909600+ entries in 0.68s
pgp: yeah - CTFC guy is kinda ridiculous -
the whole discussion was ridiculous...
dub: you
think
thats air you are breathing?
jurov: s.mpoe about
to pop
the lid?
jurov: ah it came already..my scumbag bank added it sometime after 16:00, i need
to pay sth quickly and
tomorrow is holiday
jurov: but you seem
to use
them anyway...
jurov: seems i§ll move my ops
to bitcoin.de (although
that's different can of worms)
jurov: plus another 2-3 days
to confirm
the withrawal
jurov: sepa seems
to
take longer
than usual
jurov: since
they changed account
mircea_popescu: m_p!~AndChat13@adsl-068-209-251-085.sip.asm.bellsouth.net failed
to login
to mircea_popescu.
MJR_: then i'll blog later
tonight, after i have a few drinks in me
MJR_: working on
the web app now
mod6: hey MJR_, sounds like
there was quite
the
turn out at
the open-cry market yesterday
jcpham: i want
to see 339 colo's mining rigs
benkay: good for
them, denominating in bitcoin
jurov: dunno when
that cesium clock goe
through mpex clearance
jurov: ^^me adding another mpex quirk
to code
jurov: log.error("Matched
trade date:%s earlier
than order create date:%s",
t['Date'], o.dateCreated)
kakobrekla: yeah,
that would be worth
the whatever amout of dessertspoons of btc he is asking for it
jurov: you can propose
to have webcams in sauna
then
kakobrekla: so if i
take
the virtual delegate
ticket
joecool: lulz just had
the office bugging out cause i had a VPN setup on one of
the main systems called "Romanian Wizards"
jurov: kako, what's
this
time?
kakobrekla: bitstamp is way more lame
than i
thought it was
jurov: (from sauna supernode summit
thread)
jurov: That, coupled with
the fact
that I am young and willing
to
take my life in a new direction, have pushed
towards my presence at
the Summit.
jurov: but dunno where
this came from: "The value of
the bitcoin is pegged
to AZNs"
benkay: take
that next crap back
to analysis,
tomservo
gribble: There are currently 5.0475987e+09 bitcoins demanded at or over 0.0 USD, worth 16649216.8979 USD in
total. | Data vintage: 117.2777 seconds
gribble: There are currently 5.0476089e+09 bitcoins demanded at or over 0.0 USD, worth 17084843.7602 USD in
total. | Data vintage: 0.0289 seconds
Scrat: TAT.ASICMINER
tanking
Framedragger: parseval: sure, just note
that it's usually
trivial (as in, perhaps only change
the db configuration string in your app's config)
to change
the underlying DB (installing and starting a DB server with default config is also usually really simple) - but yeah, would be interesting
to see your results eventually, good luck!
parseval: Framedragger: I'm just using
this as an intermediary store on my local box for some
testing so I don't have
to
throw in a full fledged database.. It's a simple delete and rerun until everything is nailed down with
the import. Eventually I'll move out
to mysql or postgres for
the actual application
gribble: BTCUSD
ticker | Best bid: 104.30002, Best ask: 104.90479, Bid-ask spread: 0.60477, Last
trade: 104.90479, 24 hour volume: 152353.43826341, 24 hour low: 102.00000, 24 hour high: 124.66300, 24 hour vwap: 112.03069
gribble: Next difficulty estimate | 10666629.6451 based on data since last change | 11536948.7666 based on data for last
three days
Framedragger: heheh, well if we are also worried about DB _querying_
though..
mircea_popescu: Framedragger i
think
the cause is
this : db ops may
take > 1 minute.
Framedragger: yeah, db design can
turn into a fetish, unless it
turns out significant optimizations are relevant
to business strategy and whatnot
Scrat: premature optimization
to
the point of insanity
mircea_popescu: db design might be
the most "overengineered as a hobby"
thing in it.
Scrat: unless you want
to do multimaster. heh
Scrat: can't go wrong with postgres
though
Scrat: not saying
that it doesn't work. it's just not optimal
davout: if you have
the disk space i don't see how a well indexed sql storage wouldn't be adapted
Scrat: davout: well, 100+ gigs of randomly accessible
time series
Scrat: there are DBs made specifically for
that.
there are also generics like Cassandra or mongo
that excel at it
Scrat: but you wouldn't use SQL for storing BIG
time series
Scrat: sqlite will do 100k inserts per second on crapy hardware if you group
them
Framedragger: parseval: yeah i don't
think sqlite is a good solution in
terms of scaling. if you're using ORM abstraction, changing
the underlying DB shouldn't be a prob, unless you're using some DB-specific functions. postgres is really nice for Shitloads of Data
type of situations
Framedragger: eh, i've
to import ~400gb of data into postgres, i do ORM commit every 100k inserts, it's actually quite fast :) (sqlalchemy orm abstraction). not really a problem
Scrat: I guess it's ok if you want
to get railed in
the ass
davout: (let's get
this started xD)
parseval: It'll make querying
the data a lot easier in
the end
though
parseval: Well, I'm
throwing it all into Django ORM, so
that probably adds a bit of overhead
Scrat: parseval: you have
to group chunks of 10k inserts into
transactions
parseval: I'm doing
the initial imports, it's about 1gb of json data up
through
the end of april
KRS1: I
think coke was involved somehow
KRS1: I have a pool
the best I ever did was have one stripper naked in it, nothing like
this.
Scrat: parseval: how big is
the data? I've found sqlite
to be very fast for running queries on
time series
KRS1: yeah
that one girl with
the small boobies is a must have
mircea_popescu: parseval honestly 1 day intervals would be a great
thing
to have.
KRS1: Are
these girls your friends?
parseval: sqlite seems a bit slow.. I'm going
to reprocess
the data into 5 minute intervals with
the average price and volume for
that interval
KRS1: I have needs
too..I want
to see pic 42 at least fullscreen
KRS1: And goddamnit every picture I click on
takes me
to an article with no picture..how am I supposed
to fap
to
thi
mircea_popescu: parseval actually
this sounds interesting. ding me when you have
the graphs ?
KRS1: Why do
these naked pool parties always happen in some far off country in Europe or something.
Scrat: why not? sqlite isn't
that bad for
time series
parseval: I want
to graph
the USD value for various securities over
time
parseval: I probably shouldn't have used sqlite
to store all of gox's historical
trade data..
mircea_popescu: no, i should add 0.1% of all deposits >
trilema credits
davout: mircea_popescu: you should add a BUYCREDITS command
to mpex
gribble: gigavps was last seen in #bitcoin-assets 9 weeks, 4 days, 14 hours, 49 minutes, and 24 seconds ago: <gigavps> well, we can rule out c, i have block chain
transaction
to prove what i have paid
Scrat: i cant fap
to
this
davout: well, i guess
that, it is suit
mircea_popescu: kakobrekla you know
that's just some random waiter right ?