From 285059a1bf2cb4acf51b76eaadc7ae1a59cbbfac Mon Sep 17 00:00:00 2001 From: Daniel Oaks Date: Sun, 2 Jul 2017 07:55:02 +1000 Subject: [PATCH] info: Add S2S plans/info/etc --- docs/INFO.md | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) diff --git a/docs/INFO.md b/docs/INFO.md index 52ca682e..a480bc57 100644 --- a/docs/INFO.md +++ b/docs/INFO.md @@ -1,6 +1,11 @@ # Oragono Information -Here's a bunch of misc info the Oragono server! +Here's a bunch of misc info about the Oragono server! This can include questions, plans on +how I'm going forward, how to properly use features, or why Oragono does/doesn't do +something. + +Essentially, this document acts as a braindump about Oragono while we figure out a better +place to put all this information. ## Accounts and Channels @@ -20,6 +25,23 @@ legacy interfaces to access these functions. However, it's gonna be a while befo this is specified by someone like the IRCv3 WG. +## Server-to-Server Linking (or Federation) + +Right now Oragono doesn't support linking multiple servers together. It's certainly planned, +but it's a fair while away. + +When I do add S2S linking to Oragono, I want to use it as a testbed for a new sort of +linking protocol. Mostly, I want a meshy protocol that minimises the effects of netsplits +while still ensuring that messages get delivered, and preserves the AP nature of IRC +reliability (in terms of the CAP theorem), which is something that traditional solutions +based on the Raft protocol don't do. + +Basically, I'm going to continue working on my [DCMI](https://github.com/DanielOaks/dcmi) +protocol, get that to a point where I'm happy with it and _then_ start looking at S2S +linking properly. If anyone is interested in server protocols and wants to look at this with +me, please feel free to reach out! + + ## Rejected Features 'Rejected' sounds harsh, but basically these are features I've decided I'm not gonna