- From: CCG Minutes Bot <minutes@w3c-ccg.org>
- Date: Wed, 29 Jan 2025 06:49:04 +0000
- To: public-credentials@w3.org
Thanks to Our Robot Overlords for scribing this week! The transcript for the call is now available here: https://w3c-ccg.github.io/meetings/2025-01-28/ Full text of the discussion follows for W3C archival purposes. Audio of the meeting is available at the following location: https://w3c-ccg.github.io/meetings/2025-01-28/audio.ogg A video recording is also available at: https://meet.w3c-ccg.org/archives/w3c-ccg-weekly-2025-01-28.mp4 ---------------------------------------------------------------- W3C CCG Weekly Teleconference Transcript for 2025-01-28 Agenda: https://www.w3.org/Search/Mail/Public/advanced_search?hdr-1-name=subject&hdr-1-query=%5BAGENDA&period_month=Jan&period_year=2025&index-grp=Public__FULL&index-type=t&type-index=public-credentials&resultsperpage=20&sortby=date Topics: 1. <Updates on Solid and Decentralized Data Stores> 2. <Updates on Solid and Decentralized Data Stores> Organizer: Harrison Tang, Kimberly Linson, Will Abramson Scribe: Our Robot Overlords Present: Alan Karp, Harrison Tang, Greg Bernstein, Aaron Coburn, TallTed // Ted Thibodeau (he/him) (OpenLinkSw.com), Mahmoud Alkhraishi, Vanessa, Hadrian, Alberto Leon, Nis Jespersen , Jennie M, Erica Connell, Joe Andrieu, James Chartrand, Tom S, Will Abramson, Leo, Kerri Lemoie, julien fraichot, PL, Savita Farooqui, Dmitri Zagidulin, Jeff O - HumanOS, Kayode Ezike, Kaliya Young, Benjamin Young, Sarven Capadisli Our Robot Overlords are scribing. Harrison_Tang: Hello everyone so welcome to this week's w3c ccg meeting uh today we're very excited to have Hedren uh from in Rob and uh solid and Apache software Foundation as well as uh Aaron um. Harrison_Tang: Uh the culture of the linked web storage uh group w3c link web storage group uh here to talk about the latest developments on solid decentralized uh data stores uh but before we get into the main agenda I just want to quickly go through the. Harrison_Tang: Of ethics and professional conduct reminder um just want to make sure we have constructive and respectful conversation. Harrison_Tang: All right a quick note on the intellectual property notes um anyone can participate in these calls however all substantive contributions to any ccg work items must be members of the ccg with full IPR agreements signed if you have any questions about the w3c account or the w3c community contributor license agreement uh please feel free to reach out to any of the cultures. Harrison_Tang: Are these meetings are automatically recorded and transcribed and we will publish the meeting minutes the audio and video recording in the next uh day or 2. Harrison_Tang: we use. Harrison_Tang: Gigi chat uh to to the speakers during the call to end to take minutes and you can type in Q Plus or Q minus uh to uh remove from your phone the queue and you can type in Q question mark to see who is in the queue. Harrison_Tang: Right I just want to take a quick moment for the introductions and reintroduction so if you're new to the community or you haven't been active and want to re-engage uh please feel free to just unmute and uh introduce yourself a little bit. Harrison_Tang: Just want to take a moment for the announcements and reminders so if you is if there's any new announcements uh feel free to just type in Q Plus or on mute. Greg Bernstein: I just wanted to let folks know that um for privacy preserving uh forms of credentials um particularly like the BBS crypto so we we've got some additional features that have advanced to working group documents over at The cryptographic Forum research group The CFR of the ietf so that's pseudonyms which we all want to uh give us limited linkability so we can kind of be remembered but not too much and uh some extra security features and which go under blind BBS so those have advanced to work group status we also have. Greg Bernstein: For that at the CFR we have public GitHub repos where people can now post issues and such like that so I'll put. Greg Bernstein: Send out an email to the ccg with all those links um. Greg Bernstein: I will be adding more explanatory texts to both those documents uh the more you want to see and example of cases so we're making progress we're going to get this nailed down so we can get the BBS spec out. Harrison_Tang: Thank you thanks Greg. Harrison_Tang: And by the way uh we had Greg uh here and present about BBS plus uh in the in the last few months uh so you can just search in your email inbox if you are subscribed to the email list uh for Greg uh and then it's the emails from me to see the recordings or if you uh cannot find the recording just reach out to me because uh it is a must-watch sessions because I think Greg has done a very good job explaining what BBS plus and commitment schemes and all these things uh so thanks Greg. Harrison_Tang: All right any other announcements or reminders. Harrison_Tang: Uh any updates on the work items. Harrison_Tang: I just preview uh what's coming so next week we'll have uh Drummond here to talk about uh what used to be called Global acceptance network uh trust framework now I think it's called era IRA. Harrison_Tang: And anyway if I mispronounced it I apologize for that uh and then the week after that we have uh Mike porok uh our former culture to talk about trade verified. Harrison_Tang: Uh verifiable traceability and AI in Supply Chain management. Harrison_Tang: All right uh last calls for introductions reintroductions announcements or work item related. Topic: <Updates on Solid and Decentralized Data Stores> Harrison_Tang: Great well let's get to the main agenda so again uh very very happy here to have a Hadrian and Aaron here uh to. Harrison_Tang: Some updates on the solid and decentralized data source. Harrison_Tang: so Aaron. Harrison_Tang: You want to take it away actually if you don't mind just introduce yourself a little bit and uh as well as the your work whether it's at in rump or the w3c uh link web storage group. Aaron Coburn: Thanks Harrison um. Aaron Coburn: So I will start off by just introducing myself I'm Aaron Coburn um along with hadrien I work at inrupt um I'm 1 of the co-chairs of. Topic: <Updates on Solid and Decentralized Data Stores> Aaron Coburn: The w3c linked web storage working group this is a working group that was started um September of last year so I'll get into some of the time frame that we have for this but we've um we're still in a sense getting getting ramped up um in the working group and um before starting with the working group I've I've I've been involved in the solid world for. Aaron Coburn: 567 Years um even before I came to interruptus so I've worked on implementations I've worked on specifications so I've seen kind of a broad range of of the issues that relate to this this uh this space. Aaron Coburn: Let me just um jump into I have a couple of slides. Aaron Coburn: Um there should be fairly quick I want to jump on Hedren's toes here um so. Aaron Coburn: Uh link web storage it's uh working group we've got um a number of people who meet uh we meet every week um on Mondays and I'll get into the some of the you know how you can participate. Aaron Coburn: Really the high-level goals here have to do with web-based permission data storage. Aaron Coburn: Um secondarily um interoperability across this shared storage using different applications. Aaron Coburn: Decentralization can be a part of this but it's not necessarily a part of this so I just want to make it clear that. Aaron Coburn: For some people decentralization is really key when they talk about solid but it's not necessarily a um it's not necessarily part of solid. Aaron Coburn: Or is it necessary part of uh the lws working group that may be where we go and and I think we're going to largely be pulling uh from technologies that come out of the decentralization space but um just kind of keep in mind that some of you know based on the charter based on where things currently stand those are really some of the high level goals. Aaron Coburn: Um we're still in the in the um process of of. Aaron Coburn: Curating use cases and and during uh requirements but some of the high-level requirements that most certainly are going to be front and center for this working group involved Global identifiers for entities and that's not just uh data that you put in your storage but agents uh applications other kinds of entities that that work in that ecosystem. Aaron Coburn: Secondly um Global semantics for application interoperability so like I said in the high level goals 1 of the key things here is making sure that you can have you can separate data from Storage I'm sorry you can separate data from applications such that you have different applications working on the same data and interoperability. Aaron Coburn: That kind of is the goal there is really how can you how can you help individuals or agents generally. Aaron Coburn: Become too tied to a particular application or ensure that the that the application that that data isn't too tied to a particular application. Aaron Coburn: That's where that interoperability comes in and thirdly and this is going to be really critical for um the link web storage and I think is where this is where a lot of the overlap with this group comes in is having mechanisms for controlling access this is both um authorization and authentication but. Aaron Coburn: We don't have. Aaron Coburn: If we don't have this mechanism. Aaron Coburn: It's really hard to it's really hard to to look at this about you know in terms of how you can share data how you can work with data across different apps with in different domains with different security profiles. Aaron Coburn: Um the current status we uh we were chartered in uh um. Aaron Coburn: September of last year so we're chartered through 2026 um at the moment we're uh we've been collecting use cases where in the in the process of refining those use cases so that we can derive very specific requirements um and then you know we'll be prioritizing those requirements. Aaron Coburn: Clearly based on the use cases not everything is going to be um is going to make it into the spec um so there will be some cutting that has to happen but some of the main things that will be part of this relate to Identity. Aaron Coburn: I'm putting in with identity um. Aaron Coburn: Because there's just so much overlap there but identity broadly. Aaron Coburn: Authorization and then uh ultimately a protocol that defines this. Aaron Coburn: This is my last slide um. Aaron Coburn: So we meet weekly on Mondays at uh 10:00 a.m. eastern time. Aaron Coburn: So the timing always uh aligns with Boston time so you know. Aaron Coburn: When there's a time a time change that's that's what we're we're pinned to um participation is open to all w3c members just like any other uh working group. Aaron Coburn: There's a link here I can paste this into the chat in just a moment um has more information about joining um and about the the calendar obviously the minutes are all public so anyone can can access those. Aaron Coburn: Um the other. Aaron Coburn: Thing I wanted to mention is that everyone is welcome to contribute and discuss all of our use cases which are currently in that first uh repository that first GitHub repository lws. Aaron Coburn: A lot. Aaron Coburn: Of the eventual work will go into the lws protocol repository but so far that's pretty empty we haven't we just haven't gotten to the point where we were ready to start their. Aaron Coburn: That is all that I have so. Aaron Coburn: Happy to open it up for questions. Aaron Coburn: Um or to move into the next stage. Harrison_Tang: Sounds good well before we get to the questions and by the way I prepared a couple questions and if people other people have questions feel free to just type in Q Plus uh to add yourself to the queue but before we get to that Hadrian do you mind kind of uh do a quick quick introductions about yourself that would be great. Hadrian: Uh sure and thanks Harrison for uh inviting us uh it's a pleasure to uh talk to you I don't know much about the ccg group although I recognize quite a few names in the audience right now uh some of them for from the IBEW crowd um I work for indrop for a bit over 2 years but and you mentioned that um Aaron works there as well but funny enough uh actually Aaron and I bumped uh into each other a long time ago because we're both committees at the Apes of the foundation um Aaron on Jenna if anybody is interested so turns out that we know each other for quite a while. Hadrian: Um and I came with a different background I was um. Hadrian: Uh working for a while with dog SS at uh customer comments and that's really how I got to be a user of sorts and observer of uh solid. Hadrian: We had many discussions about that and I'm sure some of you are aware of those. <aaron_coburn> Some links from my presentation: Linked Web Storage WG: https://www.w3.org/groups/wg/lws/ Hadrian: And there was some frustration that solid is not progressing well enough and kind of by accident I ended up working for for interrupt. Hadrian: And I'm very happy that we're making quite a bit of progress and I hope this uh solid idea will will catch up. Hadrian: Um can I share some slides uh. Harrison_Tang: Yeah of course thank you. Hadrian: Okay so let's see if I can share my screen. Hadrian: Can you see my screen I think you can. Hadrian: So I adapted some old slice there's not much and I'm going to Echo something that Aaron said but a bit from a different perspective because solid is actually not very easy to explain it's pretty fundamental and it's very hard to explain until you start using it and then it becomes kind of obvious um. Hadrian: Yes so you probably know that uh team is the author of solid he started it around 20145 and he incubated at MIT and then it became a community group at w3c uh in the last part of the decade and then um he started being dropped to create uh he's 1 of the co-founders to uh um create commercial products and foster adoption and last year is Aaron said the lws uh group was created uh because uh community group cannot issue a final recommendation w3c as you know so lws is probably the Final Phase in getting this specification uh to become a real standard w3c final recommendation and the motivations for for doing this for team initially were the fact that. Hadrian: The web is good and adopted as it was it wasn't really writable right so he talks a lot about web 3 as you probably know um but the web 1.0 did not it was only read only and it was actually great it was very simple it got adopted but it wasn't writable. Hadrian: you could. Hadrian: Right with your own tools in the back end the famous Pepe web server public HTML you probably know all that then tools uh came about and all that but. Hadrian: From the web it wasn't really writable. Hadrian: Because it didn't handle identities for instance that's what that was 1 of the main blockers so anyway we have to uh all came around. Hadrian: And then uh. Hadrian: What team calls web 3.0 uh always emphasized uh not to be confused for web 3 the uh blockchain world. Hadrian: You see on the slides that some of the frustrations came from the stagnating and all that however the more important um. Hadrian: Path for me was my work with doctors and they work on the intention economy which is if you want kind of a practical application uh of the technology in the sense that in order for people to engage in economic transactions or any kind of transactions socials transactions and all um they have to exchange data and data has to reside somewhere and right now because of communication issues the easiest way to exchange data is being on the same platform because exchanging data outside platforms requires protocols which don't really quite. Hadrian: Exist there are some but uh nothing um. Hadrian: Globally accepted and um that led to the idea. Hadrian: And and by the way um. Hadrian: Last time as far as I know team and Doc met was a Web Camp 2 years ago and actually team said that solid was impart inspired by Doc's ideas and his uh intention economy or portray Manifesto uh 1 of the books. Hadrian: So as I said slowly this kind of hard to explain. Hadrian: So what it does it tries to separate data uh from applications and from identities as Aaron already mentioned. Hadrian: It's a bit subtle I don't know if I have a slide for that. Hadrian: Yes it's this 1 so today the way applications are built is to have a front end and have the back end and the front end has some if you want more or less unlimited access to the back end and they are very tightly coupled. Hadrian: Solid proposes a turned around sort of model that decouples uh applications from. Hadrian: Um data allowing data to be stored in if you want this Universal data storage. Hadrian: Solid calls them ports. Hadrian: And then from their data being shared with. Hadrian: Applications and other entities so that these parts become. Hadrian: If you want a distributed decentralized back end for any kind of application. Hadrian: And if anything that's what I would like you to retain from today's presentation because it has massive implications on a lot of things for me most importantly is cost because. Hadrian: The way things are done today it's very costly for applications to be built and more importantly maintained at scale um. Hadrian: Aaron can tell you that uh indras largest customer is the government of Belgium of Flanders that provides a pod to every citizen and I think what they realized is that if you have data. Hadrian: Citizens own support and the part and they have their data in the Pod then there is no need for data exchanges between different government agencies and all that because all the data is in that reflects the relationship between the government and the citizen is in that part and the only need thing they need to build now is new uh applications to add more data and benefit from the existing data so. Hadrian: That's a very interesting model that I hope you're going to think more about and as I said if anything I would like this to be the conclusion of this uh conversation um. Hadrian: I think I'm gonna stop here there are some basic things here. Hadrian: Um Harrison said that I we should expect this conversation to become quite Technical and uh well that's in part why aren't here because it's way more technical than me um so. Hadrian: I think we should open up the rest of the time for for questions uh Harrison what do you think. Harrison_Tang: Yeah that would be great um let me uh ask a clarification question uh is what is the difference between decentralized data stores versus like link web storage or is it uh basically synonyms that means the same thing. Hadrian: You mean linked web storage is the name of the group. Harrison_Tang: Uh link web storage and decentralized data store yeah like is there a difference or we're using these terms like interchangeably yeah. Hadrian: So I should let Aaron take it but linked web storage is the name of the work group at uh. Hadrian: W3c I don't think we use it. <kaliya_identity_woman> Decentralized Web Node Hadrian: Different contexts whereas decentralized data store is a general concept uh. Hadrian: Please correct me if you have a different View. Aaron Coburn: Sure some of this I think comes from the history of solid um. Aaron Coburn: Historically solid was built on um. Aaron Coburn: A protocol called um linked um the linked data platform ldp. Aaron Coburn: Ldp itself is not particularly. Aaron Coburn: Well it's not particularly decentralized at all. Aaron Coburn: That's the first thing. Aaron Coburn: I don't think that decentralization really had any had much to much of a role in the development of ldp. Aaron Coburn: A solid background really is more in the linked data sphere. Aaron Coburn: Like I said. Aaron Coburn: There are ways of approaching solid from a decentralized perspective and you can you can you can certainly build solid as a decentralized system but it's not necessarily the case that all solid systems or all systems that come out of the linked web storage Community you know working group will necessarily be decentralized. Aaron Coburn: What we tend to find even in in interrupt. Aaron Coburn: This is especially the case for um the instance that hadrien mentioned with uh the government of Flanders or Belgium is there's more of a Federated model that's actually that tends to be more of an of of the interest um for you know for like large Enterprises for governments. Aaron Coburn: They need to they need to have a governance model that makes sense to them so. Aaron Coburn: Not so much decentralized necessarily but it can be decentralized in the sense that you can have you can have data all over the web you can have you can have lots of different pods you can have your. Aaron Coburn: User identity decoupled you can have like everything can be decoupled but it's um. Aaron Coburn: But it's not necessarily the case that it's decoupled. Harrison_Tang: Got it Dmitry you're in the queue. Dmitri Zagidulin: Hi Adrian I just wanted to add a couple more words. Dmitri Zagidulin: Uh specifically to to Harrison's question on so we have this vague notion of decentralized data storage where does solid fit in in that landscape. Dmitri Zagidulin: And the the thing that it's really important to think about with data storage is less centralized and decentralized and more storing public data versus storing permissions data and there's a pretty fundamental divide in. Dmitri Zagidulin: In let's say. Dmitri Zagidulin: Blockchain world or any sort of decentralized world. Dmitri Zagidulin: Fundamental bi. Dmitri Zagidulin: Side all data storage. Dmitri Zagidulin: Eventually is divided into either we're dealing with public only data right so that's ipfs for example many of you are familiar with ipfs internet interplanetary file system uh similar projects built on it like ceramic and some others and in general we've seen a number of protocols that are like okay you know what permissioned data storage is hard we're going to focus on public data right so for example the that project dat which I think. Dmitri Zagidulin: Uh well. Dmitri Zagidulin: Protocol I I forget what it got renamed to but it got started with we have public data sets from big science right so we've got like terabytes and petabytes of data how do we store them efficiently in a decentralized way but it's all public data. Dmitri Zagidulin: So so that's that's 1 major category of of writing and retrieving public data. Dmitri Zagidulin: And then the other 1 is. Dmitri Zagidulin: And and and the divide with permission to data storage is essentially it goes like this. Dmitri Zagidulin: There's a number of people who believe that you can ensure permission storage with encryption alone. Dmitri Zagidulin: Those are um the number of projects that are like we can do permission to storage on untrusted nodes meaning anybody could spin up a storage node and we're gonna we're gonna ensure Access Control via encryption alone. Dmitri Zagidulin: And the other half of that divide says no encryption on can never be sufficient because all encryption has a half-life you're essentially just temporarily storing data until it becomes public. Dmitri Zagidulin: So we're solid Falls in is and where how it fits into the decentralized landscape is there's going to be some some cases that deal with public data that can be fundamentally decentralized. Dmitri Zagidulin: Some cases that are low enough threat that you can say oh yeah yeah we're going to encrypt them but really it's it's temporarily uh encrypted until it becomes public. Dmitri Zagidulin: There's a number of cases where you want the data to be private and that means. Dmitri Zagidulin: Encryption with permission to access control and that's where solid solidly fits in it's it's exactly for those applications where. Dmitri Zagidulin: Can still be. Dmitri Zagidulin: Centralized you can still be decentralized you can still be a network of nodes where you storing and reading stuff from. Dmitri Zagidulin: But those. Dmitri Zagidulin: Notes have to be trusted those notes have to be able to enforce the access control the authorization part of the protocol so hopefully that uh puts it a bit more into perspective. Harrison_Tang: Ted you have a comment. Hadrian: No I just wanted to say that for those who are not aware Demitri is 1 of the original parents of uh solid at MIT. Harrison_Tang: I did know that. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There are a lot. Harrison_Tang: That's cool I feel like I know demetry for a long time but I didn't know that fact so amazing. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It's not just a couple. Hadrian: So yeah so and uh in SWS now we start we are probably starting to discuss deeds and deity actually showed uh solid working with uh did web I think quite a while ago so FYI Demetri. Hadrian: So uh go ahead that sorry. <dmitri_zagidulin> haha thanks Hadrian :) Harrison_Tang: So just to sorry uh just to summarize a demetry your comment so you're saying that a solid positioning and uh unique value proposition is on permission uh storing permission private data in the decentralized way versus like ipfs is focused on public data is that correct. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): That's not quite accurate. Harrison_Tang: Oh sorry do you mind the yeah. Dmitri Zagidulin: So uh I I would say that's roughly correct but uncertain uh time is going to give more details go ahead. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Uh yeah so having been involved in this space for quite a while I was part of the ldp working group. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Solid based itself or solid servers were initially based on. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Solid itself is more of a philosophy. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And and implementation or specification. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): That philosophy is. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): 2 things first is. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): We decentralizing the web. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And what that means is instead of there being like 3 places where you can go to have your blog hosted. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There are 3 million places that you can go to have your blog posted. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And it still can be visible. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): As visible as if it were just in those 3. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The way that gets accomplished is by Federation among other things that aggregation. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): People who have dealt with um. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): News readers in the modern sense. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Where they're picking up RSS and atom feeds and aggregating all those feeds into 1 long feed that you read. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Sort of similar to how the old use Net News groups used to work. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The other piece is if I can get my brain to cooperate. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Uh read the centralization and and making the web writable. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It's always been globally readable right there's no purpose to having a web server except to let people read it. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Also there needs to be a way for people to write it. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And that writing. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Takes many forms 1 of the most commonly understood ways is wikis like Wikipedia where somebody creates a page on a topic and other people come in and say now wait a minute that's not right or that's not accurate or here's a missing detail and then somebody else comes in and says well this language screwed up so let me fix some of that and somebody else comes in and says well this is only part of a bigger picture so let me make some links back and forth. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The whole web. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Was designed to work that way. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It hasn't been implemented that way because it gets hard quick. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): In themselves are hard to manage. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Uh graffiti you get um. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): A lot of sources that needs to be kept out or removed as you go. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): A Wikipedia used to be a lot more of a free-for-all for anybody who's been on the web for a while. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There could be editing Wars going back and forth on a page for months before somebody came in and said okay this is clearly a point of contention so there's going to be different articles that are branched from this to handle that contention. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Maybe somebody comes in and says this is clearly false data and this is clearly true data and so we're also going to Branch for that reason and and to claim 1 as false and the other a claim as true. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): But this is a big process and its evolving. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Quality implementations that exists solid server implementations excuse me um there is no such thing as solid per se it's not a monolith it's a lot of things as I said it starts out as philosophy. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Solid server implementations that exist. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Are themselves sort of centralizing. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Because it is hard to aggregate data from a bajillion servers. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It's difficult to present you with a view of a conversation if. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Software that is presenting that view has to pull all of the comments from all of the people's individual servers. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Instead of pulling them all from 1. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Right Facebook has they have a bunch of machines but they're all working together to look like 1 server. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And it's all 1 big database. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The vision of solid is to have. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): 1 person's data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): A space that in multiple spaces that they control that they say that application can read and or write this data that user can read and or write this data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): All controlled by the individual who owns that data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): This also gets complicated really quickly. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): As that new slide shows there's different kinds of data involved. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There's the data that is between you and your bank or Banks multiple Banks. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There's the data that's involved in magazines or books or movies or music. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): That you are mostly consuming mostly reading. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): But occasionally you're going to have something to say about it. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There's also government data and. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The government is or governments are the most consistent about wanting to keep your private data private and keeping the Public's data public. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And also wanting to keep their hands on all of it there's a thing called The Library of Congress which holds 1 instance of. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Most published works in the United States. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It grows infinitely. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): These Works have not necessarily been digitized because the Library of Congress began a couple hundred years ago. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): When there was no internet there was no telephone there was no recording system it was all just paper. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): That's going to evolve over time so this is a huge big pile of question. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There was a question a little bit ago about the relationship between solid and linked web servers. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): As I said before the solid philosophy. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Needs to be held by. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Um a number of different implementations. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Solid servers as they exist today. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Implement part of the solid vision. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Which might be considered a linked web server. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Those solid servers. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): As they're branded right now. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Might be compatible with the linked web server specification as it gets developed. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): They also might not. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): So that's part of why uh the working group is linked web servers which is talking about a style of server. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And not the solid working group which is a philosophical thing with a lot of different tendrils. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): I've said a lot of words and hopefully they've made sense I'll let somebody else go. Harrison_Tang: Well thank you thanks Ted. Harrison_Tang: Adrian Aaron anything to add. Hadrian: Um yes I do thank you so much uh Ted. Hadrian: Insights what I want to say is that the pecan there was a party and I think the guys in the board were asked to present slides that they never saw before it was kind of for fun and it feels like that's what they just did right now because I don't think he saw these slides before. Hadrian: Um anyway um. Hadrian: Are there any more questions very soon I I would really like to. Harrison_Tang: Yes uh I have a. Hadrian: Uh for the audience to get as much as they can from this from this meeting. Harrison_Tang: Yeah I have a kind of a question a little bit technical questions uh actually uh earlier Ted you're talking about the histories and the contacts now um a question is. Harrison_Tang: Isn't there a fundamental trade-off between latency performance uh versus uh decentralization and centralization right so to be specific um if you want to focus on decentralization and use Federation uh application use Federation to to to process this data then you'll take a long time right because you need to do aggregations and then do maybe do joins in real time it takes a while but then if you actually aggregate all of them together you'll uh to speed up the performance of such calculations then you're centralizing the data into 1 entity on aggregation layer right so how do you actually uh address this trade fundamental trade-off. Hadrian: Uh can I take this 1 uh here. Harrison_Tang: Yes uh huh. Hadrian: I'm sure that his other stuff to say so. Hadrian: to me that. Hadrian: A massive red herring because I worked in projects at massive scale in the past and it's impossible. Hadrian: To design this to Define what performance is to whom when and all that even today what we think happens on the web is completely different from what really happens you have cdns you have all sorts of stuff uh if you look at the searches we have algorithms like map reduce there are algorithms created all the time so I don't think. Hadrian: The way we think perform about performance should be a factor in the way Protocols are built and I stopped here uh Ted I'm curious what uh what your opinion. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Thanks um yes there are trade-offs at all levels of this picture. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Some cases it may make sense to optimize for presentation. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And in that case it may make sense for. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): all the. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Data of some conversation or conversations. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Held in a central location. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Sometimes that data will be mirrors of what people hold in their individual locations. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Part of what these individuals control over their data at the same time as applications presented. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): we can sort. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Sort of see this in action today in the way that um I think both IOS and Android do this. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): You to Echo your pictures from your smartphone to some storage place on the web which you can also access through a web browser on a on a desktop or laptop machine. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Those pictures do not exist only on your phone and then get presented through that Central thing. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Nor do they only exist in that Central thing and get presented by your phone. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): In both places because that is the way to optimize the use of that data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The services that store the pictures. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Are keeping them as private as you set them. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Know sometimes or well yeah some of them let you do settings so that you can say like Joe can see these pictures but Mary cannot. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Others and this will increase going forward let you use a semantic description to say that. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): My first cousins can see these pictures but my grandparents cannot for whatever reason. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): That kind of smart logic in. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Setting permissions it's called attribute-based assets control. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): it is. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): More attractive thing to use because the systems are getting better at being able to process it as sort of what you were saying about algorithms that get better. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): You'll get better over time but they're not there yet and so some of some recent centralization of the red decentralization. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): is going. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): To happen in order to make the presentation fast enough for the consumer. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It's not to say there's some performance benchmark. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It's to say that I don't want to wait more than 10 seconds for the first page of pictures to show up in my picture browser. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Whether that's a web browser as it is today or it's in iPhoto which is the web the picture browser on Ajo on Mac OS rather uh or it's the pictures on. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): IOS there's there's a whole bunch of applications that present picture data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And this is another part of the solid division is to be able to say I want to use ios's photos. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): I want to use Mac OS. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): On the same data files. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Down the road I may want to use a different application I don't know what the name of it is on Android but the data files are likely to be the same whether they're jpegs or gifts or pings or something else. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Those are standardized data formats which multiple applications know how to read and write and interoperate over. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Pictures are a nice easy example of this. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It gets more complex when you start talking about financial data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Is there a different schemas different ontologies different vocabularies that are used to describe that financial data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Having multiple applications able to read and write that data or read or write that data requires. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Some decisions about what vocabulary or ontology to use or at least what are the connections between the multiple vocabularies that exist. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): What I'll same as relations are there. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): I throw that in because it's an important thing that people as we get further into the weeds need to understand. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): All same as. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Is a relationship that says that the subject and object of the statement including Al same as the subject and object. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): What does that mean. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The identifier in the subject position and the identifier in the object position. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Same entity it's like my name is tal Ted my name is also Ted Tibido. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Those 2 uh identifiers. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Is Al same as relationship except of course that those are just strings they're not Uris and that's deeper into the weeds that we need to get at this moment. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The challenge at hand. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Build enough standardized vocabularies. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Enough application implementers. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Are you able to use. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): That enough applications will be produced that enough users will choose to use or want to use over the data that they have or will have. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And whether they have it on their local pod or is already exists in some. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Silo on the web or it exists on a local floppy disc right all these things are valid. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The challenge is to have the vocabulary is be understood by enough applications that they can open enough files that are becomes a a u. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Critical mass of users. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Users applications and data that they can interoperate that everything is Loosely coupled. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): What is Lucy coupled means. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Loosely coupled means that the banking application. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Can read any banking data not just the data produced by the bank that produced that application. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It means that I can decide that I'm changing Banks and all of my old data can move with me. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): At least my new bank can read all of my old data. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And understand it. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Translations are an element of loose coupling. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): The best loose coupling is. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): You can change things on the fly as you care and that's like using a different web browser. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): There were 6 of them give or take. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Major web browsers that you can use to read any website on the web. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Some are a little more bleeding edge than others and so browsers move faster or as sites move faster but pretty much Loosely they work together. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And you can change from chrome to Safari to Firefox on whim. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It's not going to change anything you take the link of a page you're looking at and copy it from 1 Brouse to another and it's still going to work it's going to look roughly the same. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): It might go a little faster a little slower it might show some data a little more pretty than the other that to your eyes but somebody else likes the other and that's all fine because these are Loosely coupled. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): They only connect when they need to. Harrison_Tang: By the way Ted. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): And again I've had a lot of words. Harrison_Tang: Yeah I'm sorry yeah I just want to do a time check do you have 1 last comment uh. TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): No I'm good. Harrison_Tang: Okay good good thank you thanks Ted so Adrian I know you have other slides that you want to share so do you want to proceed yeah. Hadrian: No actually that's the last 1. Harrison_Tang: Got it got it okay cool cool um so earlier tag was. Hadrian: But we can take more questions uh Harrison. Harrison_Tang: Got it okay so I I think uh earlier uh you know Ted actually mentioned a good point where uh to to make sure that this string comes true you need to have this uh ecosystem and uh where a lot a lot of pods are going uh basically a lot of adoptions right where it creates a positive reinforcement Loop so my question is like uh what are the market adoptions right now and what are the challenges and what can we do to actually drive more adoption. Hadrian: Oh man that's a that's a actually Ted said it very well I like what he said although it's a bit theoretical so. Hadrian: This is a very low conversation in reality I don't know but. Hadrian: I was thinking what what was the motivation of people paying for isps let's say back in the in the 90s because they didn't provide anything except access right you had to have that connectivity to the internet to then access websites and data and all that. Hadrian: And that's in a way the case for solid you need to have enough um. Hadrian: Either solid hosting providers or easy to set up like the Apache browser uh sorry the Apache web server was um technology to install on your Lan to host the solid pods and then you have to have that critical mass of applications that use this data from multiple Parts because you will share data with your family or communities and they will use data from your parts but this applications will provide value for the people who use them. Hadrian: And that involves writing as well like both Dimitri and and that said and once you have that critical mass and the data resides in the solid parts as I said this would create a sort of universal distributed decentralized uh back end so people will basically only write applications or algorithms to benefit from this data of course. Hadrian: Agents are a hot topic. Hadrian: Um AI is a very hot topic uh especially given the um. Hadrian: Privacy requirements um wallets are also hot and I saw Alex is on the call she's at MIT and they um so interrupt donated 1 of the wallet projects that are based on solid parts at the uh uh open Wallet foundation and Alex leads a group at MIT that did the same. Hadrian: Um so wallets I think are going to be very hot and in terms of governance and acceptance. Hadrian: I'm not sure who mentioned Gan earlier today but we became members of Gan as well and that's another thing to look at in terms of governance and um. Hadrian: Promoting more the idea of yeah the work of global acceptance networks and now they are Ira because it was incorporated in Switzerland under this name recently um. Hadrian: So I'm not sure if I gave you an answer Harrison but I think this would be the Milestones or the the points to look for. Harrison_Tang: Garden well earlier Aaron shared that in Rob has a very satisfied customer with flandes and Netherland government so what's the user value proposition uh that they chose uh to use a solid interrupt uh versus other types of solutions. Hadrian: Well I don't think there are other types of solutions with this characteristics but Aaron do you want to talk more about that. Aaron Coburn: I think 1 of the value propositions has to do with really this this question of data sharing so building. Aaron Coburn: A an ecosystem. Aaron Coburn: Where more Innovation can happen so um. Aaron Coburn: The government of Flanders is trying to do is is create a structure where the data becomes available for. Aaron Coburn: Is you know. Aaron Coburn: Oh sort of allowed entities to write applications that can interact with the data. Aaron Coburn: The data that's that's really controlled by users in this case um and and then provide additional benefit beyond what the um you know any individual. Aaron Coburn: Governmental agency would be able to to provide so that's what they're that's the the basic value proposition that that they're going with and. Aaron Coburn: You know. Aaron Coburn: Part of it has to do with. Aaron Coburn: Rather than having a lot of siloed kinds of of. Aaron Coburn: They would have a um. Aaron Coburn: Structure of data where the data itself becomes um more of a um more of an object for sharing you know across these permiss specifically permission to sharing across these different agencies. Hadrian: Uh Aaron may I point out something it may not be obvious um the fact that you have control over who the data is shared with the government data in the pot doesn't necessarily mean that you can alter or even see that data so imagine it's a criminal record it's not like oh it's in my pot I'm going to erase my uh criminal record now. Hadrian: So I'm not sure if that's obvious but. Hadrian: Texas doesn't necessarily mean that you're the owner of the data can can modify it. Aaron Coburn: Indeed it's not it's not as though I can just uh tell my banking pod that I I have a lot more money than I currently have but there's there's there's um there are a lot of ways to achieve that in terms of as an example verifiable credentials or other kinds of um. Aaron Coburn: Uh cryptographic mechanisms to show that you know has a particular bit of data been tampered with or you know who was who signed for this particular data so there's there are a lot of ways of going about making sure that a particular entity is authoritative for a particular bit of of data. Harrison_Tang: Cool thank you uh 1 last question is from the community if there's any. Harrison_Tang: All right I think we're at time uh just want to thank you Hadrian and thank you Erin for jumping on this is a great conversation I feel like I need to invite you guys to jump on in the next few months to do a follow-up because there's still some questions that I think I have that haven't been able to ask so thanks a lot. Hadrian: Uh thank you so much for inviting us. Aaron Coburn: Thank you very much. Harrison_Tang: All right this concludes this week's uh ccg meeting so thanks for attending.
Received on Wednesday, 29 January 2025 06:49:12 UTC