We've been hearing about Loops being open sourced (which would imply the ability to be federated) for months already. Just publish the thing and let the community help, @dansup@mastodon.social !

[–] Jumuta@sh.itjust.works 2 points 5 months ago (1 children)

maybe he's worried about people implementing a premature protocol?

[–] rglullis@communick.news 28 points 5 months ago

One more reason to be asking for help from the community and to be doing everything in the open.

He doesn't need to know everything. No one is expecting him to deliver flawless software. But I'd have place more trust on someone that works in the open than someone who keeps saying "next week!" out of fear of being judged by the initial quality of their software.

[–] Coelacanth@feddit.nu 11 points 5 months ago (4 children)

I know this is a contentious topic around here but what's the news of Loops' recommendation algorithm? Does it have one?

[–] gedaliyah@lemmy.world 19 points 5 months ago* (last edited 5 months ago) (2 children)

If it is open source and on-device (for personal recommendations) then I see it as a great thing.

Lemmy has demonstrated that it can really work. Quiblr even has personalized recommendations on-device (~~not open source though~~ and it's now open source).

Mastodon and others should take note.

[–] flamingos@feddit.uk 13 points 5 months ago (1 children)

Quiblr actually is open source now: https://github.com/Technicolor-Dreamcoat/Quiblr

[–] gedaliyah@lemmy.world 2 points 5 months ago

Thanks! I must have missed that one.

[–] Coelacanth@feddit.nu 7 points 5 months ago

I agree. For stuff like Mastodon and Loops I really want "For You", not a simple chronological feed.

[–] flamingos@feddit.uk 14 points 5 months ago

Irrc, a 'For You' algorithm is on the road map, but it doesn't currently have it.

[–] kobra@lemm.ee 6 points 5 months ago

Probably won’t have one for a while. Pixelfed still doesn’t have one afaik and that’s another of their projects.

[–] meldrik@lemmy.wtf 2 points 5 months ago

No. Posts are chronological.