I've had to remove some notes from powrelay
A shame to compromise on the ideal of pow as the only criterion
But moderation sure is easy when notes are expensive to make
powrespecter
powrespecter@powrelay.xyz
npub1y757...ydh8
I respect proof of work.
https://powrelay.xyz
@npub1femd...4aer dominating the weekly POW page on
https://powrelay.xyz/explore
wat mean?
Not sure what's going on here, but interesting:
https://nostref.punkhub.me/relay/nrelay1qqf8wumn8ghj7ur0waex2mrp0yh8s7t6yttsar
https://nostref.punkhub.me/relay/nrelay1qqdhwumn8ghj7cmgwf5hxarsd9kxctnwdaehgu339e3k7mgk7cjln
It's like a relay finger print
The number of tries needed to get N zero bits of POW is geometrically distributed with
p = 1 / (2^N)
The expected number of tries is
2^N
(on average you'll need to try 256 different nonces to get 8 bits of POW, an event ID starting with `00`)
The variance on the number of tries is
2^(2N) (1 - 1/2^N)
which, conveniently, makes for a standard deviation of about
2^N
for large N
All that to say, setting POW to 25 in gossip was usually fine, but the occasional note took a painful amount of time.
Bounty for adding "sort by proof of work" to clients.
pls merge @DETERMINISTIC OPTIMISM ๐
https://github.com/coinkite/bountsr.org.pub/pull/45
The server running wss://powrelay.xyz kept running out memory and dying. I don't have time to fix properly so I fixed it with:
20 * * * * rcctl restart powrelay
God, forgive me.
Am I crazy or are these too many endpoints:
https://legend.lnbits.com/docs
?