290600+ entries in 0.19s
a111: Logged on 2016-06-21 15:24 mircea_popescu:
they have pages worth 20
to 100 mb worth of shitty gifs, but can't display 10mb of a god damned s expression.
Framedragger: heh yeah, if no POST involved in phpmyadmin reply, and
there shouldn't. lol.
mircea_popescu: Framedragger notice
the interlocking magic numbers of shit!
mircea_popescu: Framedragger
that may explain
the part where
the first part fails, i guess. but not
the 2nd part. both php items (wp, phpmyadmin) should be able
to list a 10mb sql record.
Framedragger: it seems
that you can modify it in .htaccess with php_value post_max_size 20M
Framedragger: mircea_popescu: post_max_size in php.ini, apparently set
to 2M by default?
mircea_popescu: but yes i'm sure
there's "a limit" put in
there by people who don't belong
touching computers.
Framedragger: mircea_popescu: couuuld be max file upload limit set in php.ini or in webserver? but i guess could also just be wordpress /
typical php app shit..
a111: Logged on 2016-06-21 05:13 PHF: i wonder if
there are hidden implications
to making all nicks case insensitive for /log/
mircea_popescu: they have pages worth 20
to 100 mb worth of shitty gifs, but can't display 10mb of a god damned s expression.
☟︎ mircea_popescu: if
the item in question is uploaded via ssh-mysql directly, phpmyadmin (common mysql interface) dies when
trying
to say "edit"
the column, and wordpress itself fails
to pipe
the content
to apache.
☟︎ mircea_popescu: and in other lulz : wordpress interface dies if one
tries
to publish a 10mb
text article with "unknown error". apparently
this has never happened before.
☟︎ tfoxbrewster: interesting - i have
to head off, but I'ld like
to continue
this conversation
tomorrow if possible?
mircea_popescu: incidentally
tfoxbrewster : go $register <your full fingerprint>,
then
the bot will register you and i'll rate you and
then you can self-voice.
iKant: :) more on
the line of
the German Philosopher with a nod
to
the vogue of
the single letter prefix
mircea_popescu: the signature here however is prospective. "this is a signature you don't know what
to do with. at some later point, maybe you will."
mircea_popescu: in other words : signatures
there were retrospective, attempting
to build a historical chain. "this x is
that x which you respect so respect
this".
mircea_popescu: here,
there's no historical establishment.
the guy lives
through his deeds not
through a lengthy, boring story about "hey listen, back in 1920 carnegie and j p morgan were pretty cool so we should still be getting free petrocheese and chinese plastics for
this reason".
mircea_popescu: the difference between
those
two and
this
third is instructive and important : in
those instances,
the usg was
trying
to spin a narrative
that'd allow it
to claim it's still alive and relevant, in front of proof
to
the contrary from ~already established accounts~. ie, satoshi.
mircea_popescu: then
there was
the whole craig whatever incident. where
the usg assets in bitcoin (gavin & all) swore as
to
the authenticity of a gross fake.
mircea_popescu: now,
the usg assets in bitcoin (gavin & all) didn't like
this, and whined about
the signature. so i said : swear it ain't so. in writing.
they... didn't.
mircea_popescu: because
this isn't a single item, it's a
third in a
trifecta.
mircea_popescu: tfoxbrewster well, you familiar with
the satoshi/hoaxtoshi communications
this year ?
tfoxbrewster: For
that reason, I'm certain
that
the message is a fraud, written by someone who wants
to stir up
trouble in
the community."
tfoxbrewster: If
the attacker wrote
the message,
there would be no reason for him not
to provide a valid signature. If
the message were written by an impostor, on
the other hand,
they'd be unable
to generate a valid signature, and would have reason
to make
their message more convincing by adding a fraudulent one.
tfoxbrewster: Their comment: "The message claims
to contain a cryptographic signature in
the last line. ECDSA signatures of
the sort used by Ethereum are 65 bytes (130 hexadecimal digits) long, and end with '00' or '01'.
The number at
the end of his message is
the right length, but ends with '20', making it an invalid signature.
mircea_popescu: phf feature request : can log search have a parameter indicating how many later references are needed for line
to qualify ?
☟︎ mircea_popescu: (seems
to me unavoidable, 1bn (atomic)
txn ~= 100gb, yet 1bn people
that know half a dunbar's number worth of others ~= 100 Gb * however much each record
takes.)
a111: Logged on 2016-06-21 05:30 phf: ok
this
time around it
took 9.1s, 5s of cpu
time (so
that was correct). but entire wot
takes 4.1mb (i forgot
to gc at some point)
mircea_popescu: honestly, i'd be in favour of "just replace all 1 with i and all I with l and
take out all non-alphabetic characters". MIRCEA MLRCEA M1RCEA MI_R-CEA > MIRCEA and fuck
them.
a111: Logged on 2016-06-21 05:17
trinque:
thing is even if I change
the schema I'll probably replace rating with a view of same structure
mircea_popescu:
http://btcbase.org/log/2016-06-21#1486149 << honestly,
there is ABSOLUTELY no reason for rating system
to distinguish MP from mP.
there is no legitimate reason for
two different individuals
to be
this close ; nor is such nonsense supported. in fact, your scheme should not even permit m1rcea and mircea
to be distinct.
☝︎ mircea_popescu: ;;later
tell
tfoxbrewster certainly not at seven in
the morning.
a111: Logged on 2016-06-21 01:48 mod6: Framedragger: hey,
thanks for all your work with
the SSH keys.
a111: Logged on 2016-06-20 20:39 mircea_popescu: anyway, normally i'd have just let you do
the conversion locally, but since Framedragger is such an impetuous youth, we're moving fast for once :D
a111: Logged on 2016-06-20 18:52 phf: you might not notice it, because you look at it from
the perspective of release summaries and hacker news posts, but
the kind of decisions
that go into llvm are ~entirely~ driven by large corporate with large
teams and large hardware needs. it's synergistic.
a111: Logged on 2016-06-20 18:49 phf: Framedragger: despite
the rhetoric people here are extremely pragmatic. but when you
think about
these
things
there's no reason
to allow sloppy
thinking.
the question is "how much of a dick up your ass is
too much dick", and
the healthy answer is "none at all". llvm is good
technology (it gives you same kind of layering
that a decent common lisp compiler does, but for languages
that are not used
to
that kind of richnes
gribble: Bitfinex BTCUSD
ticker | Best bid: 706.0, Best ask: 706.04, Bid-ask spread: 0.04000, Last
trade: 706.04, 24 hour volume: 52974.9952276, 24 hour low: 675.5, 24 hour high: 767.77, 24 hour vwap: None
phf: ok
this
time around it
took 9.1s, 5s of cpu
time (so
that was correct). but entire wot
takes 4.1mb (i forgot
to gc at some point)
☟︎ phf: actually i don't
think
that's right
phf: entire wot
takes 9.9mb in memory, full refresh
takes 15s with an ssh hop, 5s of which is cpu
time
trinque: the wot db I got was already all lower iirc, so I have no idea what
the capitalization of nicks was originally
trinque: that way you don't have
to keep
track of my schema changes
trinque: thing is even if I change
the schema I'll probably replace rating with a view of same structure
☟︎ trinque: at which point fixing
the caps didn't matter
trinque: then realized I had
to go back
through and add lower()
to
the queries
trinque: I
think
TomServo's
the only one like
that; I could just fix it
phf: i wonder if
there are hidden implications
to making all nicks case insensitive for /log/
☟︎ phf: TomServo in person, but
tomservo in
to_nick ...
phf: trinque: perhaps having from and
to be fk with constraints is also a good idea
thestringpuller: trinque: is
there no read-only api for deedbot like
there was assbot?
trinque: hell,
that sig sauer of recent fame seems like a great gun
trinque: and
the other change
that'll affect you is I'm moving it
to another box, will let you know of
that
too
trinque: soon as I get
to it, rating's going
to get switched
to foreign keys; I'll warn you first
phf: trinque: so i
take it person/rating is what i'm looking at?
thestringpuller: when is
the dao attacker gonna cause more drama.
today has been boooring.
mod6: Framedragger: hey,
thanks for all your work with
the SSH keys.
☟︎ mod6: this eth implosion has kept
the lulztrain running
mircea_popescu: "Solidity" is one of
the most aptly named piles of rubbish for purposes of unintentional irony self-crits
mod6: dear load
that chinese food was rancid
mircea_popescu: maybe
the next sandback da could be named lish somethingortheotherheyinindiaournamesneverend
mircea_popescu: mostly quoted for
the epic lulz of "bufete de abogados". which yes, it's how
the spanishtards say law firm.
mircea_popescu: (undersigned john hat, workman without a dick : why should i pay
the
tax when i can't fuck ? warmly esteemed mr john hat, workman without a
tool : while you've got your fingers &
tongue
the law stands.)
mircea_popescu: "mult stimate ion caciula, om al muncii fara scula : cit ai degete si limba, legea
tarii nu se schimba."
☟︎☟︎ mircea_popescu: "subsemnatul ion caciula, om al muncii fara pula : de ce sa platesc
tribut, daca eu nu pot sa fut ?"