definitely don't sign up for this and enroll yourself as a subject of a prediction:
https://haruspex.shit.nexus/flash
Replying to @silverpill@mitra.social
Replying to @silverpill@mitra.social
@sun So, I think the biggest problem with this proposal is that it requires a publicly accessible HTTP server. Is there a good reason for that?
WebXDC, for example, is designed for offline first apps. I am reading WebXDC docs right now and it seems to be compatible with ActivityPub. Maybe we should work on this instead?
Replying to @sun@shitposter.world
@sun Mitra doesn't support scopes, but I can start adding them if needed.
I saw your FEP draft, by the way. Is it ready for review?
Replying to @silverpill@mitra.social
@silverpill wait, what the hell, WebXDC over fedi would be (almost) perfect, I need that!
The only friction point I can think of is how are we gonna do window.webxdc.joinRealtimeChannel ? I think Delta uses Iroh for that, and, ehh, maybe, but I'm not immediately on board with that. Maybe we can just ignore it and be fine with async apps?
@sun
Replying to @silverpill@mitra.social
Replying to @sun@shitposter.world
@sun @lnkr I think the equivalent of realtimeChannel can be implemented over ActivityPub too - client-to-client encryption is not very difficult. Iroh is not magic either - when it doesn't know how to connect to a peer directly it connects to a relay operated by the iroh team. I suspect this relay is used for the vast majority of iroh connections, so ActivityPub might be actually a better choice because AP servers can act as relays and we already have so many of them.
Codeberg.orgfeps/0806/fep-0806.md at mainfeps - My FEPsReplying to @silverpill@mitra.social