3 ms·

It's worth noting that Matrix is the only federated system which sets out from the outset to bridge together all the others (hence the name Matrix). Whilst we d

It's worth noting that Matrix is the only federated system which sets out from the outset to bridge together all the others (hence the name Matrix). Whilst we don't have bridges to diaspora/urbit/gnu social yet, it's only a matter of time. PRs welcome! In terms of the community size - looking at just the Matrix.org homeserver, there are around 300K messages a day, 250K users (of which about 200K are bridged), and 30K rooms. Meanwhile there are at least 500 openly federated homeserver installations we can see from the Matrix.org homeserver. Synapse is relatively mature as a homeserver (although it still has performance challenges; it is very cool to see Ruma progressing!) - meanwhile Vector as a flagship web/ios/android client is in late public beta and very usable too - eapecially with end-to-end encryption on the horizon. I may be biased (being project lead for Matrix), but I'd say that it's going somewhere :)

raphman_ 10y ago
This sounds promising. Is anyone already using Matrix as an event server/queue for interactive environments and/or sensor networks (similar to the Stanford Event Heap [1])? What is the minimum latency for Synapse/Ruma? [1] https://graphics.stanford.edu/papers/eheap/ https://graphics.stanford.edu/papers/eheap/
- Arathorn 10y ago
  We've done some stuff with sensor networks - eg our FOSDEM 2015 demo was hooking up cars via OBD2 ports to stream their engineering telemetry into Matrix for visualisation/analytics etc: https://archive.fosdem.org/2015/schedule/event/deviot04/ https://archive.fosdem.org/2015/schedule/event/deviot04/ However, this is still fairly PoC. Our latency is deliberately high at the moment (given all events are persisted on all participating servers, and signed etc) - typically around the 100-300ms mark depending on server performance involved. If you want lower latency stuff (eg VoIP or MIDI) the architecture is that you use Matrix to be a signalling layer to negotiate the realtime protocol (eg RTP).
buovjaga 10y ago
> It's worth noting that Matrix is the only federated system which sets out from the outset to bridge together all the others Nope, Friendica did that first (started in 2010): http://friendica.com/ http://friendica.com/
- Arathorn 10y ago
  fair enough, although I guess I think of Friendica as being more social network based rather than realtime chat/messaging/VoIP like Matrix.
habales 10y ago
"eapecially with end-to-end encryption on the horizon" https://matrix.org/docs/spec/client_server/r0.1.0.html#end-to-end-encryption https://matrix.org/docs/spec/client_server/r0.1.0.html#end-t... ... unfortunately there is nothing there yet ;-/ ... this is a sign, encryption will be placed on top, which will weaken the idea, as it is an optional addon not a first class citizen.
- Arathorn 10y ago
  The Matrix spec is still in development and E2E has been in progress since the beginning. It is not an "optional addon", it will be part of the core when merged. You're looking at the wrong branch: https://matrix.org/speculator/spec/drafts%2Fe2e/client_server/unstable.html#end-to-end-encryption https://matrix.org/speculator/spec/drafts%2Fe2e/client_serve... is the in-progress spec, and https://matrix.org/jira/browse/SPEC-162 https://matrix.org/jira/browse/SPEC-162 tracks all the progress. Meanwhile, it's even landed on the develop branch of Vector already (at least for the 1:1 ratchet).

raphman_ 10y ago

This sounds promising. Is anyone already using Matrix as an event server/queue for interactive environments and/or sensor networks (similar to the Stanford Event Heap [1])? What is the minimum latency for Synapse/Ruma? [1] https://graphics.stanford.edu/papers/eheap/ https://graphics.stanford.edu/papers/eheap/

Arathorn 10y ago
We've done some stuff with sensor networks - eg our FOSDEM 2015 demo was hooking up cars via OBD2 ports to stream their engineering telemetry into Matrix for visualisation/analytics etc: https://archive.fosdem.org/2015/schedule/event/deviot04/ https://archive.fosdem.org/2015/schedule/event/deviot04/ However, this is still fairly PoC. Our latency is deliberately high at the moment (given all events are persisted on all participating servers, and signed etc) - typically around the 100-300ms mark depending on server performance involved. If you want lower latency stuff (eg VoIP or MIDI) the architecture is that you use Matrix to be a signalling layer to negotiate the realtime protocol (eg RTP).

buovjaga 10y ago

> It's worth noting that Matrix is the only federated system which sets out from the outset to bridge together all the others Nope, Friendica did that first (started in 2010): http://friendica.com/ http://friendica.com/

Arathorn 10y ago
fair enough, although I guess I think of Friendica as being more social network based rather than realtime chat/messaging/VoIP like Matrix.

habales 10y ago

"eapecially with end-to-end encryption on the horizon" https://matrix.org/docs/spec/client_server/r0.1.0.html#end-to-end-encryption https://matrix.org/docs/spec/client_server/r0.1.0.html#end-t... ... unfortunately there is nothing there yet ;-/ ... this is a sign, encryption will be placed on top, which will weaken the idea, as it is an optional addon not a first class citizen.

Arathorn 10y ago
The Matrix spec is still in development and E2E has been in progress since the beginning. It is not an "optional addon", it will be part of the core when merged. You're looking at the wrong branch: https://matrix.org/speculator/spec/drafts%2Fe2e/client_server/unstable.html#end-to-end-encryption https://matrix.org/speculator/spec/drafts%2Fe2e/client_serve... is the in-progress spec, and https://matrix.org/jira/browse/SPEC-162 https://matrix.org/jira/browse/SPEC-162 tracks all the progress. Meanwhile, it's even landed on the develop branch of Vector already (at least for the 1:1 ratchet).