- E'veryn started slightly bending the rules and ran the five dungeons available to her unsynched. She will continue to do so until/unless she can't.
- Lea mostly ran cleanup around what portion of the world she has access to at the moment, only to realise someone else has to find her a pass to enter Bergen Mine; she did, however, bring Eva one step away from ARR level cap on Reaper, and possibly opened a boss door for Team Arias.
- Samus also managed to open a door for Kyuli and Zeke, but got herself stuck in a predicament until someone gives her a way out.
- Alucard tore through his father's castle, getting Eva to that ARR level cap and opening up a little more of the Shroud for her (and also Eulmore but lol). He also managed to find a new ability and Mouthful Mode for Kirby, but alas, not the one he needs.
- A hilarious number of further Waddle Dees have been rescued.
- Some overlay bugs have been squashed.
- A fox is ready to awake from his nap next time. Will he find a path forward?
Also, here's some output from the Typescript-based middleware I wrote that parses the Archipelago server output and communicates it to SAMMI via a crappy API server. SAMMI can't do websockets as far as I know, but it can do HTTP requests just fine, even if it's kind of jank to do. When the middleware detects a check ("Slot x sent abcxyz to Slot y"), it records that data to the server instance, then outputs it like this so that 1) I know it at least read a check correctly, and 2) so I can read it more easily. It also saves everything (in its original format) to a log file, and I have more functions that can parse those logs for manual data verification after the fact.
It's probably not proper to make SAMMI read from an API every quarter second, but if it's running on localhost, it shouldn't be too big of a deal, right? Anyhow, the API provided by the middleware has two endpoints; one to report current totals, and one to report how many checks there've been since the last time that endpoint was called (or HTTP 204 if none). Both report the number of checks received and sent by each game, as well as the total number of checks as reported by that endpoint.
In short: it's all a hot mess, and at least some of it works. The only way to know for sure is to actually run it, and it's too damn late tonight to try that, 'cause I'd pretty much have to generate a new multiworld for testing purposes and set all that up.