951200+ entries in 0.704s
mircea_popescu: <Namworld> No amount of
testing would have found
the issue.
Namworld: I'm
talking about people making such claims... not about me making
the same claims =/
jborkl_: people were all up in arms about having 80
to 1000 unconfirmed
transactions waiting. with NO
testing, people made larger blocks -
that broke BDB
Namworld: THAT'S EXACTLY WHAT I'M FUCKING CLAIMING
THAT
THE ISSUE AIN'T ON .8!
mircea_popescu: the issue is with berkley db, which was on
the way out
smickles: personall, i include previous version interaction in
the
testing of one version
Namworld: This is
the only
thing I'm
trying
to say.
Namworld: Not
that more proper
testing would not have found
the issue (proper with many versions).
That more
testing of .8 alone would have not changed anything.
jborkl_: Yeah, me
too- I don't like being
the first
to change- bit me
too many
times
Namworld: Don't you get it? I'm
trying
to refute people who
think
the issue is with .8 and
that .8 alone should have been
tested more...
MJR_: Namworld: I
think what
they mean is
that part of
testing a new release is
testing it's interactions with other versions for backward compatiblity
mircea_popescu: jborkl_ i've always been slow
to upgrade, specifically for lack of faith.
smickles: Namworld: i believe
they do
test for interaction with older versions
mircea_popescu: nobody is saying "shouldhave been
tested more by itself in a vacuum"
Namworld: I know... but people everywhere say
that .8 should have been
tested more so
the issue would have been found.
jborkl_: really, only
test a part of
the code against itself?
mircea_popescu: unauthoritatively my it says berkley seems
to choke in
this case over about 10k records
Namworld: I'm not saying it couldn't have been found, I'm just refuting
that doing more
testing on .8 would have given any result.
smickles: he only
tested 500k blocks, and it happened on a 900k block
mircea_popescu: smickles gotta give props
to gavin, he's one of
the few
that are actually contrite enough.
smickles: Namworld: gavin said he would've caught it if he had
tested a larger block on
the
testnet
jborkl_: I ran .7 and .8 versions, If I had run
testnet and made a 1mb block I would have most likely broken .7
mircea_popescu: obviously not
the amount of
testing
they were willing
to do in between chat session and harassinfg
the userbase found it
Namworld: The issue was in .7, if
they
tested .8 forever,
they'd never have found
the issue if
they didn't use .7 in
the
testing.
smickles: something gavin was
testing, just not
toroughly enough
smickles: Namworld: depends how you look at it no?
the iss presented with
the interaction of .8 and .7
mircea_popescu: any statement
that claims "no amout of
testing" is a lie by its nature
Namworld: No amount of
testing would have found
the issue. .7 would have needed
to be
tested.
smickles: gavin had admitted
to
testing 500k blocks
mircea_popescu: in a sense
this backhandedly proves how strong
the idea is
Namworld: they can still mass object
to changes and refuse... but
the longest chain will be
the one with most miners.
mircea_popescu: jborkl_
the idea is strong,
the implementation horrible
smickles: jborkl_: I do wish
they had
tested .8 better.
tis could've been caught
smickles: Troic_: beauty of
the system,
they have a financial incentive
to cooperate
jborkl_: That
they encouraged people
to change
the block size and not knowing it was going
to break BDB was a big fuck up
Troic_: it could have been colossal, if
the pool ops weren't so cooperative
smickles: yeah, and it's not likely
that many lost btc due
to incidental double spends
Namworld: It wasn't
that colossal... just a little forking
jborkl_: It is holding up well considering what a collosal fuck up
this was
gribble: BTCUSD
ticker | Best bid: 43.41654, Best ask: 43.41666, Bid-ask spread: 0.00012, Last
trade: 43.50000, 24 hour volume: 166958.61630911, 24 hour low: 36.65000, 24 hour high: 48.46900, 24 hour vwap: 43.77197
mircea_popescu: "It's about as bad as everywhere else on
the planet, I guess, with
the single exception of Romania," Asia Pacific Network Information Centre's chief scientist, Geoff Huston, said in an interview with ZDNet Australia.
mircea_popescu: In Australia, IPv6 adoption is almost non-existent, reports Josh
Taylor.
gribble: BTCUSD
ticker | Best bid: 41.51000, Best ask: 41.99900, Bid-ask spread: 0.48900, Last
trade: 41.99900, 24 hour volume: 146092.87031245, 24 hour low: 36.65000, 24 hour high: 48.46900, 24 hour vwap: 44.35230
kakobrekla: unaffection comes
to bitbet so naturally i wont be posting any signes.
Bowjob: well, i watched it from 45, drop
to 42.5 .. i see each cent dropping
gribble: BTCUSD
ticker | Best bid: 42.50000, Best ask: 43.00000, Bid-ask spread: 0.50000, Last
trade: 43.00000, 24 hour volume: 143829.74796654, 24 hour low: 36.65000, 24 hour high: 48.46900, 24 hour vwap: 44.38319
Ukto: looks like
trading on btct is back
midnightmagic: More likely it was a side-business for coin laundry, but none of
the miners could ever (far as I know) figure out where
the coins were coming *from*.
midnightmagic: nah it's not. It's an example of hashrate for sale, and successfully for sale for
that matter.
mircea_popescu: i don't
think so. gpumax dissapeared with pirate, as
the sane people were saying back in february 2012.
midnightmagic: I
think gpumax is
the counterexample
to
that assertion (that
they'd make more by mining)
mircea_popescu: is
the main strategic purpose of having a devteam in
the first place.
mircea_popescu: keeping money ideologically neutral, implementing
the old pecunia non olet
mircea_popescu: the same applies ideologically
too.
the network won't get attacked for as long as it manages
to not stupidly offend.
mircea_popescu: just leave it stand as it is.
the
theoretical response
to a 51% attack is
that
the attacker would make more by mining.
midnightmagic: Well if
that's not what you meant, perhaps you should define
things more narrowly. :)
mircea_popescu: midnightmagic
that's fine, but you pervert
the choices when you add riders such as "one is really vanishingly small"
midnightmagic: .. but you defined
the
two choices earlier. I'm not agreeing
that
the choices match reality. I'm just hanging out in your hypothetical land because you invited me.
mircea_popescu: sure, irrelevant ventures aren't part of
the discussion, if bitcoin stays alpacca we're wasting our
time with
these concerns/
midnightmagic: Even btcexpress preferred
to leave his attacks on
the altcoins.
mircea_popescu: that;s brancing
the discussion off
the germane path
tho
midnightmagic: The gambling servers at least are very popular
targets. Extorting
them is almost a Russian pastime.
midnightmagic: mircea_popescu: Well, not in my personal experience. But perhaps
that's out-of-date
these days.
mircea_popescu: pigeons if we shouldn't,
there should not be call for doing it.
mircea_popescu: midnightmagic not so. you will be surprised
to find attackers routinely forego hitting higher payoff cooler
targets
pigeons: ok maybe we should just encode files and upload
them
to
the blockchain ala namecoin
too
midnightmagic: Also, if
the money is on
the side of
the all-use-cases-are-valid (i.e. gambling and drugs, hypothetically)
then obviously
the money is where
the attacker goes.
mircea_popescu: same
thing pigeons. as long as you're making
that call your ass is
too
tight.
midnightmagic: I'd preferentially attack
the ones who appeared
to be more philosophically aligned with
the notion of attack.
pigeons: by
the way people dont
think using bitcoin for gambling is spam,
they
think using bitcoin
transactions
to signal a lost bet is spam
mircea_popescu: suppose
there is one attacker such as you've modeled him.
mircea_popescu: suppose buttcoin has some devs
that
think all usecases are equally valid, and dorkcoin has some devs who believe using
the coin for gambling is spam.
midnightmagic: Yeah but what has
that
to do with anything? Is
this your way of saying you're done arguing?
mircea_popescu: looser in
the ass ? it means being used with being fucked in
that orrifice.
mircea_popescu: if
the devs were a little looser in
the ass, would
the sort of econ savvy attackers stay
that side or join
this side ?
mircea_popescu: and you must never presume political acumen on
the part of anyone.
mircea_popescu: but selecting
the coin
to attack is not a
technical problem.
midnightmagic: you must presume
technical acumen on
the part of
the attacker.
BingoBongo: midnightmagic Win32 killed everything on Windows... Crosschatter happened before, some altcoin late last year did it (no halving fork) and had
to be euthanized.
mircea_popescu: midnightmagic no. because attacker wouldn't know which
to attack.