Synonym will be initially hosting: homeserver and Pubky App
Overview
Pubky App is a decentralized social media application built on Pubky Core. It serves as a working reference implementation demonstrating how to build social applications on the Pubky protocol.
The data models and validation rules are formally specified in the pubky-app-specs repository, which defines structures for users, posts, tags, bookmarks, follows, and feeds. This specification ensures interoperability between different Pubky App implementations.
Live Application & Development Status
- Live Demo: https://pubky.app - Production instance currently operational
- Current Repository: github.com/pubky/pubky-app - ⚠️ Deprecated (MVP codebase, not for learning/contribution)
- New Development: github.com/pubky/franky - Work in Progress (next-generation client, expected summer 2025)
- Build Compatible Clients: Use pubky-app-specs as the authoritative specification
Note for Developers: The original pubky-app repository presented at BTC Prague is deprecated and marked as “very hacky” by the team. If you want to build a compatible social client or learn from production-quality code, start with the pubky-app-specs specification and wait for the franky repository to mature. The current codebase at pubky-app is transitional and should not be used as a reference.
Key aspects
- Data Ownership: Users have full autonomy over their data, hosting it on independent homeservers that are decentralized and distributed across the network. This approach enables users to maintain control and ownership of their data, while also ensuring data sovereignty and privacy.
- Profiles: The system employs a decentralized data storage approach, where post, comment, and like data are stored in association with user profiles.
- Aggregators collecting social graphs
- Feeds of followings’ activities
- Searching profiles and posts
- Notification delivery through application backends
- Distributed moderation through user blocking
Components
The Pubky App is a complex system that can be broken down into two main components: the backend and the client. These two pieces work together to provide a seamless user experience.
Backend: The Data Organizer
It collects and organizes data from various sources, processing it into a usable format.
Client: The User Interface
It is the part of the Pubky App that you interact with directly. It’s responsible for taking the organized data from the Backend and presenting it to you in a visually appealing and easy-to-understand way.
MVP Architecture
The early versions of Pubky app take some shortcuts over the Pubky Core design. The MVP app is centralized, therefore we saved time and complexity by aggregating functionality into fewer components. The main two components are the Homeserver and the Indexer
- The homeservers fulfils the function of
data stores, republishing users keys to PKARR and it acts also as an identity-provider (Oauth-like sign-in). Users maintain a trust relationship with the homeserver. - The
Indexerfulfils the function of the backend for the Pubky App. Pubky Nexus is the production implementation of this indexer, providing real-time social graph aggregation, high-performance search, and a comprehensive REST API.