Nostr Web Client

Agreed 💜 Leaving users in charge of their experience is key.

I think a simple “I don’t care about” with a few pre-made filters like trump, Bitcoin, NFTs, NYT, etc during on-boarding to enable client side filters like below is a reasonable approach.

Keeping in mind just because someone may post about a topic like Bitcoin 1 in 20 posts, does that mean you shouldn’t follow them for their unrelated content? Filtering topic based content is more valuable at the content than author level - unless the author is strongly tied to a topic - like a Bitcoin Price bot or something.

nostr:note1mlh0wuqh7kdcjueaq7qedfaeunq9kqy02hpf2yaslekxl5sz5tzsjczaev

Reply to this note

Please Login to reply.

Discussion

elsat 2y ago

Great point. Keyword mute is on the bench cc #[5]

corndalorian 2y ago

I agree with the idea of enabling user-controlled content filtering Anything suggested as an “I don’t care about” topic also suggests a particular slant for the app/network, especially if the suggestion is an already polarizing topic like Trump or NFTs. It’s a dangerous slope. If my onboarding process said “you can filter out content like ‘Trump’ or ‘Biden’” then I’m immediately going to assume the political/ideological leanings of the app creators and perhaps the network. I don’t think those kinds of suggestions should be made, but I do think the tools to filter them should be there, if that makes sense.

Blake 2y ago

Yep. I don’t have a specific way to generate suggestions of what new users may want to see or filter. Typically noisy (by volume/content) or controversial are the topics I’d imagine make sense. However, even just a toggle to filter NSFW by default is a start - that could include 20 or so common explicit hashtags or words. The art will be a balance.

One way could just be a past 24 hour summary of popular/common words or hashtags based on the relays you connect to.

Another could be a guided tour on how easy it is to add a filter - like long-tap a hashtag, or highlight a word/phrase.

Maybe some kind of “see less off Y”. Maybe with randomness. Maybe this is where custom shareable filters and dynamic client evaluated views fit.

I do think Nostr has lost a lot of people just after onboarding due to global being Wild West and it being less easy to discover or port over who your current follow lists from elsewhere.