YouTube Video
https://media.ccc.de/v/matrix-conference-2026-102447-making-matrix-feel-like-whatsapp-without-centralizing-it
Gua is an experimental Matrix-based messenger built around a deliberately provocative question: can a federated, end-to-end encrypted network feel as simple as WhatsApp, and what do you have to "centralize" to get there?
The project is an experiment built around the Matrix ecosystem. It builds on Element clients for iOS, Android, and Web, Synapse, and Matrix Authentication Service, with targeted Gua-specific forks where the user experience requires it. The goal is to keep the robustness of Matrix federation while hiding the complexity that usually scares off everyday users: homeserver choice, full Matrix IDs, server-specific login details, and federation topology.
Gua is an early-stage architecture and product-design case study about making Matrix federation usable for people who would otherwise never adopt a federated messenger.
Matrix gives us federation, interoperability, end-to-end encryption, and the possibility of communication networks that are not controlled by a single vendor. But the average user does not want to think about homeservers, server names, Matrix IDs, login methods, federation topology, or trust boundaries. In many countries, messaging products such as WhatsApp became dominant partly because onboarding is extremely simple: enter a phone number, receive a code, and start talking.
Gua explores whether a Matrix-based messenger can offer that kind of simplicity without giving up the federated architecture underneath.
The prototype uses Element-based clients, a Gua identity service, a fork of Matrix Authentication Service, and a new component called "Gua Resolver". The Resolver identifies where an account is hosted or selects a homeserver for registration. Authentication uses OpenID Connect, while messaging continues through the standard Matrix APIs and federation. Synapse itself remains unmodified.
A brief demonstration introduces the user experience, followed by an architectural walkthrough showing how account routing and authentication work without placing the resolver in the message path.
The second half addresses the trust model. In today’s deployment, one operator maintains the resolver, a signed list of eligible account-hosting servers, and the associated routing policy. The client authenticates the resolver over HTTPS but does not yet verify the signatures on its routing data.
The proposed next step is a network of independently operated Gua nodes, each with its own identity provider, resolver, and signing authority. Signed records could be shared across operators and verified by clients, allowing organizations and communities to operate their own infrastructure while preserving a simple experience for users.
This seven-minute talk uses Gua to explore the architectural and trust trade-offs involved in making Matrix feel familiar without reproducing central control.
Sarah L. Silva
https://cfp.2026.matrix.org/matrix-conference-2026/talk/FEBUFM/
#matrixconf2026 #LightningTalks
Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/
Gua is an experimental Matrix-based messenger built around a deliberately provocative question: can a federated, end-to-end encrypted network feel as simple as WhatsApp, and what do you have to "centralize" to get there?
The project is an experiment built around the Matrix ecosystem. It builds on Element clients for iOS, Android, and Web, Synapse, and Matrix Authentication Service, with targeted Gua-specific forks where the user experience requires it. The goal is to keep the robustness of Matrix federation while hiding the complexity that usually scares off everyday users: homeserver choice, full Matrix IDs, server-specific login details, and federation topology.
Gua is an early-stage architecture and product-design case study about making Matrix federation usable for people who would otherwise never adopt a federated messenger.
Matrix gives us federation, interoperability, end-to-end encryption, and the possibility of communication networks that are not controlled by a single vendor. But the average user does not want to think about homeservers, server names, Matrix IDs, login methods, federation topology, or trust boundaries. In many countries, messaging products such as WhatsApp became dominant partly because onboarding is extremely simple: enter a phone number, receive a code, and start talking.
Gua explores whether a Matrix-based messenger can offer that kind of simplicity without giving up the federated architecture underneath.
The prototype uses Element-based clients, a Gua identity service, a fork of Matrix Authentication Service, and a new component called "Gua Resolver". The Resolver identifies where an account is hosted or selects a homeserver for registration. Authentication uses OpenID Connect, while messaging continues through the standard Matrix APIs and federation. Synapse itself remains unmodified.
A brief demonstration introduces the user experience, followed by an architectural walkthrough showing how account routing and authentication work without placing the resolver in the message path.
The second half addresses the trust model. In today’s deployment, one operator maintains the resolver, a signed list of eligible account-hosting servers, and the associated routing policy. The client authenticates the resolver over HTTPS but does not yet verify the signatures on its routing data.
The proposed next step is a network of independently operated Gua nodes, each with its own identity provider, resolver, and signing authority. Signed records could be shared across operators and verified by clients, allowing organizations and communities to operate their own infrastructure while preserving a simple experience for users.
This seven-minute talk uses Gua to explore the architectural and trust trade-offs involved in making Matrix feel familiar without reproducing central control.
Sarah L. Silva
https://cfp.2026.matrix.org/matrix-conference-2026/talk/FEBUFM/
#matrixconf2026 #LightningTalks
Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/