#growNostr
StayFriends (German classmates.com) lets you write to an old classmate for free, but he has to pay to read it. LinkedIn ate their lunch by accident. There is no working "find your class of '96" tool left anywhere. I want one and I think it is the best normie onboarding funnel for nostr nobody has built yet.
The constraints that make it fun:
1. Your classmates are not on nostr and never will be on purpose. The word "nostr" shows up only after the user used the tool: "btw, this login also works in Amethyst/Jumble/Damus".
2. Google has to be able to find it. Crawlers do not open websockets, so relays are invisible. A branded njump fork with one aggregate route ("everyone who tagged school Q123456, cohort 1986") does the indexing, and the kind templates can go upstream. Now, somebody googling his name can find the tool and might engage.
3. It should not become a doxxing honeypot. About other people only what a yearbook says: name, school, cohort, class letter.
Un-paywallable by design. Claims live on relays, the indexer is MIT, anyone can rerun it. The data is never worth less to the user over time.
Generic model ("I was at <geohash> during <period>"), schools-only v1, because school is the one category where ~30 people knew each other by name and the privacy math works. Festivals, company units, city blocks could fit the scheme.
StayFriends (German classmates.com) lets you write to an old classmate for free, but he has to pay to read it. LinkedIn ate their lunch by accident. There is no working "find your class of '96" tool left anywhere. I want one and I think it is the best normie onboarding funnel for nostr nobody has built yet.
The constraints that make it fun:
1. Your classmates are not on nostr and never will be on purpose. The word "nostr" shows up only after the user used the tool: "btw, this login also works in Amethyst/Jumble/Damus".
2. Google has to be able to find it. Crawlers do not open websockets, so relays are invisible. A branded njump fork with one aggregate route ("everyone who tagged school Q123456, cohort 1986") does the indexing, and the kind templates can go upstream. Now, somebody googling his name can find the tool and might engage.
3. It should not become a doxxing honeypot. About other people only what a yearbook says: name, school, cohort, class letter.
Un-paywallable by design. Claims live on relays, the indexer is MIT, anyone can rerun it. The data is never worth less to the user over time.
Generic model ("I was at <geohash> during <period>"), schools-only v1, because school is the one category where ~30 people knew each other by name and the privacy math works. Festivals, company units, city blocks could fit the scheme.
81❤️1
