- From: Rui Zhao <rui.zhao@cs.ox.ac.uk>
- Date: Thu, 20 Nov 2025 00:55:26 +0000
- To: <public-solid@w3.org>
Dear Solid CG team, Good morning/afternoon/evening. Unfortunately I wasn't able to join the two meetings in the past two weeks, ans see the slightly complicated timezone issue (sorry I tend to think Anywhere on Earth...). I fully acknowldge the potential impact, but would still like to send this one out, also to share my two cents for some of the questions. Info about myself: I'm Rui Zhao, started to work on Solid since 2022 after joining University of Oxford's EWADA project as a postdoctoral research associate (and been in this role ever since). I mainly work on the research side, from complicated architectures/systems around Solid to small use cases of Solid, to investigate how Solid can be applied in different settings, and the boundary of the current protocol's capability. During the journey, I developed some Solid-related libraries and tools both for my own interest and to test my understandings of technology too. I've also become an active (if I may say) member of different Solid-related groups, from SolidOS, to Solid CG, and to LWS WG. (Would be willing to give some pointers to relevant papers. But in general, my research is on data autonomy, data-flow governance and privacy.) Despite being an academic researcher primarily, I also keep an eye on the practical challenges that Solid faces -- both the tecnology itself, and for creating a (larger) community by attracting outsiders. I've talked with some people, and heard their opinions and the blocking factors. Before Solid, I've had several times of active searching of decentralized technologies, especially social tools (SNS, IM, etc), and have used pretty-much all of them by then. Solid has a quite unique position for that, on two main parts (apart from RDF): 1) all data are stored in one place, so different apps/services can make use of any of them; 2) there exists (in principle) a data interoperability mechanism for cross-app (cross-client) data sharing. But that also makes Solid very disadvantageous, as it's very different from what people are used to nowadays -- platforms (combined "service"/"function" and "storage"). (And we don't want to lose Solid's unique stance.) Thus, the client-client interoperability is indeed an important factor, and I fully agree and support efforts for that. Besides creating standardard vocabularies (and/or ontologies) and (simple) wrappers around them, I would like to emphasize the importance of a "modal": to actually create a set of applications that demonstrates the power of the interoperability approach. This has been a question for many, and having a demostration of Solid's most-promised (and distinct, compared to other PDS == Personal Data Stores) feature of interoperability would be a very important step towards showcasing to and attracting outsiders; I also learned this the hard way for my research. It both proves the technology does work, and provides references on how that can be done, and hopefully generalizable. In addition to that, one thing I believe missing is the lack of convenient libraries / tools / resources for developers. At the moment, developers need to decide from scratch what they need and want to do for their Solid App -- that is not an easy job, both for Solid's design (i.e. PDS), and also for RDF. I've repeatedly heard how one should use RDF to express data, and where to put data in Solid Pods, and confused on where to start with; different people also got different preferences and suggestions. Of course, we would still like to keep that as an option, but we can move forward towards a particular direction: to create tools/libraries for building "typical" apps. That means, to create libraries for building apps of some specific feature requirements. The most prominent is single-user apps, where all data are stored to the person's own Pod; then, there can be friend-aware apps, where some data are in the user's own Pod, and some are in friends' Pods; then social apps, where (data) interactions happen between the user's and friends' Pods; alternatively also "collective" apps (e.g. social media), where index-like mechanism is needed, and there needs ways to keep track of interactions / activities. They have different underlying data handling and data *locationing* requirements, and the library/libraries should abstract this away, to provide the developer a peace of mind in that the data can be handled corectly. The "peace of mind" is what I see most important here, rather than performance (etc). So in general, I feel the CG should st up some guidance goals for people to think about and potentially work on. From time to time, we may want to reflect on how well the goals are met, and if any synergy can be made. Of course, it can only be "guidance", due to the nature of the CG; but it would be better if somehow the CG (or together with some other groups) managed to work on it together (even just for a small fraction of time). The discussion above about client-client interop and convenient libraries are both two possible common goals that I see are very important. I understand there are many affairs that the CG can work on, more than the two practical questions above, such as protocol evolvement and experimental features. But afterall, all of them has one major common goal: to make Solid appropriate for more (or all) types of services/apps. Of course, some documentation needs to be done for that, and we may also develop some tools for hosting/viewing/editing such documentation along the way; to start with, maybe reusing existing tools is necessary. The goals shall come from the group's discussion, and some relations (e.g. prority, dependency) will be put between them; I think the chairs can put some small efforts in maintining the documentation from the discussions. If I were elected, apart from regular CG activities, I would try to push the CG to spend some effort in this direction, and take the responsibility to maintain the document to start with. So, that's my two cents. Hope that's useful, and looking forward to any comments. Best regards, Rui -- Rui Zhao Postdoc @ EWADA (Ethical Web And Data Architecutres in the Age of AI) Human-Centred Computing, Dept. Computer Science University of Oxford https://me.ryey.icu
Received on Thursday, 20 November 2025 00:55:35 UTC