log☇︎
950900+ entries in 0.614s
mircea_popescu: past this point.
mircea_popescu: and ideally remove the current devteam's license to release new clients.
mod6: it shouldn't be up to one dude to pay for all this shit...
mircea_popescu: jurov atm the plan ios to push for spec
jurov: what do you plan to do, then?
jurov: mircea_popescu one thing i don't understand, you aren't going to hire any devs yourself cause it isn't worth the conflict and drama?
mod6: i hate that site
mod6: awe fuck it, i guess that's long ago anyway.
mod6: i gotta look through btctalk for that *sigh*
mod6: like on upgrade from version 4 to 5 or something? there was a night where they did an upgrade and the similar thing happened.
mod6: didn't this happen or something very similar to it before?
mod6: so can we rewind the tape just a bit...
Lyspooner: The Church of Bitcoin and the Vatican Bitcoin Foundation have met Luther Popescu
mircea_popescu: so much for testing
mircea_popescu: jgarzik sipa: I don't think so
mircea_popescu: sipa jgarzik: have we seen a block which affected 5000 transaction index entries?
mircea_popescu: i mentioned something about how they;'d better
benkay: would you point me at that?
benkay: you mentioned something about the dev team starting work on a spec?
benkay: especially to those of us accustomed to flushing idiots out of systems with economics.
benkay: it's a transient state. frustrating to be sure.
Lyspooner: what about those of us who have been running 0.9
mircea_popescu: this sort of shit belongs in onboard code for vibrating dildoes at the most.
mircea_popescu: and more importantly bitcoind doesnt die with "failed to create transaction"
mircea_popescu: nobody gets to say i can't,
mircea_popescu: if the block can take it i can make it.
mircea_popescu: i want to make a 750kb tx.
benkay: mircea_popescu: what do you mean by arbitrarily-sized transactions?
mod6: i still <3 you guys tho :)
mod6: ok i gotta walk away for a bit and think about this some more.
mod6: at least the SSL code they copied looks to be better: BN_CTX *ctx = NULL;
mod6: i just hate looking at the client code. maybe its just their style.
mod6: for starters who does this? Dbc* pcursor = db.GetCursor();
mircea_popescu: due to overwhelming miner vote.
mircea_popescu: that thing stil around ?!
jurov: lmao DUE TO OVERWHELMING REQUESTS we had to remove PayPal & Local Pickup options.
jurov: let's get to budapest for buncha asics and outmine them http://primeasic.com/order.php
mircea_popescu: i am still steaming over the fact that i can't create abitrary size txs
jurov: incidentally, they still don't have sane interface, too
jurov: well, they preferred to entagle themselves in multisig txs at the time
mod6: they should back away after this
mod6: thats obvious to everyone in the world who pays attention
jurov: prolly they don't have time to test properly xD
mod6: yeah, its WAY too large imho
jurov: wasn't hard... couldn't understand why that function isn't exposed already
jurov: like my two independent wallets prototype: https://github.com/jurov/bitcoin
jurov: mod6 i'm not interested in arguing. if i had time, i'll do a fork with some very useful function and wait for them to come
jcpham: i'm doing much better in the poll today
MJR_: ain't no one got time for that indeed
kakobrekla: oh i read that "in summary the coinbase is defenetly useful" and got a heart attack almost
mircea_popescu: a summary of the codebase is definitely useful.
mircea_popescu: mod6 i don't know who's listening to them.
mod6: aint noone got time for that
mod6: i'd love to take the time to audit that shit, but who wants to fight with LukeJr and the others?
jurov: mod6 good for ya. i did touch it and it didn't hurt.
mircea_popescu: MJR_ the bitotter thing he's working on
mod6: so i'll just my retard ass out of there
mod6: i never wanted to touch bitcoin's client or anything else because the #bitcoin-dev is filled with pompus retards
mircea_popescu: that;s usually a good step towards specification.
mircea_popescu: mod6 actually once you're done with the iphone app, a very good use of someone's time is to go through the codebase and summarize it
mod6: slice and dice shit out of there that needs to be replaced.
jurov: mod6, maximal packet size (or MTU) is hard fact baked into ethernet from the beginning, not some broker earlier implementation
mod6: i dunno, i guess i'd have to spend some time in there. and the client needs to be combed through
pigeons: yes perhaps, but currency liquidity flowed throug it before and will continue too,
mircea_popescu: once they get their altar broken for reals.
mircea_popescu: pigeons on superficial analysis ripple seems just the place where the curerent bitcoindev rats will sucrry to
mod6: im simply saying, look at how this problem was solved
mod6: im not saying about the transport, or anything
pigeons: mircea_popescu: cause people are transitioning from doing it via the private standalone django app to the new ripple but it is being done now
mod6: but when the size of UDP gets over 65,535 bytes (need to double check), it simply converts it to TCP and sends it that way. something can be learned from this, nothing more.
mircea_popescu: look what happened in ny when the clearing corp got flooded
mircea_popescu: MJR_ except on the day it gets attacked.
MJR_: I think a central clearing corp, that only needs to transfer the diff between exchanges will be the best way
mircea_popescu: mod6 of course, the blockszie issue wasn't hitting the transport layer
mod6: jurov: we can learn a thing or two, perhaps, about how DNS solves the issue. UDP is meant to be fast, and arbitrarily large since forward/reverse zones can contain an large amount of entries.
mircea_popescu: MJR_ something like that.
MJR_: I can't remember your exact quote mircea_popescu but something like "if you want to use crypto currency but don't understand cryptography..."
pigeons: really? its second to -otc in stupid
MJR_: when you think that people on coinbase are probably still in the innovators (before 18%) stage...wow
MJR_: I think bitcoin-assets is the bleeding edge of the bleeding edge
mircea_popescu: MJR_ im not even sure at this point.
MJR_: I think it's multicast tcp?
jcpham: unicorns are definitely walking in hamster wheels though
mircea_popescu: mod6 originally i thought it was udp p2p
MJR_: I wrote a program to monitor how often we need to fail over to TCp for our data feeds
jcpham: you access mpex through open standards
MJR_: know exactly what you are talking about
mircea_popescu: versus the keep the specification equal to the unreadable codebase.
mircea_popescu: keep the specification open and the implementation proprietary
mod6: any thoughts on that? anyone know what I'm referring to?
mod6: and really... the UDP (TCP when needed) is a model that the guys should be looking at to solve this blocksize issue
MJR_: but also...I really think node and branch will be needed
mircea_popescu: people would just mindlessly replicate it (see bitfinex) then we'll be well fucked
mircea_popescu: mod6 one of the key reasons why i keep the code closed.
mod6: similar to rootservers
mod6: not the SAME code, but similar to avoid attacks upon a weakness found in one
MJR_: I really think a gpg based order system would be nice
mod6: if they were modeled after mpex, that'd be a start
MJR_: but even better opportunity for someone to route flow between exchanges
mod6: all independant of eachother that aren't scams