Oddbean new post about | logout
 #OtherStuff is our chance to break away from the Influencer Paradigm and turn to an Information Paradigm. 
 #OtherStuff is our chance to break away from the Influencer Paradigm and turn to an Information Paradigm. nostr.fmt.wiz.biz 
 I’ve been thinking about this a lot. The core innovation of nostr is its signed event relayed model. This allows us to break apart existing systems and rebuild with new patterns that don’t lead to capture. 
 Yeah, instead of having one, neat table full of data in a big data center, copied to other big data centers, we split everything up into small, coherent pieces, replicate them, and disperse them, and then we pull it together to be displayed according to some construct.

We can pull it together any way we like, so long as we don't store it in a construct. If you store it constructed, you have to programmatically deconstruct and reconstruct it, every time you want to use it. Which is stupid. 
 Very similar to a UTXO, it is a 'construct' that doesn't actually exist as a cogent thing - only constructed in everyone's database.  
 A sort of abstract object, that doesn't persist, but only exists for as long as the client uses it. And the data storage doesn't contain that object, it contains the items that the object consists of and the pattern for building such objects. That means you can change out the items, and the next time you build the object, it looks different. 
 Database tables were a tuple abstraction. Now we have a signed event abstraction which we can build higher order constructions that look like database tables.  
 Errrr no, it's just a text string.  It's largely inferior to almost any other data structure that's even been invented, but it just about works.   
 It doesn’t have to replace databases writ large. Just the private sensitive stuff. I have already implemented a mini-database system for my wallet component and it works just fine. 
 It works quite well, actually.
Strings are an underrated data type. 
 Stringified JSON data for the win! 
 doesnt scale 
 Everything scales. 
 One of the reasons OtherStuff has been a dud, is because we've been pressured to transfer influencing to our clients.

If you open #Alexandria, it'll show you "stuff from your follows".
But what if your follows are the Primal Top 50 (because what else would they be 🙄 )?
Well, then you see a blank page. No entries. Alexandria ded.
Or there'd be 42 editions of "The Bitcoin Standard". 😅 

If you went by relay, topic, label, etc. you would eventually see lots of stuff, from npubs you might not know, but it would be the sort of stuff you were _actually_ looking for. Because you were looking for INFORMATION, not INFLUENCING. You were looking for SOMETHING, not SOMEONE.

OtherStuff will get nowhere, fast, so long as we aren't willing to admit that the person who owns the information you want, may not be Odell or Jack or ThatHotChick. It might be npubbunchofnumbers and it might be the only thing that npub ever produced. 
 The goal, in other words, for #Alexandria, is not to just show you "stuff from popular npubs" (although we'll do that, too). It's to allow you to type in "please show me ABC book" or "please show me XYZ article" or "please display everything you can find about breeding patterns of the prairie vole" and then go HUNT IT DOWN FOR YOU, WHEREVER IT IS AND REGARDLESS OF WHO WROTE IT, and be like,

https://media.tenor.com/aDN42XJePB0AAAAC/here-it-is-i-found-it.gif 
 Yes, but they'll probably often be different ones, than the ones you already had.