We are growing and #hiring a full-time and a freelance Senior Rust Engineer to help us build the next generation of private & secure messaging.The new Messaging Layer Security (MLS) protocol is at the core of our work. We are building a modern messagin...
-
@phoenix_r_d yet another chat protocol reminds me of: https://xkcd.com/927/
-
@krille @phoenix_r_d That’s why we helped start the IETF MIMI working group to get everyone around the table
-
@raphaelrobert @phoenix_r_d That solves the federation-problem but let me guess you are going to implement a new c2s protocol instead of using Matrix, XMPP or whatever we already have?
-
@krille @phoenix_r_d Yes, neither the matrix protocol nor XMPP protect metadata well enough unfortunately.
-
I knew it. the typical story: you think you can do it better than everyone else. Increase fragmentation instead of improving what already is available. Spoiler: everyone is thinking the same and in the end you will re-invent the weel over and over again instead of bringing us forward.
If you want to hide metadata, just help Matrix with encrypted state events.
If you want MLS, help Matrix with their shift to it (already planned but lacks resources)
-
@raphaelrobert @phoenix_r_d or help xmpp. At its core it is very minimal and flexible. You can invent new ways there to hide metadata, share it with the community, write XEPs and try to help everyone, adopting it.
There ARE opportunities here. Of course starting from scratch makes more fun, but in the end you will run into the same walls, the others already ran again. Then you are forced to do compromises. And others will say, that you did bad decisions and they need to start from scratch
-
@[email protected] picked my interest, I will keep an eye on this.
-
@raphaelrobert you picked my interest, I will keep an eye on this.
-
@krille @phoenix_r_d We've been having extensive discussions about this in various IETF groups, with both Matrix and XMPP folks. We are very open to either adhering to existing standards or helping shape new ones. Sometimes you have to start fresh though, that's exactly what we did with MLS and MIMI. I don't see why this couldn't be repeated for s c2s protocol. Either way, join the working groups and propose your ideas!
-
Hi @delta
I wanted to be sure you are aware there are now multiple project to implement MLS in Rust. @raphaelrobert is working on OpenMLS at @phoenix_r_d. It's being adopted to implement DM in #nostr by @1739d937dc8c0c7370aa27585938c119e25c41f6c441a5d34c6d38503e3136ef
-
@raphaelrobert @phoenix_r_d @krille Is ActivityPub still being considered for MIMI?
ActivityPub for Interoperable Messaging
The MIMI working group is chartered to define tools that messaging providers can use to interoperate with one another. The W3C ActivityPub protocol is already widely used for several use cases that resemble the MIMI use case. This document examines whether ActivityPub might be a good baseline for providing the sort of interoperability that MIMI intends to achieve.
(bifurcation.github.io)
It is a mature and widely used standard, but at the same time it's a blank slate because it doesn't have e2ee layer yet. Seems like a perfect foundation for developing a next generation of private & secure messaging.
-
@silverpill @phoenix_r_d @krille No, and I don’t remember that it ever was seriously proposed.
The MIMI documents are here: https://datatracker.ietf.org/group/mimi/documents/ -
@silverpill @raphaelrobert @phoenix_r_d @krille ActivityPub is actually a poorly designed protocol. Seems like there were no senior engineers in the team at the time of making it. And now it's too late.
-
@raphaelrobert @phoenix_r_d @krille Are you saying that draft-barnes-mimi-aim report is not serious? What is considered a serious proposal then?
-
@bonkers @raphaelrobert @phoenix_r_d @krille Do you know some other social protocol that organically grew to support millions of users and has more than a hundred of implementations? This is how a well designed protocol looks like.