XMPP: When a 25-Year-Old Protocol Becomes Strategic Again

After twenty-five years, XMPP (Extensible Messaging and Presence Protocol) is still here. Mature, proven, modular, and standardized, it may well be the most solid foundation available today to build the future of messaging.

And now, XMPP is more relevant than ever: its resurgence is driven by European digital sovereignty efforts, renewed focus on interoperability, and the growing need for long-term, vendor-independent infrastructure.

Against this backdrop, the recent funding round around XMTP (Extensible Message Transport Protocol), a newly launched blockchain-based protocol marketed as a universal messaging layer, raises questions. The name clearly evokes XMPP, yet there is no technological or community connection. And while XMPP could easily serve as a transport layer for blockchain-integrated messaging, XMTP chooses to ignore this legacy and start anew.

So the real question is:
Why rebuild from scratch when a solid, extensible foundation already exists?

A Protocol That Never Went Away

XMPP is an open protocol for real-time messaging, designed from the start for federation and decentralization. Standardized by the IETF (RFC 6120, 6121, 7622…), it has powered mission-critical systems for decades: enterprise communication, mobile apps at scale, online games, IoT control platforms.

What makes XMPP especially powerful is not just its architectural simplicity, but its modular extensibility. The protocol evolves through an ecosystem of open specifications (XEPs), covering:

• End-to-end encryption (OMEMO, OTR)
• Multi-device synchronization (Message Archive Management)
• Group chat with subscriptions (MUC and MUCSub)
• PubSub (XEP-0060) for real-time data and events
• Interoperability bridges (SIP, MQTT, Matrix)
• And more…

XMPP has never stopped evolving. Dozens of new extensions are proposed every year. It remains one of the most adaptable foundations for building secure, federated, and future-ready messaging systems.

XMTP: A New Protocol with a Familiar Name, but a Different Approach

XMTP is a blockchain-native messaging protocol developed by Ephemera. It aims to connect wallets and dApps, leveraging decentralized infrastructure (libp2p, IPFS-style storage) and cryptographic identities.

The ambition is clear: to build a censorship-resistant, peer-to-peer messaging layer for Web3, rooted in crypto-native identity and cryptography.

However, the naming is misleading. In an older interview, XMTP co-founder Matt Galligan said the name is a blend of SMTP and XMPP. It was chosen to evoke familiarity, perhaps even as a tribute. But the result is confusing: XMTP is not an extension, evolution, or even distant cousin of XMPP. There is no shared architecture, no interoperability, no community overlap.

Why This Matters Right Now

This naming issue would be minor if it weren’t happening at a critical time for protocol design. Governments, especially in Europe, are actively exploring how to regain control over digital infrastructure. Messaging is central to this effort, especially with upcoming interoperability mandates, data sovereignty requirements, and the need for long-term maintainability.

XMPP is uniquely well-positioned to meet these needs. It is mature, open, extensible, and governed through transparent standards. It has a community of engineers, operators, and developers actively maintaining and evolving it.

Instead of inventing closed messaging stacks around new ecosystems, the more pragmatic move would be to build on robust, extensible layers like XMPP:

• Need to integrate blockchain identities? XMPP can map public keys or wallet identifiers through custom namespaces or JIDs.
• Need cryptographic message-level guarantees? XMPP already supports message metadata, signatures, and encryption.
• Need better privacy ? XMPP can be run over privacy-preserving transports like Tor.

In short: XMPP can serve as a transport layer for Web3 communication without discarding two decades of protocol maturity.

I understand that the main focus of XMTP is to prevent censorship, but this really a situation that can be mitigated efficiently with XMPP. You can for example run your own server or develop a fully decentralized approach that you can leverage as needed (e.g. xmpp-overlay).

Yes, there is still work to be done. For example, integrating MLS (Messaging Layer Security) into XMPP would provide a strong foundation for interoperable, end-to-end encrypted group messaging. But that only reinforces the point: Why ignore what’s already working and extensible?

Use What Works

New ideas are always welcome. Innovation matters. But messaging protocols are infrastructure. Reinventing them lightly, is not harmless, especially when it is done without acknowledging existing efforts.

Instead of multiplying disconnected stacks, we should double down on what works.

XMPP is here. It works. It evolves. It can be extended, adapted, and integrated, even into blockchain-native systems, without sacrificing openness or interoperability.

That may be its most valuable trait today: Still standing, while so many overengineered protocols have come and gone.