Trying to manually clean up 30k pubkeys into a browsable feed is too much. I can't do it. Some hops out on WoT, people start following all sorts of spammy, weird, and nasty things. I just can't get ahead of it.

nostr:nprofile1qqsw2feday2t6vqh2hzrnwywd9v6g0yayejgx8cf83g7n3ue594pqtcpz9mhxue69uhkummnw3ezuamfdejj7qg4waehxw309aex2mrp0yhxgctdw4eju6t09uq36amnwvaz7tmwdaehgu3wvf5hgcm0d9hx2u3wwdhkx6tpdshsd8m8kr nostr:nprofile1qqs8eseg5zxak2hal8umuaa7laxgxjyll9uhyxp86c522shn9gj8crspz9mhxue69uhkummnw3ezuamfdejj7qgjwaehxw309ahx7um5wgerztnrdakj7qgkwaehxw309a3x2an09ehx7um5wgcjucm0d5hsvlnggv ...I tried. I'm going to break the stream on aplaceinthesun, after I figure out where I'm going to read & write in replacement. Then, I'm going to try expanding the allowed keywords on aquaticchickentheory and see if I can do something interesting with that. The dual allow criteria should make it relatively simple to clean up popular hashtags. I can't currently republish nor stream recipes there anyway, so its basically just an occasional reposted picture of soup, as it is. πŸ˜…

Reply to this note

Please Login to reply.

Discussion

All good, yeah I think more modes are needed (that leverage lists and follows) to make the relay be more of your use here. The raw feed is good for discovery but yeah, not maybe what you (or I) want for more curation. I've got some stuff cooking on this. and also @david was talking about introducing more data into the brainstorm feed likes and zaps for example (similar to spatianostr), which could also be a way to do more styles of feed.

Yeah, we're not quite there on "clean" discovery or strict curation yet, as far as user level goes, but it's getting there. I'm not absolving myself of these experiments, just taking a different approach for now

I changed my mind. I blocked kind 1 :)

hahah! checkmate, dawn wins πŸ˜‚

For a minute, i felt like a genius πŸ˜‚ I'll see how it shapes up after a while.

Your personalized whitelist is about 30k. Would it help if brainstorm were to make some smaller personalized lists available? How about your top N pubkeys, not all 30k? Maybe top N, next top N, next top N, etc? Or how about top N pubkeys but not including pubkeys you’re already following? Which would be a way to discover novel follows.

It would be relatively easy for me to add these lists to the API so relay tools could download them.

That might work. I'll break my dependence on the experimental setup and think on this. I need to go through my Brainstorm and see how far out we make it before things start getting really weird. Then we can try something else with it. Top N pubkeys minus follows might be really good.