Discovery on the fediverse has always been its Achilles’ heel. Finding people, content, and communities across a decentralized network is fundamentally harder than on centralized platforms. FediDisco (Fediverse Discovery) is a project aiming to change that. This guide explains what FediDisco is building, how it approaches the technical and social challenges, and what it means for your fediverse experience.
What You’ll Know by the End
- What the FediDisco project is and who is behind it
- The technical approach to federated discovery
- How FediDisco balances discovery with privacy and consent
- What changes to expect in your Mastodon experience
- How instance admins and developers can participate
The Discovery Problem
On a centralized platform, search is straightforward: one database, one index, comprehensive results. On the fediverse:
- No single server has all the data
- Each Mastodon instance only knows about content it has seen through federation
- Privacy norms favor consent-based indexing
- Resources for indexing are distributed across thousands of independent operators
The result: searching for a person, topic, or conversation often fails because your instance simply does not have the data.
What FediDisco Proposes
FediDisco is building infrastructure for opt-in, privacy-respecting discovery across the fediverse. The project addresses several interconnected challenges:
People Discovery
Finding people based on their interests, expertise, or activity:
- Profile indexing (opt-in) across instances
- Topic-based discovery (“find accounts that post about photography”)
- Improved account suggestions during onboarding
Content Discovery
Finding posts and conversations by topic:
- Cross-instance hashtag aggregation
- Full-text search for opted-in public content
- Trending topic detection across the network
Community Discovery
Finding instances and communities:
- Instance directories with rich metadata
- Community maps showing topic clusters
- Recommendations based on your current follows
Technical Architecture
FediDisco’s approach is designed to be decentralized (matching the fediverse’s nature) while being practical:
Distributed Indexing
Rather than building one central search engine, FediDisco proposes a protocol for search nodes that:
- Receive opt-in content from instances
- Build local indexes
- Respond to search queries
- Federate search results with other nodes
This means search can be run by anyone, similar to how email or web search can have multiple providers.
Consent Layer
Every aspect of FediDisco respects consent:
- Users explicitly opt into having their content indexed
- Instance admins decide whether to participate
- Search results respect user visibility settings
- Users can revoke opt-in and have their data removed from indexes
Integration Points
FediDisco integrates with existing infrastructure:
- Mastodon’s built-in search UI
- Third-party clients via API extensions
- Instance admin dashboards for configuration
- ActivityPub extensions for discovery metadata
How This Changes Your Experience
If FediDisco succeeds, your fediverse experience improves in several concrete ways:
Finding people gets easier: Search for someone by name, topic, or interest and find them even if they are on a different instance than yours.
Discovering content improves: Hashtag feeds become more complete because they draw from a wider network of opted-in content.
Onboarding smoothes out: New users can find communities and accounts to follow based on their interests, not just what their instance happens to know about.
Federated timelines become richer: Better discovery infrastructure means your instance learns about more content, enriching your timeline experience.
Privacy and Consent Design
FediDisco has learned from the fediverse community’s strong feelings about privacy:
What Gets Indexed
- Only public posts from users who opt in
- Profile information from users who enable the “discoverable” flag
- Hashtag metadata from opted-in posts
- Never: private posts, DMs, unlisted posts, followers-only content
Control Mechanisms
- Per-user opt-in/opt-out controls
- Per-instance participation decisions
- Delayed indexing (content is not instantly searchable)
- Right to removal (opt out and your data is purged from indexes)
Transparency
- Clear documentation of what is indexed and how
- Public audit of search node behavior
- Open-source implementation for community review
For Instance Admins
If you run a Mastodon instance, FediDisco participation involves:
Enabling the Feature
- Update to a Mastodon version that supports FediDisco integration
- Configure which search nodes to connect to
- Set default opt-in policies (recommended: opt-in per user)
- Monitor resource usage from search integration
Resource Considerations
- FediDisco integration adds modest network traffic
- Local caching of search results uses some storage
- The benefit (better search for your users) typically outweighs the cost
For Developers
FediDisco opens new possibilities:
- Building search nodes: Run your own search infrastructure for the fediverse
- Client integration: Build search features that leverage FediDisco APIs
- Analytics tools: Help communities understand their reach and growth
- Moderation tools: Better search helps moderators find and address problematic content
See our developer notes for technical guidance on building with fediverse APIs.
Timeline and Adoption
FediDisco is being rolled out gradually:
- Phase 1: Core infrastructure and protocol specification
- Phase 2: Integration with Mastodon and major clients
- Phase 3: Broader adoption across fediverse platforms (Pixelfed, Lemmy, PeerTube)
Adoption depends on instance admin decisions and user opt-in rates. The project’s success will be measured by the quality of search results, not just deployment numbers.
Common Mistakes
- Expecting instant comprehensive search: FediDisco’s effectiveness depends on opt-in rates; it will improve gradually
- Conflating discovery with surveillance: Opt-in search is fundamentally different from invasive tracking
- Ignoring the project because “search should not exist on the fediverse”: The community’s privacy values are respected by FediDisco’s consent model
- Expecting it to work exactly like Google: Decentralized search will always be less comprehensive than centralized search; the trade-off is privacy and autonomy
- Not opting in if you want to be found: If you create content for an audience, opting into discovery helps that audience find you
Frequently Asked Questions
Is FediDisco the same as Fediscovery? They are related projects in the same space. The fediverse discovery ecosystem includes multiple initiatives. Terminology varies across community discussions.
Will FediDisco work with Pixelfed and Lemmy? The long-term goal is cross-platform discovery. Initial focus is on Mastodon, with expansion to other ActivityPub platforms planned.
Can I opt out of being discovered? Yes. Discovery is opt-in by default. If you do not want to be indexed, you do not need to do anything. Your existing privacy settings are respected.
Does this change how hashtags work? Hashtags continue to work as before. FediDisco enhances their effectiveness by aggregating hashtag data across more instances, making hashtag feeds more complete. Check our best tools guide for clients with good hashtag support.
How is this funded? FediDisco is a community-driven project with funding from the fediverse community and supportive organizations. It does not have a commercial business model.
When will I notice a difference? Gradually, as your instance and the instances you interact with adopt FediDisco. The improvement is cumulative — the more participants, the better the results. Follow our articles hub for updates.