1 of the.
Dmitri_Z._(pres.): 1 of the open specifications the the technical report that came out of the Social Web working group and is now being stewarded by the community group so as you can imagine there were a lot of discussions both in the existing fediverse community in uh wider developer community in general so things are getting serious.
Dmitri_Z._(pres.): As you very well know in the ccg.
Dmitri_Z._(pres.): Authentication identity authorization still very difficult.
Dmitri_Z._(pres.): Uh and although.
Dmitri_Z._(pres.): We've got some more Tech to experiment with to try and integrate to see if we can if it can help our efforts in the Social Web space.
Dmitri_Z._(pres.): And aside from uh.
Dmitri_Z._(pres.): Basically stuff you can sign.
Dmitri_Z._(pres.): We have very real and pressing uh challenges with with moderation and portability uh challenges that the community and and the community group are working on actively uh that we hope for your help in.
Dmitri_Z._(pres.): What's hard why isn't everyone using the fediverse well in the relative scheme of things.
Dmitri_Z._(pres.): So pretty new.
Dmitri_Z._(pres.): Uh just a lot of people just haven't heard of the fediverse a lot of people don't know the name uh Mastadon or activity Pub or fed averse or anything like that right so we have a lot to go uh we have far to go uh publicity wise.
Dmitri_Z._(pres.): but also.
Dmitri_Z._(pres.): Work to do.
Dmitri_Z._(pres.): Development and standards side.
Dmitri_Z._(pres.): For those very motivated to switch from traditional social media.
Dmitri_Z._(pres.): Decentralized social media.
Dmitri_Z._(pres.): Uh they they do run into their real problems of hey traditional social media.
Dmitri_Z._(pres.): Builds a moat around.
Dmitri_Z._(pres.): The followers graph.
Dmitri_Z._(pres.): Have a list of followers or friends on 1 platform and I want to migrate to another.
Dmitri_Z._(pres.): There's no way uh traditional companies are like oh sure yeah let us let us helpfully map your uh your followers from from 1 platform to another so that when you migrate you don't have an empty list right let let us ease the transition that is the exact opposite of uh the the current business model and so that presents presents an interesting challenge during waves of migration away from Twitter we saw some really ingenious solutions to this where people put in their profile description hey uh I'm off of Twitter but I'm over here on the fediverse and then people build uh scrapers and tools that.
Dmitri_Z._(pres.): Index people's profiles and and you could put in your.
Dmitri_Z._(pres.): Your old username.
Dmitri_Z._(pres.): and it.
Dmitri_Z._(pres.): Would say okay.
Dmitri_Z._(pres.): Of your friends be these are the ones that migrated here's where you should sign up with them right so so some very Creative Solutions.
Dmitri_Z._(pres.): That kind of had to hack around this lack of uh portable social graph and that that is 1 of the things that we're trying to.
Dmitri_Z._(pres.): Um to to solve here.
Dmitri_Z._(pres.): Uh in a CG.
Dmitri_Z._(pres.): Uh the next major problem is that okay so say as a user you've heard about this decentralized Social Web you're like okay I want to try it out.
Dmitri_Z._(pres.): you go to some.
Dmitri_Z._(pres.): You run into the various usability challenges in the decentralized world such as.
Dmitri_Z._(pres.): you're not.
Dmitri_Z._(pres.): Case with the choice of okay uh which which server.
Dmitri_Z._(pres.): Which uh community of users do you want to join now the whole thing interferes.
Dmitri_Z._(pres.): I from 1 server can follow everybody else on other domains other servers that's not the problem.
Juan Caballero: if people want to see the "daunting first step" for themselves :D
Dmitri_Z._(pres.): But they but I still have to pick 1 I I still have to create an account somewhere before joining uh and and that the the choosing of the server has traditionally been a fairly difficult problem um a friction barrier for users to overcome although that's gotten better in recent years for example um some of the flagship.
Dmitri_Z._(pres.): Implementations like Mastadon like Firefish like um.
Dmitri_Z._(pres.): Have started recommending like.
Dmitri_Z._(pres.): Yes you should choose um you should choose a domain you should choose an initial instance server.
Dmitri_Z._(pres.): If you don't have strong opinions here's 1 you might start with.
Dmitri_Z._(pres.): So this is why for example a lot of people following from I don't know New York Times article or a wired article they hear about mastadon.
Dmitri_Z._(pres.): Based on has started.
Dmitri_Z._(pres.): Recommending or at least.
Dmitri_Z._(pres.): Educating people that okay if you don't know what what else to join join Master on social.
Dmitri_Z._(pres.): Now this made onboarding easier but as you can probably imagine the this also uh resulted in some grumbling of okay but now.
Dmitri_Z._(pres.): To vendor locking.
Dmitri_Z._(pres.): It's still on.
Dmitri_Z._(pres.): We can get.
Dmitri_Z._(pres.): Back to that.
Dmitri_Z._(pres.): If people are.
<bumblefudge> a familiar problem to the wallet-chooser domain...
Dmitri_Z._(pres.): What else are we working on what what are the problems are we trying to solve in the Social Web uh.
Dmitri_Z._(pres.): Area access control is still hard for for various reasons right either encryption based access control or traditional authorization policy Access Control such as what solid or webdev or even Google Documents has.
Dmitri_Z._(pres.): Uh This Thread is only for my followers or my friends or whatever the uh the terminology is at that particular implementation.
Dmitri_Z._(pres.): Countries that we can.
Dmitri_Z._(pres.): Another thing that's on everybody's mind in the decentralized social space is of course moderation.
Dmitri_Z._(pres.): Which is very painful and very relevant these days uh because several countries like European Union the US others have passed legislation uh trying to combat.
<bumblefudge> terrorism content as well
Dmitri_Z._(pres.): Sexually explicit material material dealing with minors uh oh all sorts of things that should be moderated.
Dmitri_Z._(pres.): But by passing that legislation.
<bumblefudge> often more aggressively policed
Dmitri_Z._(pres.): Large instances large service providers and small ones uh equally but obviously they don't have equal uh means and funds to deal with it so there's a lot of interesting conversation there's a lot of friction there's a lot of.
Dmitri_Z._(pres.): Solutions being feverishly worked on in the area of moderation uh because well it's it's a problem everywhere.
Dmitri_Z._(pres.): The traditional mainstream social.
<emelia_s.> For CSAM, IFTAS is currently working on a solution, and I'm the tech lead for it.
Dmitri_Z._(pres.): Platforms just have entire Departments of people moderating and of course burning out and and having post-traumatic stress disorder and and so on and so forth so it's a huge problem in Tech in general.
Dmitri_Z._(pres.): It's definitely something that we have to deal with in uh Social Web space.
Dmitri_Z._(pres.): Is surprisingly difficult or maybe not surprisingly right well you've all you've all been in the area of decentralized system you know that.
Dmitri_Z._(pres.): Uh in any decentralized system is hard let alone 1.
Dmitri_Z._(pres.): That has any sort of authorization or Access Control.
Dmitri_Z._(pres.): So that's that's an ongoing conversation and and there's interesting Solutions there.
Dmitri_Z._(pres.): And then of course.
Dmitri_Z._(pres.): Why do we get in this in the first place because we want choices we want portability we want standards.
<bumblefudge> competing proposals/designs for hashtag-federation and global search were the main topic at the 2019 activitypub conference in prague that I attended...
Dmitri_Z._(pres.): And if I join 1 instance let's say it was the first 1 that the commanded I followed some article.
Dmitri_Z._(pres.): On social or.
Dmitri_Z._(pres.): Firefish social or something else.
<emelia_s.> There's not much I can disclose for now on IFTAS's CSAM detection tooling, but we're working on it actively
Dmitri_Z._(pres.): But I'm like okay this is great but uh here my friends started up.
Dmitri_Z._(pres.): Instance over here I want to join that.
Dmitri_Z._(pres.): The promise of the fediverse is that I'm able to do that I'm able to take my posts take my followers and following lists uh take my sort of identity and profile and settings and all that and and migrate to another.
Dmitri_Z._(pres.): Another server to another domain.
Dmitri_Z._(pres.): And it is possible but as you can probably imagine there's a lot of technical difficulties there's a lot of edge cases there's a lot of stuff for us to work through uh which is 1 of the reasons that um.
Dmitri_Z._(pres.): Social Web.
Dmitri_Z._(pres.): Group have start.
Dmitri_Z._(pres.): So why why were we especially excited to uh talk to this community here uh to the ccg.
Dmitri_Z._(pres.): Identity and authentication is difficult and a part of the Social Web Community is looking at DS.
Dmitri_Z._(pres.): Uh incubated by by this uh very ccg.
Dmitri_Z._(pres.): To see if that can help with some of the challenges there.
Dmitri_Z._(pres.): Uh the other direction that that's currently being worked on is an older uh protocol called webfinger uh that where usernames look like emails essentially uh and are used to discover the initial profile which contains keys and service endpoints should be familiar.
Dmitri_Z._(pres.): Some interest towards deads.
Dmitri_Z._(pres.): Due to the portability and migration problem in addition to access control in addition to moderation uh there's a lot of incentive to be able to sign everything to to use digital signatures data Integrity style to.
Dmitri_Z._(pres.): Signpost for again.
Dmitri_Z._(pres.): For audit logs for authorship for portability for lots of different reasons so we have a couple of in progress specifications the fap 390 and the ADB 32.
Dmitri_Z._(pres.): That are using some of the tech incubated in the ccg to apply it to the Social Web.
Dmitri_Z._(pres.): Another conversation that's just really starting up just just starting to get going is this notion of okay so if we can do did for Athenian Identity or something did like using um webfinger and these these actor profiles what about access control.
Dmitri_Z._(pres.): Wouldn't it be great if we had some sort of uh Access Control policy language or.
Dmitri_Z._(pres.): you know.
Dmitri_Z._(pres.): Uh authorization capabilities such as zap LD in in this group you can uh which is another uh basically equivalent to to zaps uh that's that's currently being done outside of w3c or maybe a uh synthesis of the 2 uh a best of all worlds.
Dmitri_Z._(pres.): And of course.
Dmitri_Z._(pres.): 1 of the challenges of.
Dmitri_Z._(pres.): Migrating content so I make a couple posts on 1 dumaine and I want to bring it to another what happens to.
Dmitri_Z._(pres.): To the links to the previous domain.
Dmitri_Z._(pres.): How rights also the links will have to be moved.
Dmitri_Z._(pres.): Be great if in addition to the classic domain based urls.
Dmitri_Z._(pres.): The posts that I make the comments that people leave on those posts the likes that people send had portable identifiers.
Dmitri_Z._(pres.): Uuids or better yet content addressed identifiers just a bit more in the purview of the ccg and so on so lots of tech need needs figuring out uh in that area as well.
Dmitri_Z._(pres.): Yeah so I want to leave lots of time for um other colleagues from uh the Social Web working group to to to weigh in on what they see is the.
Dmitri_Z._(pres.): Are the primary challenges the stuff that we're working on things that we need help with uh but I want to make sure to give the call to action uh if you find this stuff interesting.
Dmitri_Z._(pres.): come join.
Dmitri_Z._(pres.): During the social.
Dmitri_Z._(pres.): Community group during the mailing list I've got a user-friendly Forum uh if you're.
Dmitri_Z._(pres.): Then uh join the fep repo where some of these specs are being worked on as well.
<bumblefudge> ...contribute to a FOSS implementation!
<bumblefudge> hehe
Dmitri_Z._(pres.): And if you haven't tried it out uh dip your toe in the whole Fetter verse uh.
Dmitri_Z._(pres.): Space get a f account uh try it out yes you'll you'll encounter some of the challenges uh that we mentioned but.
Dmitri_Z._(pres.): It's also really really fun and and uh there's a lot of momentum in behind all of this.
Dmitri_Z._(pres.): So yeah so I'll pause here on this.
Dmitri_Z._(pres.): Uh on on the slide here and uh invite.
Dmitri_Z._(pres.): Um aan um Juan anybody else to to hop on the queue.
Dmitri_Z._(pres.): What else did I miss.
Dmitri_Z._(pres.): Bob please go ahead.
Bob Wyman: Yeah uh you kind of touched on it a number of times but a very specific thing okay that's relevant to the credentials um working group is is the fact that we need credentials um.
<bumblefudge> yeah
Bob Wyman: In profiles right now or in essentially it's a social system but there's no way within that social system of being able to find out who the heck anybody is um or or what their attributes are for instance if somebody is like flaming on about.
<tantek_çelik_(he/him,_mozilla)> lol what was that laughter?!?
<bumblefudge> me!
Bob Wyman: I don't know climate science I'd like to know that they're a professor at a university or they have a degree in the subject or they have some kind of qualification at the moment I I have no mechanism in the in the social network of of discovering that.
<bumblefudge> jitsi is wierd, man
<bumblefudge> it laughs when you click the laughing emoji-react
Bob Wyman: And uh so a lot of the a lot of the stuff that the credentials Works uh group works on I think we we need uh in in the Social Web not for commercial purposes but just so we know who the heck we're speaking to um.
Bob Wyman: Anyway I think uh the other thing if I could mention to sort of to add on you spoke about the issue with search you know people wanting to it's not a technological problem there to a great extent it's more a question of what what people want so it's a question of.
<bumblefudge> huge +1 to the "green checkmark" idea -- linkedin uses VCs to display who has a signature from an employer, fediverse could do the same
Bob Wyman: Of of Rights um 1 of the issues that I think needs to be dealt with is the question of of text and data mining.
<tantek_çelik_(he/him,_mozilla)> is there a link to notes or the slide deck?
Bob Wyman: Now for instance we have European a lot of the users of the social system do not want their content to be text or data mind um but we now have things like European laws that say that unless you explicitly state in your data uh some provide some sort of signal that uh that you're having text and data mining that uh a large number of people have an assumed right to do that uh irrespective of copyright law that's that's European law um so 1 of the things that we don't have is a way of making.
Bob Wyman: A statement.
Bob Wyman: About rights um particularly the the right to uh uh control text and data mining.
<dmitri_z._(pres.)> notes are auto-transcribed here in jitsi chat
<dmitri_z._(pres.)> and that gets published afterward
Bob Wyman: And that's that maybe isn't that relevant to this group although you know it would be good to have it as a as a credential uh the TDM rep or the Texan data mining representation.
Bob Wyman: I think it's representation uh working group with the w3c is is is probably a better context for that but it does relate.
Bob Wyman: The z cap and some other kind of similar issues and odl and that sort of thing anyway so but the the the important thing I want to state was in the social system in the Social Web we need credentials we don't have any credentials now we really need credentials.
Dmitri_Z._(pres.): Thank you so much Bob uh really great Point uh let's see who's up next Evan.
<bumblefudge> C2PA and content credentials have been mentioned, but I think it's early for standardizing
<kaliya_identity_woman> come to the FediForum!
Dmitri_Z._(pres.): Thank you so much yeah join join the task force is uh definitely need help there uh Amelia go ahead.
Dmitri_Z._(pres.): Oh oh sorry sorry.
Emelia_S.: Is at the moment.
Dmitri_Z._(pres.): I I just just saw kalia's uh point that uh feder form clear hop on the Queue and tell us about 34 I'm sorry go ahead go ahead Emilia.
<bumblefudge> when is the next fediforum? format is familiar to those that remember the online IIWs :D
Emelia_S.: Yeah uh so 1 Thing worth mentioning is that uh at the moment in order to participate in the fatty verse you do have a public private key pair.
Emelia_S.: Uh which is used for HTTP HTTP signatures there's also the identity through web finger and things like that.
Emelia_S.: So perhaps did or some other.
Emelia_S.: Related credentials specification might help with regards to that IE owning your credentials rather than uh.
Emelia_S.: having this.
Emelia_S.: Of a managed them.
<bumblefudge> yup, custodial keys and webfinger DO present a straightforward upgrade path/interop point for self-managed keys and/or identifiers
Dmitri_Z._(pres.): Thanks yeah that's a that's a really good point that the feds already uses cryptography for authentication and identity by just having a key pair so the part of the on-ramp part of the migration uh sort of proposal is okay what if we can translate those key pairs into DS and and go from there.
Dmitri_Z._(pres.): Let's see who's who's up next Ted go ahead.
<bumblefudge> no "breaking change" upgrade without an upgrade path is realistic IMHO
TallTed_//_Ted_Thibodeau_(he/him)_( So I'm hearing competing interests.
TallTed_//_Ted_Thibodeau_(he/him)_( To put it mildly.
<dmitri_z._(pres.)> competing interest? in social networking? :) :)
TallTed_//_Ted_Thibodeau_(he/him)_( This is all very much Reinventing the old uh nntp or Net News protocols and interactions and.
TallTed_//_Ted_Thibodeau_(he/him)_( which was.
TallTed_//_Ted_Thibodeau_(he/him)_( Sort of a.
TallTed_//_Ted_Thibodeau_(he/him)_( And infinitely replicate discussion threads across the internet uh.
TallTed_//_Ted_Thibodeau_(he/him)_( Pipe for that.
TallTed_//_Ted_Thibodeau_(he/him)_( Is enormous I don't think it is possible anymore to carry a full Newsfeed because the bandwidth that you need is more than the internet provides.
TallTed_//_Ted_Thibodeau_(he/him)_( setting that.
TallTed_//_Ted_Thibodeau_(he/him)_( We have wishes for searches and the reason we wish for search is because we want to find discussions that are of Interest.
TallTed_//_Ted_Thibodeau_(he/him)_( But in order to have a search that works we have to be able to quote unquote data mine because that's the only way to build an index.
<evan_prodromou> It doesn't work like NNTP.
TallTed_//_Ted_Thibodeau_(he/him)_( The devil would love to hate or hate to love because they have scraped so much.
TallTed_//_Ted_Thibodeau_(he/him)_( Data mined so much data.
TallTed_//_Ted_Thibodeau_(he/him)_( from a.
TallTed_//_Ted_Thibodeau_(he/him)_( Across the web.
TallTed_//_Ted_Thibodeau_(he/him)_( So that we can find information but we can't necessarily find the information that we're interested in because people have learned the tricks of keyword poisoning Pages such that things come up when we don't want them to.
<tantek_çelik_(he/him,_mozilla)> Dmitri can you share a link to the slide deck?
TallTed_//_Ted_Thibodeau_(he/him)_( And make it harder to find the information that would do want to find.
TallTed_//_Ted_Thibodeau_(he/him)_( Um the Federated discussion thing.
TallTed_//_Ted_Thibodeau_(he/him)_( Blue sky and the others uh the the sort of parallel micro blogging.
TallTed_//_Ted_Thibodeau_(he/him)_( It was semi standardized a decade ago when Twitter was really hot uh and when Twitter was killing all of its third-party client software.
TallTed_//_Ted_Thibodeau_(he/him)_( Killing the the API access so the third party software had no choice but to die.
TallTed_//_Ted_Thibodeau_(he/him)_( uh and.
TallTed_//_Ted_Thibodeau_(he/him)_( Their tools less usable.
TallTed_//_Ted_Thibodeau_(he/him)_( The same thing I hear is happening with what exists right now as far as the quote unquote Fedders and blue sky and others.
TallTed_//_Ted_Thibodeau_(he/him)_( They're sort of integrated but the presentations that they offer all differ.
TallTed_//_Ted_Thibodeau_(he/him)_( Similar data structures.
TallTed_//_Ted_Thibodeau_(he/him)_( It is very difficult to have a unified presentation the closest thing that happened.
TallTed_//_Ted_Thibodeau_(he/him)_( In the last 20 years was the RSS and atom explosion.
TallTed_//_Ted_Thibodeau_(he/him)_( Does have a semi standard format of data.
TallTed_//_Ted_Thibodeau_(he/him)_( but the.
TallTed_//_Ted_Thibodeau_(he/him)_( Is very very different depending on which quote unquote client application you use.
TallTed_//_Ted_Thibodeau_(he/him)_( And how usable those applications are varies wildly there are open source ones there are closed Source ones there are pay for it and not pay for it or pay for it with cash and pay for it with ads.
TallTed_//_Ted_Thibodeau_(he/him)_( It's all a jungle right now.
TallTed_//_Ted_Thibodeau_(he/him)_( And finding ways to.
<dmitri_z._(pres.)> ActivityPub is /very/ similar to RSS and Atom! Same sort of deal - standard data model, lots of different clients can innovate in the UI
TallTed_//_Ted_Thibodeau_(he/him)_( Not to pave that jungle but to tame it somewhat to make it easier to consume to make it easier to travel.
TallTed_//_Ted_Thibodeau_(he/him)_( That's where the game is right now and finding a way to make that pay.
TallTed_//_Ted_Thibodeau_(he/him)_( People who are.
TallTed_//_Ted_Thibodeau_(he/him)_( That's the people are devoted.
TallTed_//_Ted_Thibodeau_(he/him)_( The standards we.
TallTed_//_Ted_Thibodeau_(he/him)_( All need roofs and food.
Dmitri_Z._(pres.): Thank you so much Bob.
Dmitri_Z._(pres.): Uh clearly I think you're up next.
<tantek_çelik_(he/him,_mozilla)> Dmitri Z., I'm sure you mean ActivityStreams is similar to Atom, since it's an evolution thereof
Kaliya Young: Um sure I think Johannes joined the call um so maybe he can also speak to it but the fetty Forum.
Kaliya Young: Catalyzed by Johannes and I'm helping him with it um using as Juan said the format that we use for virtual AWS using open space technology.
Kaliya Young: The community of folks working in all things fed averse um connecting to each other Johannes do you want to.
Kaliya Young: Share more about your inspiration and what you hope can come out of it.
Kaliya Young: If he's here.
Kaliya Young: He told me he was here on the back Channel but maybe he's not.
Dmitri_Z._(pres.): No worries uh Johannes when when Dana joins us can hop hop on the queue.
Kaliya Young: Yeah but basically all things fed up for him and we make the agenda the day of the event so everything's welcome.
Dmitri_Z._(pres.): Yeah so that's uh it's a great conference great way to dive into uh Fiverr and the tech so so please join.
Dmitri_Z._(pres.): Uh let's see it's we've got Bob is that a is that a repeat do you have another comment.
Bob Wyman: Yeah I just wanted to clarify in response to what Ted was saying there's no question that search is often useful however in the social world it's it's something that large.
Bob Wyman: Subset of the users.
Bob Wyman: Um are actively in opposition to having their content be subject to search and it's because and it's 1 of the reasons they use.
Bob Wyman: Activity part at the moment because it cannot be searched it's been found by a number of shall we say minority groups.
Bob Wyman: Either racial or gender sexual whatever um that.
Bob Wyman: If when search is available it becomes possible for bullies and uh uh trollers to discover um or to impute the Identity or the characteristics of individuals and then and then abuse them um and.
Bob Wyman: And for those groups they are very vocal about the fact that they do not want.
Bob Wyman: You know as Ted was saying search is generally useful and it's provided a lot of benefit to us all it provides a lot of benefit to those who are opposed to having their content be searched as well in other kinds of content but we need to respect um the fact that.
<dmitri_z._(pres.)> yeah, Search is hard on a /good/ day, let alone when there's access control and privacy involved
<bumblefudge> it's worth mentioning that many core designers of the protocol were explicitly informed by histories of online abuse and optimized for coziness and safety which the commercial social web was NOT providing at the time
Bob Wyman: Uh 1 1 Wrights 1 has rights over what right and and uh uh we need to find ways to uh provide search while respecting uh the authors uh the right holders uh desire not to be searched or or or data mind okay that's all.
<bumblefudge> i can't remember offhand if it was listed as a goal in the use-cases docs, but it was there
<tantek_çelik_(he/him,_mozilla)> Dmitri can you stop presenting the slide so we can see the full grid of folks on the call?
<dmitri_z._(pres.)> of course
<tantek_çelik_(he/him,_mozilla)> thank you!
Harrison_Tang: Yeah so I have 2 questions um the first 1 is uh how do you actually do search without a centralized index so that's a search related question the second 1 is about uh the feed the news feed so uh I think most of the uh the biggest uh draw to uh social networks is the most engaging part the features the newsfeed or or maybe in Tik toks is the uh ever going uh videos and for building an engaging Newsfeed you need the user data right like essentially these algorithms basically uh do supervised learnings toward uh engagements like clicks and Views so how do you actually create a Federated uh news feed um how do you actually even uh create a Federated uh data model for user engagements just curious if anyone have thoughts on that.
Dmitri_Z._(pres.): Uh David go ahead you're up.
Dmitri_Z._(pres.): Uh cleia do you mind muting.
David Mason: Can you hear me.
David Mason: Oh wonderful hey everyone it's been a while so um yeah so I actually um been researching um.
David Mason: Set of Earth and activity Pub and activity streams to hopefully make a recommendation for a project.
David Mason: And uh I initially asked you know is there a doc a document or something that would be this is how you would create an application.
David Mason: And it didn't really seem to exist and then I started digging in and realizing that what there are is a lot of um discussions around all the different directions that's going in and all the problems and um.
David Mason: You know.
David Mason: I'd still like to.
David Mason: Find a way.
David Mason: Where I could today make you know legitimate uh uh application in this realm or at least um something that would be forward compatible uh but you know I've seen a lot of issues um that really make me hesitant around um making any kind of recommendation um other people have discussed different aspects of this but I mean the privacy concerns um around you know pill pillaging uh information I mean it's unsafe for some people today and I think for everyone in the long run if anyone's you know building these uh you know these models of people uh I think they can still do it with uh uh thought of verse um you know there's Bad actors uh you know including companies changing hands so good 1 day they're about the next there's box there's breaches there's.
David Mason: Laws that change.
<evan_prodromou> Not true
David Mason: There's a servers that centralize a lot of information I mean as far as I understand Mastadon um.
David Mason: I I appreciate that probably some of my points are not correct because it's quite tense but uh as far as I understand um on Master on um.
David Mason: Some of the like direct messages are not actually private so somebody running uh a master on node can read other people's uh what are some kind of times called private messages um they can see who's subscribing to what um so um but at the same time I think actually for a system that you know is existing today uh it really should kind of be about the graph um and it shouldn't you know trying to be returning to be uh 200 uh Twitter it should expose as much of the information the valuable and useful information uh of the graph that people are on uh to the end user and uh to to applications and so on so it's a really hard Balancing Act.
David Mason: And I do.
David Mason: Apologize for all the things.
David Mason: It's very difficult to understand all this um uh but that's basically what I've been thinking about uh when trying to make a recommendation thank you.
Laurin_(he/him,_ActivityPods): To someone want to respond to 1 of the open issues first maybe before I open a new 1.
Dmitri_Z._(pres.): It's fine I think people are hopping on the Queue to respond so go ahead.
Laurin_(he/him,_ActivityPods): All right um so I'm working on a project called activity pop which is the activity pots sorry which is a general purpose activity pop server that supports um.
Laurin_(he/him,_ActivityPods): Blood pods for your personal account.
Laurin_(he/him,_ActivityPods): And right now I'm.
<dmitri_z._(pres.)> yeyyy solid + activitypub!
Laurin_(he/him,_ActivityPods): We're starting to implement the cap LD for um.
Laurin_(he/him,_ActivityPods): An authorization scheme to support certain kinds of inbox actions for users for example.
<bobwyman> We need a Rights Expression Language that can express grants of rights, and withholding of rights, for any object in the SocialWeb. If you can sign something, you should also be able to make a statement about the rights that you grant or withhold.
Laurin_(he/him,_ActivityPods): I assume an authorization capabilities to someone to send me a specific kind of activity like um and other except for following a person.
Laurin_(he/him,_ActivityPods): Um and I was wondering if anyone else has been working on this um or has a billions on it and um.
<dmitri_z._(pres.)> @Laurin - I'm also VERY interested in exploring capabilities (both for Solid and the Fediverse)
Laurin_(he/him,_ActivityPods): I'm not very familiar with the process standardizing this so for now it's process of um yeah checking out how to go on about it and I created an issue on our post on the social help Forum um I'll post in the link and I'm happy for anyone to respond there.
<dmitri_z._(pres.)> also yeyy Spritely!
Dmitri_Z._(pres.): And I think you.
Emelia_S.: I'm pretty sure learning is implementing the paper from sprightly if I'm not mistaken.
Dmitri_Z._(pres.): Evan I think you were also on the queue in addition or what was that.
<bumblefudge> friend of the podcast/CG Christine Lemmer-Weber was an editor of the AP spec and runs SPritely AFAIK
<emelia_s.> It's not private in the sense of end to end encrypted
<bumblefudge> there are more private-only implementations in progress - bovine, for example
<david_mason> past legal issues, I think a bot one way or anohter could access this information
<emelia_s.> i.e. server admins _can_ read your DMs, we just trust them not to.
<tantek_çelik_(he/him,_mozilla)> ^ this
<bumblefudge> yeah, David is right, privacy and its liabilities are measured by results and fake accounts or malicious/sloppy mods are important to factor in
<bumblefudge> it's a complicated topic!
Dmitri_Z._(pres.): Thanks Evan uh I wanted to hop on the Queue to talk about search but I do see we've got uh 5 now 4 minutes uh to the top of the hour uh so uh I want to give uh Brent and uh Evan more more time to speak uh yes search is hard come join the conversation uh we can talk about Solutions.
Brent Shambaugh: Yeah Demetri I I was actually interested in search of how you can have people who don't want to be found and maybe activity Club got in there and people that do want to be found in the same space.
<emelia_s.> I have a small little thing on credentials
Dmitri_Z._(pres.): Yeah it's a it's a really fascinating and valuable and challenging um go ahead Amelia.
Emelia_S.: Uh so there was mentioned before of uh credentials in in particular for verifying the identity of journalists and things like that.
Emelia_S.: Um I am currently well I say currently it's probably within this year hopefully uh working on uh identity verification for the fediverse.
Emelia_S.: Um and I have a couple of ways started doing it that integrate already with existing Fede versus software but verifiable credentials for instance would be very beneficial here.
Emelia_S.: It's just a fact that we don't currently have them widely available on not strictly widely available to the General Public.
<dmitri_z._(pres.)> @Emelia - I'd love to chat with you about it!!
Emelia_S.: So 1 of the things that I have to look at there is actually charging some small amount of money in order to verify someone's identity through 1 of these third-party services that do the AI ml.
<peter_z> Has anyone mentioned rel=me verification??
<bumblefudge> ^ not yet
<bobwyman> Let's not do something that isn't compatible with Verifiable Crediential!!!!!
Dmitri_Z._(pres.): Thank you uh Peter and Chad is asking is anybody mentioned real me verification uh no we haven't yet that's another uh really useful tool in our toolbox uh come over to the group we'll tell you all about it.
<emelia_s.> If you're interested about that, you can reach me at
Dmitri_Z._(pres.): Uh yeah I think uh hand handing it back over to uh to Harrison Bob go ahead.
Bob Wyman: I was.
<tantek_çelik_(he/him,_mozilla)> thanks bumblefudge
Bob Wyman: But the thing.
Bob Wyman: Needing to pay somebody um.
Bob Wyman: I I I I think at the credentials were group will back me up on this there are many ways to verifiable credentials without having to pay anybody.
<kaliya_identity_woman> He was talking about traditional identity proofing
Emelia_S.: Uh sir specifically here um the issue is that a lot of the ways of getting a verified credential of some sort uh don't exist globally so for instance you've got certain US states that do Implement verifiable credentials.
Emelia_S.: They're not always easily accessible to the General Public.
Bob Wyman: Right right but that's a verifiable state credential.
<tantek_çelik_(he/him,_mozilla)> OTOH buying a domain name is available worldwide
Emelia_S.: There's also there's also then some requirements from certain laws egh verification laws where you actually are required to use government ID in a certain way.
Bob Wyman: Right right we should.
Emelia_S.: And it doesn't necessarily mesh with the stuff that they're doing with verifiable credentials in driver's licenses or things like that um so the current standard within that or industry is actually to use uh services like jio um ID now Etc.
Emelia_S.: All of those.
<tantek_çelik_(he/him,_mozilla)> Thank you Dmitri Z! Appreciated!
Bob Wyman: Right we shouldn't go in the in great depth on this the important point is that yes some verifiable credentials are the require money or great amount of effort to get them but not all do so we can still use verifiable credentials even if people don't have the aren't using the kinds that you have to pay money for.
Emelia_S.: Right so there are many ways that you can implement this and uh.
Emelia_S.: at the end of the.
Emelia_S.: It's easier to implement 1 credential solution first and then add others.
Emelia_S.: Than uh trying to support all the many many many many different versions of credential management out of the box straight away.
<bumblefudge> anything first, open platform second?
Emelia_S.: That's my kind of product the answer.
Dmitri_Z._(pres.): Thank you everyone uh over to Harrison.
<bumblefudge> any port in a storm if the lawyers and regulators are already knocking on the door!
Harrison_Tang: Well thank you Dimitri and thanks everyone uh so.
Harrison_Tang: I think these are great conversations and uh if you have any questions or want to continue those conversations please feel free to join the uh Social Web uh working group and uh you know I think we probably should do another 1 of these sessions I think we have so so many great discussion today well another 1 of these sessions like in a few months so thanks a lot uh this concludes this week's ccg meeting and thanks to Ministry uh for the in the discussion.
<david_mason> thanks everyone!
<evan_prodromou> Thanks all!
<emelia_s.> @kaliya fyi, my pronouns are she/her