@srs @bmann interesting, thank you! moa.party could perhaps be used to implement this? Assuming Twitter won't let us grab or interop the whole timeline for a user, but individual users could opt in into a timeline experiment that processes the opted in subgraph in Mastodon land. A series of servers that essentially are better Twitter views.

@srs @bmann client side mods are likely much easier to pull off, though, at least initially. I guess I just really like paths that eventually could lead to full [[siphons]].

@srs @bmann then again it might be I just don't think in client side terms; presumably a twitter client could also siphon out a timeline while rendering? Could twitter ban this? What if it's just a browser extension that scrolls for you? It seems hard in the limit.

@srs @bmann anyway, I guess I jumped straight from moods to revolution :)

@flancian @srs viva the agora!

@flancian @srs I think opt in to siphon curation leads to re-usable data streams.

Wait a second. Push a PR to my GitHub repo???

@bmann @srs right -- a la [[promnesia]]. You could run local siphons, no need to go through an [[agora]] first as that has privacy implications; the [[agora]] can be the "social" target after private curation/filters previous to the data getting to your public github repo.

@karlicoss has *a lot* ready in [[promnesia]]: anagora.org/node/promnesia

@bmann @srs it would be easy to just directly fetch stuff from [[agora server]] to the user, but that means the user has to trust an [[agora]]. There should always be an option to do filtering yourself and only then submit to a social platform, I think.

This is where Markdown on Git comes useful: it is such a simple convention to follow, I think.

@bmann @srs if you put Markdown on Git, the [[agora]] can ingest your content no matter its provenance.