we had terrible onboarding. so bad that we created a manual "suggested users" list, which was effectively king-making. the verified badge followed. all centralized, manual, inherently biased decisions.

search is the only thing that really scaled, and really mattered. more investment in that would have dramatically changed things for the better at Twitter. this is still the case.

the only thing the centralized internet (google, reddit, Facebook, twitter) solved was the discovery problem. solving that for decentralized protocols is a massive win for the free internet.

nostr:note10n24cghex503ppu78ug69sheyzvtvppeggmpulteksaw7cze8uhs7rcx33

Reply to this note

Please Login to reply.

Discussion

If you had to prioritize, search or zaps first?

Search.

thank you so much for the elaborate answer 💜

I’m trying to understand, there currently no good way to search decentralized protocols? Like none?

try to find a specific note of mine mentioning word X from date Y only using damus, snort, coracle or amethyst

• don’t use any external search clients

Yeah doesn’t work. I’m no dev so forgive the dumb questions.

I’m guessing you’re referring to not using nostr.band

Why can’t that be used for search?

What’s holding it back?

analogy:

i’m saying that you wouldn’t go on google to find something that you’d want to find on twitter as you’re using it

Coracles search is pretty good.

I’m not very Smurf but why can’t this be used in clients:

https://advancednostrsearch.vercel.app/

☹️ I'm not very Smurf either. I really wanted to be one of them as a child. 🫂

😜😂

I understand.

you’re gonna make me code something

Outside of using NOSTR.band this simple thing is so hard to dom should be a part of every client imo

Why no external clients? That’s like every website trying to build their own crappy search engine when they could just plug into Google search.

Don’t threaten me with a good time

Nostr.band and nosWhere are decent tools, but need to be better.

And yet the Internet was never supposed to be centralized; this is what began the downfall.

🤫 i miss u

🤙💜⚡️

onboardstr.vercel.app !!

#🔎

And the search came from a company, Summize, which had a real time social search engine which they launched using twitter data. It really made twitter work. Buying summize was one of the critical early decisions at twitter which has been forgotten but was so important.

https://techcrunch.com/2008/07/15/confirmed-twitter-acquires-summize-search-engine/?guccounter=1

Indeed. All possible due to the open api.

Again, it’s nostr’s advantage right now. Every other service is closing down access unfortunately.

Can a search index of post content be decentralized yet reliable and fast?

Is the index the right place to slice the problem?

Is a centralized search service the only practical solution?

I agree search is essential, and at present the search available in amethyst is great but unpredictable.

If we start indexing the web based clients will get negative score by the crawlers for "duplicate content". This will be something very interesting to watch and a b..ch to solve. Whichever client gets notes indexed first will get positive score and each client after that negative score.

#twitter history from expert-micro-nano-blogger of the world - he uses 86 words 570 characters - understand importance of it. #plebchain #[0]

Nostr is so cool, if you decide to dev something, you will FOMO on other tools 😂

Concur. But calling discovery the "only" problem that centralization solved makes it sound like a minor problem. It's the principle obstacle. Start with discovery first when you take on the decentralization challenge.

Divide and conquer is how we win.

Without search, nostr is indivisible at scale. It remains one big community that falls under its own weight.

The "other stuff" is a multiplier, it requires social to be greater than zero.

GM ☀️

nostr:note17u7admqvzsv3ynjee35g3m5gst0k8rksn3pv065awtjp7xzupeys25h5y5

Isn't this still a problem?

The first thing people usually see is Jack when coming to Nostr, and fumbling their way through clients. If I didn't know better - I would assume he's the king of Nostr.

The ability to discover things is... Moving slowly. But at least it's moving.

NGL... A lot of people only stay because of "celebrities" - basically Jack and Snowden. And others use them like a small shield of clout protecting parts of their slow progress from ridicule.

"Oh this sux..."

"Yeah but we got rich and famous people that might talk to you to keep it alive."

That's terrible. Not sustainable. (Imo)

-- personally, I mostly stay because it's not mainstream, and the 'like' buttons can be any emoji... And it's not character limited.

nostr:nevent1qqs0w0wkasxpgxgjfevuc6yga6yg9hmr3mgfcsk8a2wh9eqlrpwqujgpz3mhxue69uhhyetvv9ujuerpd46hxtnfdupzpq35r7yzkm4te5460u00jz4djcw0qa90zku7739qn7wj4ralhe4zqvzqqqqqqyvl5fku

Zaps are stupid.

Lmao why?

Hmmm a suggested user list with the same 20 people? Sounds familiar 😂

Using clients without trending lists and recommended users list fixes inFLUenza and king making crap.

I wonder if nostr search should still be centralized for its efficiency advantages, but completely open source so that anyone could spin up an alternative at any time.

“You are here” diagnosis on Damus.

🫂

📝

I'm still a relatively new developer. Beyond algorithms, what resources would to recommend to someone like me to study so try to contribute you this area?

So much alpha lol

It wasn't that hard to get on to NOSTR, but I am also a bitcoiner who understands the concepts of nodes and private keys. Maybe that makes a difference.

NOSTR struggles primarily due to weak network effects and nuanced benefits.

Although I think it achieves its goal perfectly as a protocol for freedom centric communication. We have to be careful what we wish for - sometimes things are amazing just the way they are.