i like how your dog has an arrow at the back of his head, that's pretty unusual.

quite funny how it folds its tail down so much ... little dogs often are like that tho

chronos is one of my favourite electronic artists... i don't even know how to describe the style, it's like dark ebm psytrance

https://youtu.be/VKv6NT8zEVo?list=OLAK5uy_k7YxLu93UPX8qokJQvHun_lYkhm_YjEDc

yeah, literally thai baht

lamers

i think that the mindset of the psycho corporate mind manipulators muddies the waters and that actually things will organically gel a lot better than you think in a self emergent system, just like brain cells are cooperating with each other and even two whole hemispheres with different biases cooperate in brains

the control of signal propagation exerted by the silo owners is to benefit their own interests, and it is very similar in its influence on the human social network as a tumor is to a brain - it consumes space and resources, and it gets in the way of connections forming between systems

i see, you can back that up with some investigation then i presume

this is the latest work from a french distributed systems programmer with the handle lesterpig

https://arxiv.org/pdf/1911.03291.pdf

it is a distributed database byzantine fault tolerant, eventual consistency model, and it's something that i've been interested in getting into more since 2018 when i first learned about it

calvin looks cool and i'm gonna research both...

the purpose of this is to find a way to make byzantine fault tolerant database replication federations for nostr data storage backends, that don't use shitcoin pBFT derivative trash that is outdated and crappy

i'm building such an interface now on a grant given by such a project but my bigger picture is to figure out how to solve the consensus problem for *at least* federations of nostr relay operators, preferably one that requires minimal manual interaction from the operators as a collective, and minimises politics and other kinds of human byzantine faults in the equation

some more on distributed database replication protocols:

https://betterprogramming.pub/demystifying-consensus-algorithms-and-their-implementations-c52f8aca3020

yeah, this is why i parameterised this part in my ingestr event mirroring tool... because if you constrain the time enough it gets through, but it's a different thing for each relay so it needs to be customisable per task

i also ran up against the issue of "too many concurrent subcriptions" on relay.primal.net so i added a pause parameter to give the relay time to expire the closed subscriptions

limits really annoy me... i just implemented a mirroring app and i think if the relay implements NIP-11 it can specify a max limit and you can stick with that but what happens when you make a filter that matches more than that limit within its constraints (mainly, since/until) how do you get the rest of the results??? protocol doesn't specify that, so i have used a more conservative strategy where you can define windows that you query on and specify your limits so you don't miss any of the results in that window

it's one of the several flaws i percieve in the protocol spec so far

there should be a special message in the ok envelope or something to indicate more results and some rules about how the results are cut into segments when they exceed the relay coded limit maximum, like, sorting by time, or else it's very undefined behaviour and results in difficult problems with varying implementations and configuration settings

but as far as i can tell the only thing that differs with it from leveldb, boltdb, badgerdb, rocksdb is its support of some kind of distributed replication integration

rocksdb is good, and bolt and badger are derivatives but only go versions and only embedded

there is also lmdb, which is used by strfry as well, so i'm wondering why foundationdb instead of one of these non-tech-company-developed C++ abominations that do the same thing except for replication

it looks also like the replication strategy is primitive, probably RAFT or Paxos derived, but it doesn't specify which is a smell for me

yep, more of them need to actually implement this, i've done basic implementation of it though not codified into the library yet (requires a little extra handling, i have added a signal channel that a client can wait on or timeout waiting for auth, it doesn't currently exist in the go-nostr codebase, but is now present in my forked, reorganised version)

there are two types of uses of the "humor" technique of sarcasm

one is that of the humor of psychopaths and sociopaths belittling their opponents

the other is that of the oppressed individual who doesn't give a fuck about being punished because their life is meaningless due to their oppression by psychopaths

this is a distinct type of humor from irony, which is self deprecating, and is usually easier to detect due to the divergence from the normal character of the individual

whereas sarcasm is usually expressed deadpan and is not funny when the one using it is actually a vain and cruel individual

psychopaths can't tell the difference between sarcasm and irony, non psychopaths can

indeed i can think of ways to process out the parts of a filter and segregate the results so filters that have partial matches can be partially satisfied by the cache and in some cases sufficient to produce the result

caches are hard tho... but i still don't get the choice of foundationdb

nice, not heard of foundationdb before - gonna have to look into that

i have a modified go reverse proxy running on my mleku.online which serves up my nip-05, my gitea instance and forwards requests back to my relay that is still in testing, that i keep running most of the time as my personal as well as testing relay

i assume this proxy has some kind of caching intelligence perhaps? kinda difficult to really cache filter results unless you can isolate the boundaries, perhaps it could if it stored the filter/results sets with a "current time until" so if a query matches inside that since boundary it can filter it... but maybe that's overkill

lol, sorry but fiatjaf's self aggrandising self-praise doesn't include irony so no, humility is not something he's cultivated adequately yet

being right is more important than being seen as being right

those who say it is "unfair" or "salty" or "ok" or other perjoratives to point out faults in the deeds of others are more concerned with being seen as being right when they clearly are not to the person who is doing the detective work on the evidence of errors

if you have nothing to hide you won't be bothered about people finding your errors, you will thank them

only people with ego problems where they can't face being wrong fight you about this stuff and they are the most important people to investigate and uncover their errors because usually somewhere in the pile you are going to catch them in deliberate error, also known as evil

so happy to see this

yeah, monero got delisted from binance

turns out that it's also useless as a payment instrument in a relatively short term too, i feel bad for any poor bastard that was trying to save in monero at this point, oopsie

#bitcoin only

nostr:note1fr4qmweqpkt5q94wevr3cmceeuvydv4ygvhcrhqe8q0src85p7xsrsmns2

makes me think of something i saw some time back, something about the idea that the annunaki were massive consumers of gold and they were enslaving humans to dig it up

something maybe about how they needed to eat gold because of genetic defects? that's just the thought that's coming to me now

not really meaning to be a brag but i'm pretty sure i've emitted more latin characters into the public arena than a lot of the big names and i doubt he is any different

i mean, i've measured my output and i have probably averaged over 5000 words a day for at least half of my life, that's several PhD theses worth of text, a stack of books as tall as myself (5'10")

people can, if they can touch type, and are talkative, blather more now than ever before in history, and i hope but i doubt that my record can all be seen, a lot of it can be, thanks to various archives and mailing lists and forums, but many of them died and my most numerous words were often written in public forums that didn't keep logs

i suppose that steiner was probably quite chatty whenever he was in company too tho

nostr:nprofile1qyd8wumn8ghj7urewfsk66ty9enxjct5dfskvtnrdakj7qgmwaehxw309aex2mrp0yh8wetnw3jhymnzw33jucm0d5hszxthwden5te0wfjkccte9ekkcettw5hx7mnvd9hx2tcqyqalp33lewf5vdq847t6te0wvnags0gs0mu72kz8938tn24wlfze66qnzkp this is a demo of the same bug you are experiencing

https://v.nostr.build/87qG.mp4

it is choking on 10002 events in both cases i demonstrate

either the khatru based server side is not sending an ok, or the go-nostr client side is not receiving the ok and the context deadline is expiring