208000+ entries in 0.076s
mircea_popescu: <asciilifeform> 'costco', for u.s. folks. sorta like a meatspace 'amazon prime' << used to shop at their equiv in ro. "selgros".
mircea_popescu: <mats> NATO would also presumably be happy to sell their enemies ammo for the right price << god knows romania has been doing it for a decade straight.
mircea_popescu: in this sense, the few captive "us trainers" they have work exactly like the mongols'
mircea_popescu: of their time (at their prime, that is), on their own. << in no way did they do this. they merely went around with a horde of gypsies, chinese, etc. heck, the siege of byzantium used a hungarian cannonmaster.
mircea_popescu: in other news : i am satisfied that the 363730 fork and the 363730-363736 dead chain contain nothing remarkable from a "hey guise they switched out the blokchanin" perspective. so i'm letting this rest.
mircea_popescu: cazalla ah the mystery unravels, they wanted the guy as ceo so as to buy some legitimacy for something stupid they want to do.
mircea_popescu: they've decided they wanna hang out with wintel and the rest of the "good guys in their own mind" club
mircea_popescu: and it was not even relayed (as it's not in fact relayable)
mircea_popescu: but the truth, in general, is that closed systems have a very short shelf life. you can get away with it for a matter of years, maybe, if you're very good at managing it. not decades.
mircea_popescu: had no particular expectation to see anything like this yet, either.
mircea_popescu: neway, needless to say, i'm actually impressed. for once world of shit averted.
mircea_popescu: i have nfi where i'd be had b-a never happened and foundation never happened and etc, trying to make sense of this situation
mircea_popescu: yeah you know what, this work actually made a difference.
mircea_popescu: ascii_field not fully understood yet, but it seems to exploit a buffer overflow in their own fucktarded code.
mircea_popescu: it didn't actually work, obviously, but i must say without all the work the foundation put in, it probably WOULD have.
☟︎ mircea_popescu: relatedly : i am reasonably satisfied the ~950 kb single tx a few days ago was trying its best to make pre 0.10 clients unable to verify the chain anymore.
mircea_popescu: by now a multitude of people have copies, of that chain.
mircea_popescu: gernika_ it's up and down as i have the 363736 issue examined.
mircea_popescu: well if you email it anyway... why bother with bcrypt ?
mircea_popescu: ben_vulpes everyone emails passwords in plaintext. EVERYONE
mircea_popescu: bitcoin is open for everyone who can keep the whole thing. the whole blockchain, the whole mempool, etc.
mircea_popescu: thestringpuller the notion that bitcoin is open "to everyone" does not include fromfuctorhead above.
mircea_popescu: ascii_field his point was "who expects". apparently, buyer.
mircea_popescu: but it is i think the first chance to actually improve on extant code by writing something since the start of this project. rather than just cutting warts off and sewing the crater together
mircea_popescu: i didn't mean that isolated. i just meant, "and it never amounted to anything" isolated
mircea_popescu: you may either present : a) the whole mempool (current best behaviour) or b) a selected fraction (hopefully future best behaviour)
mircea_popescu: because you present an arbitrary fraction of the current mempool, there is no incentive for anyone to peer you.
mircea_popescu: not worth doing the other one. you'll just end up isolated on the network.
mircea_popescu: ideally the -so much, +so much and min_node_value are to be set in config file by user.
mircea_popescu: ascii_field nothing wrong with keeping a finite ip db and flushing the lowest entries as needed, just like iptables works.
mircea_popescu: this way, our nonstandard behaviour can at least be excused, nodes will continue to peer us for economic reasons, and your dream fo someday someone mining off a pogo may come closer.
mircea_popescu: g accepted txn. then it refuses connectioon to shit nodes. permanently. it still reads from them, but sends nothing.
mircea_popescu: so, how about, something like this : 1. pogo acquires max memory to be used for tx pool ; 2. pogo accepts txn into this until full ; 3. pogo calculates per byte value for eacxh accepted txn ; 4. pogo accepts new txn only if their per-byte value constitutes an increase to its pool ; 5. pogo keeps a score on all nodes ir EVER SAW, by IP! as to what txn it sent : -so much for sending unacceptable txn, + so much for sendin
mircea_popescu: if we're going to do this, which is nonstandard, a) better make it good.
mircea_popescu: ascii_field i dun parse that, are you disagreeing or wut ?
mircea_popescu: if i recall this was discussed, i was spouting things like "sort the tx pool by value per byte, kick out bad txn and nodes that propose them, always offer a buttered selection to askers"
mircea_popescu: im not setting mental limits on the basis of implicit conventions.
mircea_popescu: "ideology" ? yeah, that's right : usg's fucking ideology.
mircea_popescu: "Please understand that you mentioned you accept "Bitcoin" as a method of payment. Some online payment companies are fairly new and inexperienced and they do not provide sufficient fraud protection to members. When we review payment methods to determine whether they are allowed on eBay, two key factors that we consider are "Security and Safety" and "Ease of use". Please do not offer this payment method."
mircea_popescu: and the infortmation of WHO READS IT is not usg's business.
mircea_popescu: Itile you mean maximum so far ? i dunno, something like a week on contentious bets iirc.