https://frames.weponder.io/api/polls/3247/frame

50% of people think tips are spammy sometimes. 10% always.

Wonder if there is a better convention that allows Farcaster clients to better classify / collapse tips into a more useful UX for those that want to see them and hidden for those that don't want them.

Parsing $[token] replies with some regexes could work but a bit naive.

Tips were always meant to grab mindshare via visibility.  Any mechanism to reduce visibility will reduce usage.

This means the goals of leaders of tipping communities are at odds with non-tippers.

TBH I think mute words are the best solution to allow these groups to coexist.

Startup artist seeking inspiration.  Champion of /bcard

Haha that makes it your problem, eh?

Assuming they ARE churning because of tips, regex is the fastest way, but this problem will keep popping up with new kinds of content.

How about change « report cast » to « categorize » and have options for spam/tip/scam, etc?  Then you can train AI models based on user feedback and show posts based on user settings.

For instance default group « tips » behind a link (similar to « see more ») but allow users to adjust which categories they see.

It’s a shit ton of work for not a ton of value IMHO but you asked.

Most people won't do the work for mute. They churn (silently).