165700+ entries in 0.098s
mike_c: hm, what happened in
the last couple days? difficulty adjustment?
mike_c: and it will just spit out
the hand numbers
Birdman: and i
type just
that string after
the rest at
the same
time? "whatever.py C:/handhistories python hands2.py c:\directory > logfile.txt" ?
mike_c: well, if
that's all you want just:
Birdman: i meant
to have it save results
to a
text file, how do i do
that>
mike_c: you just want it
to spit out
the hand numbers? sure.
Birdman: rather, save it
to a
text file
Birdman: any way for me
to run
this and have it spit out hand numbers it finds?
Birdman: well
thank you
this is crazy helpful!
Birdman: then may as well keep
them
mike_c: still 4,393 bad hands. down from 5,013. so doesn't cut out
too many
Birdman: ok so i suppose we could get rid of any
that have
the value "seat256" in
the summary and any
that are missing
the summary in general
mike_c: well, seems like 80% of
those are only going
to be missing one player, so filtering by
that won't help you
Birdman: which i did by searching
through with "T1" cause
tournaments all start with
that and no cash games would
Birdman: 262953 / 4914 / 3966 respectively, after i deleted all
tournaments from
the histories
mike_c: run
the updated version I gave you and
tell me all 3 numbers (processed, bad, files)
mike_c: 4,412 of
the 5,013 bad hands are only missing one player
Birdman: but
to be honest its probably good
to have
the corrupted hands
to be able
to request
them from
the server
Birdman: yeah something like
that,
trying
to get it
to filter out
the corrupted hands from
the hands with
the bug im looking for
mike_c: um, not following. but you're saying you want
to see
the ones
that are only missing one player?
Birdman: not sequential, i mean
the ones
that repeat
the hand number
Birdman: it seems
to me
that
the bug im looking for is most likely
to appear in hands processed
that aren't sequential as
the sequential ones are most likely entirely fragmented and
the ones with only one player missing are
the ones im most likely looking for
mircea_popescu: well depends observably for whom. what, you're going
to compare events with alt-events-as-reasonably-imagined ?
mircea_popescu: i
think
this proves roosevelt is alive (might as well).
Birdman: oh also what would i
type
to have it dump
the output into some
text file?
mircea_popescu: asciilifeform aha, old man went in on a
ticket of no-compromise, compromised errythang.
Birdman: alot of
these hands are missing
the summary in general
Birdman: bit over 26k and before a bit under 30k, also i
think
this guy's hand history is amazingly fragmented in a
ton of spots so im
trying
to get a more solid chunk of data
to run it on
mike_c: and how many did it say before you pulled
the
tournament files?
Birdman: and if im not mistaken it lists
the player who is missing in
the summary and will list
the same hand
twice
to show if another player was missing in same hand so
that extra fat
that could be
trimmed but im not asking you
to fine
tune
this or anything lol,
thoughts on it maybe not showing all results for lack of space of something?
Birdman: so i went
through and deleted all
the
tournaments cause
those are not relevant, and i got seemingly different results. I
think maybe its displaying less
than all
the hands it has found? i only saw four cash game hands
the first
time, deleted all
the
tournaments, and now i see a
ton of cash game hands
Birdman: mike_c so it went
through all
these hands and says it processed ~30k of
them, but what am i actually looking at here?
the hands it lists in
the prompt are
the ones
that
the discrepency is found?
mircea_popescu: the notion
that spain has or could ever fucking recover...
mike_c: ok. it's a syntax error because
the script was written for python 2, not 3. easy
to change
though, hang on.
Birdman: python 3.6 (32 bit) and if i run
that it brings me
to a command prompt pretty much
mod6: i aspire
to get
that periodical
mircea_popescu: just like if on an imaginary island you had a farm
that produced all
the milk but none of
the meat,
the return would be mostly in milk and not much in meat.
mircea_popescu: seems
to be obvious and necessarily
true, iunasmuch as bitcoin has 100% of
the world savings market share and about a millionth of a percent of
the world capital allocation market share.
mod6: but mostly,
the savings rate far exceeds
the yield of capatal allocation; which is obvious. and will stay
that way for sometime as fiats -> 0. someday,
that'll be different.
mod6: <+asciilifeform> in
this light, doing anything whatsoever other
than 'hodling' is raw deal. <<
there are exceptions
mircea_popescu: anyway,
transparend qpq seeing how microsoft is losing
the global battle for relevancy.
mod6: <+mircea_popescu> let
them learn
teh brass
tacks. << everyone has
to, ya.
mircea_popescu: asciilifeform well no,
tank batallion is a different story.
this is a
textile maker, it has no further productive externalities
than what it gets for cloth
mircea_popescu: the important parts : spent 200 per share
to go from 60
to... 68.
mircea_popescu: whjole fucking point of investing, at least in
the buffett understanding of it (apolitical, or in his case naively copacetic-political, columbia floating over farmer drawn hills) is
to identify
the corps with
the most favourable capital discounts and
to buy
those while selling
their opposites.
mircea_popescu: the only reason everyday joe doesn't encounter capital discounts is because a)
too small and b)
too stupid. otherwise even someone running a small video rental comprehends
the difference between accounts receivable depending on from whom, and so on.
mircea_popescu: or in
the immortal words of some schmuck, "nah. if i give you 50k now i'm wasting it".
mircea_popescu: FOR
THE REASON
that
the implicit discount on its money is different.
mircea_popescu: this is because
the "same" money is discounted on
the basis of "in whose hands it finds itself" --
the 1 people are expected
to lose most of it under couch cushions ;
the 2 people are expected
to
take over
the world with
that scrawny one billion.
mircea_popescu: whereas in case 2 it will be put at so massively a premium so as
to make 2 stock outstanding worth more
than 1's.
mircea_popescu: both
these items made
the same unexpected billion. nevertheless, in
the case of 1
this will be discounted
to about a
third of its value, resulting in minor share appreciation c.p.
mircea_popescu: consider : 1. large corporation, worth 1
trillion. analyst expected profit year end is 75 billion. company beats, makes 76 billion.
mircea_popescu: goldbug looks at bar of gold,
thinks "value here", you look at benjie
think "value here".
this is however not so, capital is an abstract not a concrete item.
mircea_popescu: asciilifeform your notion of money is broken in
that you misunderstand its materiality.
☟︎ mircea_popescu: why shouldn't whatever spare cycles his brain has be used
to improve
tmsr-rsa ?
mircea_popescu: yes
they buy and make some
things.
this is of
the nature and substance of clouds.
mircea_popescu: anyway, rolling back up
this many branched loop :
there is nothing
to be gained by mentally denominating capital in uniform standard dosidoes, nor by mentally denominating labour in redditard.
mircea_popescu: no but we're all under a discreet obligation
to promote
the herdemocracy! everything else is
theft and corruption and racism and so on!
mircea_popescu: fortunately,
there's no kako-nubbins party
to argue
that hey, as btc price is chiefly supported by
the republic, which has a lot
to do with mp, it is
therefore mp's job
to idemnify "the users" as imagined by kako-nubbins for "this loss" resulting from his you know, unauthorized use of resources for purposes not priorly reported
to
the almightyly important kako-nubbins party!
mircea_popescu: yes. but
that's not under dispute,
the "convince
them" part is.
they were convinced, and it didn't even
take
that much in
terms of machine gun chains.