Discussion
the quality of NIPs have severely degraded, tbh…
originally it was core specifications. then some small apps.
then we saw the boom of devfluencers that build and abandon apps.
specs now only focus on very niche use cases and don’t try to be interoperable or open to other use cases
a few months ago my access to the NIPs repo was removed without any notice.
as someone who has been here before the hype and have developed a good amount of the standards and infra here, it’s pretty disheartening to see Nostr turn into the messy “ecosystem” (if you can even call it that anymore) that it is now
I am most likely going to begin making my own group of standards, the “standardize before implementation” approach never works with NIPs anymore with them being stuck in review limbo alongside reviewers trying to impose their extremely narrow views on how they think Nostr should be
Every dev has this thing of building their own standards. Then they realize their stuff doesn't work unless they implement what others do.
So we just host our own nips repo then? I already run a mirror. I'll merge your patches XD
Yes.
These will contain proposals like my Nostr Login proposal, moderation and translation service interfaces, request cursors and a bunch of other things I will implement on my own Nostr services and that others would benefit from also implementing.
Vendor specific details if they were to exist would be in another folder.
you can write your nip on nostr https://nostrhub.io/
🤞
You can just have your own nips repo 😄
i'm interested in discussing the finer points of any proposals you make. i also have my own designs that i'm working on. you saw my first draft http api, i'm refining it now and will have subscriptions implemented probably eod tomorrow. uses a single listener for the client with a relay assigend ID and then subscribe/unsubscribe methods to activate/deactivate filters on the listener.
i don't really care that much to discuss issues around message types. i think you know i'm opposed to the whole kinds field altogether for the same reason as i oppose the numbered nips. i think document types should follow the HTTP standart.
nips have devolved to mix protocol and application concerns. id undo most of them and freeze it. less is more. it is what it is.
Kind 1 sucks. It's fully broken. The threading system requires clients to do so much fixing or miss half of the conversation.
when are they going to deprecate some NIPs and make "Nostr 2025 Standard" like programming languages do?
the beauty of Nostr is that you can do that right now yourself
Agreed. Kind 01 is terrible. That's why we don't use it in any app. Total nightmare.
What's your thread spec then?
🅰️ 30040 with a thread category?
🅱️ a kind 11 with deep threads?
Other? None?
I'm open to anything better than kind 1.
Or kind 1 with 1111 replies, that too.
Kind 11 with kind 1111 replies.
With logic to display replies of the author to his OP as an actual :thread: ?
Still my most pressing need with this: interior on the actual thread display.
Kind 11 OPs and threaded Kind 1111 replies.
Is that what you meant?
No, I mean :thread: UI:
🟣 Author
| Text of OP
🟣 Author
| Text of auto-reply 1
🟣 Author
| Text of auto-reply 2
Etc...
Then replies from others..
Twitter threads concept.
Ah, I see. I'm calling the tab "Discussions" and just using kind 11 for the feed, instead of Kind 1. And I have a sort of bulletin board or Reddit design.
But, you are right. Threads would be a string of kind 11 from one person. 🤔
Everything, including 30040 with 1111 replies, except Kind 01. Kind 01 gets Kind 01 replies. I only use them on the Events page, not in feeds.
give an example