Many people understand this as "retraction". Terminology might help resolve the debate over whether you can "delete" something in nostr. nostr:note192w58wm5zck35gw66ve26s00fujtkxul95rz2uzfta64xf5q2xdsxzye5u

"Retract"?

Markets are complicated, and liquidity pools can lead to weird behaviors. Could you ask a more specific question?

I actually called it "X" this morning, and it seemed fitting.

One real, and one imaginary

The fact that you can pull down a project that you just learned about and ask Claude to implement a technical spec that you haven't published yet in a language that you haven't used in years is... disorienting. #llms #ai

The real question is how we will find the right events

Backblaze B2 will store a petabyte for $72k/yr

From Snow Crash:

“The business is a simple one. Hiro gets information. It may be gossip, videotape, audiotape, a fragment of a computer disk, a xerox of a document. It can even be a joke based on the latest highly publicized disaster.

He uploads it to the CIC database — the Library, formerly the Library of Congress, but no one calls it that anymore. Most people are not entirely clear on what the word “congress” means. And even the word “library is getting hazy. It used to be a place full of books, mostly old one. Then they began to include videotapes, records, and magazines. Then all of the information got converted into machine-readable form, which is to say, ones and zeros. And as the number of media grew, the material became more up to date, and the methods for searching the Library became more and more sophisticated, it approached the point where there was no substantive difference between the Library of Congress and the Central Intelligence Agency. Fortuitously, this happened just as the government was falling apart anyway. So they merged and kicked out a big fat stock offering.

Millions of other CIC stringers are uploading millions of other fragments at the same time. CIC’s clients, mostly large corporations and Sovereigns, rifle through the Library looking for useful information, and if they find a use for something that Hiro put into it, Hiro gets paid.”

Same people that ran BBS's, or Hotline https://en.m.wikipedia.org/wiki/Hotline_Communications

This would be an interesting LLM-bot task. Use graph analysis to find statistically interesting nodes, then see if a few hundred Claude tokens can figure out what the connection might be and post a note with the results.

The problem right now is that we're only paying attention to the new entries. We need clients that explore the graph in deeper ways. Nostr has that early web energy.

This type of complexity is a common problem in object oriented programming. "This" type has a method that accepts a "that" type, and it works well until there are too many combinations of types. Any time a new "that" is created, the original code needs to be updated. This is where OOP reaches for interfaces, so the "that" type can be anything that claims to be a "that". This works well for even longer, but it still tends to fail at scale. Any time a new behavior is require, the interface changes and now every implementation needs to be updated.

I have a soft spot for old systems, and one of them is Common Lisp. CL was standardized before OOP, and one of the ways it deals with this is multiple dispatch. Instead of "this" type of object having a method operating on "that" type, the function is declared abstractly in a namespace. "Intersect" would be a function that finds the intersecting parts of its arguments.

Now when someone makes a new "that", they include code that knows how to intersect "that" with as many "this" types as they find useful. No original code needs to be changed. Later, someone working on "other thing" needs to intersect two types that no one has written intersection code for, so they write their own implementation that is applied to these two other pieces of code, and neither of them needs to be updated.

Complexity can quickly get out of hand, but it's also something that we've been thinking about for a long time. The more we learn about history, the more opportunities we have to find good solutions to new problems.

A subkey proposal: https://github.com/nostr-protocol/nips/pull/1450

Is there, though?

Kick the football, Charlie Brown!

Encryption is the way. If we can prevent relays from knowing what's in a message, they can't censor it. Unfortunately they also can't index it, and only people with keys can read it. There are problems to be solved, but the benefits of relays are worth it.

This chart is also misleading when it says "Password hash: bcrypt". Bcrypt is used with a work factor that should be set as high as practical for your server. Not listing the work factor means that the absolute numbers listed are meaningless.

Tor is a web browser, so you use a web client. https://nostr.com/clients lists a few.

"Providing cryptology services aiming to ensure confidentiality without certified declaration"

Is this where we're at now?