756800+ entries in 0.478s
mircea_popescu: people seriously
think Judy Collins was hawt at any point ?!
jcpham: fill all bitcoin bars with
tungsten immediately
mircea_popescu: keiser with his
tv hordes is still smarting and so on.
mircea_popescu: as
the sore holes of mtgox,
the foundation, etc etc scattered all over
the forum neatly attest.
mircea_popescu: remember,
this is bitcoin.
the powers
that are can
tear
the powers
that wannabe a new one quite easily
mircea_popescu: in other news,
tea party runs off with seat in always-democrat constituency
ThickAsThieves: "With millenniums of history behind it as a hedge against debasement,
the key
to gold’s success is
the stability and predictability of its demand."
ThickAsThieves: they spell out
their own doom without realizing it, as usual
mike_c: then
the eye socket
thing
mircea_popescu: then mp fucks
their skulls and pisses in
their eye sockets
mircea_popescu: ThickAsThieves starting
to get a little pannicky, are
they :D
mircea_popescu: KRS1
there's no such
thing as gay dogs. dogs, much like people, will fuck other male dogs if no female dogs are available.
KRS1: And some even cross
their legs.
KRS1: I knew it..there was like
this dog once who always walked kind of sideways...ever since I first saw it I kind of wondered.
ozbot: Can a dog be gay, lesbian or bisexual? New
test says yes - Chicago animal welfare | Examiner.com
benkay: trough of disillusionment?
trough of disillusionment.
ozbot: Dismayed by men, 47-year-old UK woman 'marries' pet dog : Europe, News - India
Today
benkay: "gay" is a human
thing, KRS1.
HeySteve: shouldn't have joined #dogecoin
then
jcpham: gay dogs porn was something i didn't entirely want
to see
today
KRS1: who knew
there were gay dogs..i kind of wondered since I once saw a boy dog
trying
to hump another boy dog.
KRS1: There ya go if you are looking for
the red rocket.
benkay: i guess it's hard
to see his penis in
that photo
artifexd: Also, for
the record, I
think
that increasing
the block size right now is a dumb idea, not necessary and counter-productive
to
the desired result.
artifexd: I have a prior engagement
that I have
to attend
to. I will
think on
this.
mircea_popescu: artifexd perhaps because it's
the only one
that satisfies some other criteria we don't know about.
mircea_popescu: (and for
the record, i'm not arguing
that
this would be a particular venue of attack, i am merely showing how my
thread length/surface metaphore is much better
than your lock/lockbox metaphor)
artifexd: Because you still have
to find
the birthday collision. Why would an attacker
try
to create a block 1 gigabyte long (which still has
to be valid) in order
to clash with a block 1 megabyte long?
artifexd: Your point is
technically
true, but practically worthless.
artifexd: However, your point
that limiting
the maximum length in order
to limit
the possible collisions is irrelevant as well because
there is already a specific place for you
to
twiddle
that will have
the exact result you would want.
The nonce/timestamp.
mircea_popescu: an infinite block size is worse
than
the "1st char of
the md5" hash scheme i just proposed.
mircea_popescu: my point is
that expanding
the block size is not unlike shortening
the hash length.
artifexd: If
the hash was one letter, sure. 256 bits? "A lot" is being conservative.
mircea_popescu: which is exactly why
the hash power and
the block size interact
to create
the chain cryptographical security
mircea_popescu: but an attacker with a lot of hash power could create
two such blocks and
then swap
them in and out
to create dead
txn
artifexd: You are essentially looking at an attack
that is by definition maximum difficulty.
artifexd: Because when
the hash of a block is "discovered"
the contents of
that block are spread around
the world. An attacker would have
to not only find a collection of valid
transactions and a sundry other pieces of data
that collide with
the exact same hash but also convince someone else
that
the replacement was
the actual block.
artifexd: I don't argue with what your saying. I would instead say
that
this particular manner of attack would be monumentally stupid in
the case of bitcoin.
mircea_popescu: in other words :
the larger
the block is,
the easier for a collision
to be found
mircea_popescu: but mircea_popescu is shorter
than
the original block.
mircea_popescu: should i claim
thje actual block really was "Jesus hates blacks &^ faggots" ?
mircea_popescu: now, inasmuch as your blockhash is 3, how do you know
the actual block (Women are not people.) is what hapepned
artifexd: If we assume
that md5 is a "good" hash function,
then how much data you put into it doesn't have any bearing on what you get out of it.
Thus
the first character remains, essentially, random regardless of whether
the input is 200 bytes or 200 megabytes.
mircea_popescu: we'll have
to work at
this. suppose your hash function is defined as
the first character of
the md5.
artifexd: Making
that reference harder
to forge is where
the security comes from. How much happens in between each reference point has no bearing on how vulnerable or reproducable each reference point is.
artifexd: Solidifying a group of
tx into a block doesn't provide security in
the same sense
that 100
twigs in a bundle are stronger
than 100
twigs seperately. It provide security in providing a stable
time reference for when
tx happened. Completely different domains.
ozbot: Bitcoin exchange for high-frequency
traders is launched -
The Wall Street Journal - MarketWatch
artifexd: This will
take a bit
to
type. Patience.
artifexd: I'm fairly well versed in block/transaction/hash/etc mechanics and I am
trying
to wrap your metaphore around
that structure. It doesn't make sense. I normally have a high amount of respect for what you say. In
this case
though, I
think you're wrong. Or, I don't understand
the variety and
thus need more lurking.
mircea_popescu: this is fundamentally what hashing does : it
turns
the problem of "cut
this
thread" into "cut
this surface"
assbot: [HAVELOCK] [B.MINE] [PAID] 2.44287593 BTC
to 3`707 shares, 65899 satoshi per share
mircea_popescu: artifexd your lock metaphore is broken.
the block size is
the length of a square strip of material
artifexd: mircea_popescu how does doubling
the block size quadruple
the work needed
to get equal security? If a certain lock on a certain box provides XX amount of security, how does increasing
the size of
the box make
the lock worth less?
mircea_popescu: let
the people raise
themselves up
to it, rather
than having it lowered
to
their present convenience.
mircea_popescu: generally speaking opening any one good
thing
to more people shittifies it.
Neil: Well, we'll get
to see!
Neil: Pools are afraid of orphans, and
that's
the way it should be.
mircea_popescu: but
the securing of
the chain is an expense external
to
them and easily externalisable
Neil: Not at all, or
they'd have been doing it for
the last year.
mircea_popescu: to wit, people have an incentive
to mine large blocks fopr
the direct payout
mircea_popescu: and no, it is not valid
to pretend "block size should be set by
the market", because of
the problem known as
the disaster of commons
mircea_popescu: the idea is for
this system
to be self sufficient, and in
this sense
there must always be a rarity of
txn slots in blocks with regard
to
txns
Neil: mircea_popescu: It'd be
the end of bitbet with fees
that high.
mircea_popescu: if
there are 10k slots made just as soon as
there's 10k
txns looking for a slot, you'll not see any fees.
mircea_popescu: if
there are 1k slots and 10k
txns vieing for a slot, you may see high fees
mircea_popescu: cryptography aside,
there is
the economic problem, :
the only way
to extract rents is rarity.
Neil: These
things should be chosen by
the market, just like fees. I doubt we'll get blocks much bigger
than 1MB for a long
time
to come, even if
there weren't a limit now.
mircea_popescu: a doubling of
the block size roughly quadruples
the hash needed for equal security.
mircea_popescu: currently blocks are secured
through
the process of mining.
the hash count per block is a rough proxy for its cryptographical security.
San1ty-Work: VanCleef: I heard some rumors
that
they are planning
to deploy 1600 PH
mircea_popescu: there are
two good reasons not
to increase it. one is
that
the larger
the block,
the more expensive its securing
Neil: Can't argue with
that.
mircea_popescu: well
then here's
the caper : it's beyond fucking stupid.
Neil: It's
the right
thing
to do.
mircea_popescu: Neil so you believe
the block size will increase. why ?
Neil: artifexd:
The occasional 900 KB blocks rarely go over 1200
txns
San1ty-Work: What's
the general consensus here on AM and
their Gen 3 chips?
artifexd: 1meg block allows for (ideally) 5000
transactions (at 200 bytes/tx).
To get 12.5 btc in fees, each
transaction will have
to pay .0025 btc in fees.
Them
thar are some hefty fees.
Neil: Well I believe
the block size will increase. But also with say ~1k
txns per 1MB block, I don't see people paying an average fee of 0.0013 BTC. Especially as I expect BTC
to be significantly higher
than
today.