The Nostr NIPs are becoming a train wreck. Removing stuff in use is the best way to cause a mass exodus of devs. Google found this out with early versions of Angular. Its the kind of things devs (rightly) moan about for years. And of course completely unnecessary. Developers cant build on quick sand. But OTOH if there's a small group that just wants to do their thing, and dev exodus is acceptable, that works too. They are a talented bunch, but I feel more devs = better nostr. We'll seeeee...
Discussion
I agree. I don't believe in taking things out or even changing things for a live protocol. Just deprecations and strict additions.
Unless this is just an experiment and we are planning to start over... but people aren't treating it as an experiment, they are pushing to get more people onto nostr.
i am a kind4 disrespectoor, but i may be capitulating i dont know. i have to use ndk and have custom super-gossip protocol to deliver kind1 or kind4 anyway. eg. dm4 doesnt work unless you send it to the target users read relays, same for kind 1. but outbox pure-ism i believe just delivers em to my write relays 😂 where they will never get or respond there... and then who knows where the reply will go, usually only to damus relay, which is extremely limited now so, ya. inbox so quiet these days till people get the 44 setup..
basically i will just keep sending kind1s instead of dms until someone says stop.
😂 re: "basically i will just keep sending kind1s instead of dms until someone says stop"
I absolutely will not simply replace NIP-04 with NIP-44. The referenced PR is a fine example of whats going wrong behind the scenes.
c'mon, man. the ceo of nostr says that's the way we should go. why won't you just conform?
Its fine if the NIPs change, I don't mind unless its a mandatory breaking change. I say grow a thick skin to those who can't handle constant breaking changes - that's the way its been on a lot of lightning and a lot of projects in the space.
Dev talent is super thin, globally. We need to work harder on getting more young devs involved, none of us are gonna live forever.
Yes, this is exactly how to lose a developer base, as so many projects have found out the hard way
IMO if people still want to use NIP4 let them use it, but in my head this design is wholly unsatisfactory.
Whether NIP44 is the long term solution is up for the market to decide; the encryption enhancements and analysis is worthy enough to give the spec a shot at something better than what we have now.
I don't know if this is going to impact the existing developer base but I'm willing to bet that they are hardcore enough to take it; they're not soft at all. ;)
I still didn’t know why they use numbers if JSON really don’t care if they used NAMED_STRINGS in place for kinds.