Didn't we already figure out why storing binary in json is a bad idea back in december? I mean seriously?
Discussion
Yes! ๐ถ๐พ๐ฏ
I wasnโt here lol pls fill me in
NIP-94 and NIP-95 ๐ถ๐พ๐ซก
Wait 95 passed ?
Amethyst is plowing ahead! ๐ถ๐พ๐ซก
Wtf ๐คฃ whyyyyy, is that normal ?
It shouldnโt be normal, if we want to develop a robust protocol! ๐ถ๐พ๐ซก
Certainly, agree with you! Hopefully he stops tryna get ahead of himself & the protocol ๐ซก๐ซ๐
Thatโs why I am running around and campaigning for people to take a notice! ๐ถ๐พ๐ซก
IMHO Vitor can do what he pleases. If it serves Amythest users somehow, it's fine. It will have no effect on my client or my relays or their interactions with other people's clients and relays. It's additive and ignorable, right? Am I wrong?
If it was only about client. The changes are done to the common protocol, that is a problem IMHO! ๐ถ๐พ๐ซก
Isnโt it the way around ? It affects all other users as they canโt see the visuals hence impacting interactions on Nostr as a whole ?
Precisely! ๐ถ๐พ๐ซก
It only affects amethyst users. All other clients will simply not see their image posts. Damus will likely hide empty posts with unknown event references since it looks horrible.
The code change only impacts amethyst. But the negative impact/aftermath is on all other clients and users.
-ve impact on users :
When users canโt see content, they are at a loss. If this becomes frequent, Nostr becomes contentless and pointless.
-ve impact on clients :
Client devs cannot be stopping everything and implanting new changes everytime one client makes whimsical changes. Hence you have every right to be upset
Use cases of NIP 94
The various use cases Vitor brought up should be explored and tested. But not yet implemented until all clients agree as it impacts everyone.
Consensus
There has to be some ground rules that requires consensus among Devs when it cannot be done on NIP itself. For example :
1. Major changes in protocol
2. Unintended NIP implementations
3. Changes that impacts users on other clients
We already survive with lots of impacts like this already. Gossip can't see zaps, but it is still usable. Many clients can't see all the posts of the people they follow, but they are still usable. Many links can't be followed because the client cant find which relay that event is on. We can't count all the 'likes' or all the zap receipt totals in this decentralized model. Polls don't show up in gossip. Some people don't see note "subject" lines. Etc.
Fiatjaf's saying "embrace the chaos" applies.
I do think we do need very good interoperability where core functionality is concerned, and I would like that core functionality to slowly grow and harden. But there will always be a large set of optional features only available in a scattered set of clients and/or relays. I'm counting NIP-95 as an optional thing that I won't participate in.
YES, users will be confused about all of this. I don't know how to solve that problem. Forcing everybody onto the same page isn't a realistic solution, we are cats and you cannot herd cats.
Probably having a stable protocol that takes into consideration the adoption rate? Not like this is new! We changed common protocols before. ๐ถ๐พ๐ซก
Check dms #[5]โ ๐ซ
Yep I see them. Can't reply to DMs with gossip though, may have to spin up one of those web clients that wipes my contact list ๐คฃ
Oh mannnn ๐คฃ๐คฃ๐คฃ donโt worry reply here itโs no problem donโt lose your contact list!!!
Reserve 4AM EST for me. That's my 8pm. Now I have to find nostr nests (or hit me up on telegram @mikedilger)
Sounds good!!! Canโt wait!! & donโt worry you donโt have to sign in just edit your profile when you join, Iโll message you on telegram to walk you through it. Very simple ๐ซก๐ซ appreciate you! https://nostrnests.com/frogtalk
Yea Iโm with you on distinguishing feature and function, and having an interoperable core functionality. I think not having features like zaps, likes, DMs is diff from functionality of not being able to view the content (unless user intended for it to be private, based on relays).
Example of practical use cases : if media were to use an amethyst userโs content as reference - nobody else can view it. Eventually it ends up forcing ppl to use 1 client which defeats the purpose decentralisation and multi-client environment.
If Amythest uses the feature in a way to just link files in normal text note posts, then then yes it would be a problem.
But I think he's trying to replace github with it, so it won't break the normal client interactions.
If I could go back in time I would push to make nostr a binary protocol. Validating the JSON signature is not easy because JSON is not data... it's code, and there are multiple ways of encoding the same thing so you have to normalize it before you sign or validate a signature. ๐คฎ
๐คท๐ผโโ๏ธ https://github.com/nostr-protocol/nips/pull/112