devyn changed the topic of #elliottcable to: 22:53:14 <+whitequark> also there was a fragment about Swiss embassy being located on top of a 1000-ft pole, inside which there was a gigantic arms exhibit
<devyn> yorickpeterse: hahaha
<devyn> yorickpeterse: (I promise it's not me)
yorick has quit [Remote host closed the connection]
<devyn> HOLYSHIT
<joelteon> what's that
<devyn> Firefly in Starbound
<devyn> joelteon: hahaha
<devyn> 5.times.inject("") { |s| s << rand(0x500) + 0x50 }
<devyn> whee
<joelteon> alexgordon: what would you use in C++ for a multidimensional array?
<joelteon> a vector of vectors?
<alexgordon> nooooo
<joelteon> that's what i thought
<alexgordon> good question though
<alexgordon> joelteon: depends on the usecase
<alexgordon> usually I build one myself
<joelteon> ok
<joelteon> i see
<alexgordon> it's not difficult
<joelteon> because then you do destructors
<alexgordon> in fact you can just use one std::vector and just do the index math yourself
<joelteon> yep
<alexgordon> remember to use operator() and not operator[] for indexing
<alexgordon> because it can take two arguments
<alexgordon> joelteon: what is this multidimensional array containing?
<alexgordon> often I *will* use a vector of vectors (though usually with a struct inbetween) but that only applies to "jagged" multidimensionals
<alexgordon> where different rows can be different lengths
<joelteon> no, it's 4x4
<alexgordon> 4x4 exactly?
<alexgordon> if you know the size at compiletime, then it's even easier
<alexgordon> use std::array
<alexgordon> no need to do any of that pesky dynamic allocation
<alexgordon> hm I wonder if it's possible to have a template argument that defaults to another template argument
<alexgordon> TO THE C++ TESTER
<alexgordon> (chocolat)
<alexgordon> apparently it IS
<joelteon> dude, C++ is actually pretty decent
<joelteon> I'm kind of understand what the fuck is going on
<joelteon> understand ING
<whitequark> >C++
<purr> Let it be known that whitequark loves >C.
<whitequark> >decent
<whitequark> no.
<whitequark> just no.
<whitequark> >C--
<purr> Let it be known that whitequark is indifferent to >C.
<alexgordon> --C++
<alexgordon> -- C++
<purr> Let it be known that alexgordon hates C++.
<alexgordon> ++ C ++
<purr> Let it be known that alexgordon loves C ++.
<alexgordon> -- C ++
<purr> Let it be known that alexgordon is indifferent to C ++.
<alexgordon> -- C --
<purr> Let it be known that alexgordon hates C --.
<alexgordon> ++ C++
<purr> Let it be known that alexgordon is indifferent to C++.
<whitequark> the fuck exactly you are doing
<alexgordon> enjoying my delirious state
<whitequark> is it global and mutable?
<alexgordon> joelteon: as I've been saying for fucking YEARS, if you simply avoid the bad features, C++ is a good language
<alexgordon> it's terrible to write with other people, if they have not seen the Truth (and insist on using virtual member functions, for example)
<alexgordon> C++11 is actually spectacularly good. they added some bullshit (e.g. custom literals) but you don't *have* to use that
<joelteon> custom literals sounds cool
<joelteon> i'd take it
<joelteon> does c++ do its own varargs
<alexgordon> custom literals are really stupid. I've never used them
<alexgordon> I should clarify, it's custom *string and number* literals
<alexgordon> I guess there might be a use for them if you're making say a bignum library
<alexgordon> but I hardly ever use bignums...
<alexgordon> joelteon: this is what I came up with BTW (please make allowances for my delirious state) https://gist.github.com/fileability/7738bb498ea9e98b246c
<joelteon> delirious huh
<joelteon> that's cool though
<joelteon> i like it
<alexgordon> extra credit if you can implement addition, subtraction, negation and multiplication!
<joelteon> i don't want to
<alexgordon> so what does your 4x4 array contain then?
<alexgordon> I just looked at an extremely pink website, and now my whole world has gone green
<joelteon> tiles
<alexgordon> (no it wasn't barbie.com)
<prophile> was it prawn
<prophile> hardclaw prawn
<alexgordon> yes and no
<prophile> david cameron will be very upset
<joelteon> but then
<joelteon> that's
<joelteon> wait does that mean the contents of the array have to be known at compile time too
<joelteon> o_o
<alexgordon> prophile: to say nothing of mumsnet!
<alexgordon> joelteon: no
<joelteon> how do I make an array
<joelteon> I can't figure it out
<alexgordon> {{1, 2}, {3, 4}}
<alexgordon> maybe?
<joelteon> what if I want to zero-initialize it
<alexgordon> that should work
<alexgordon> Matrix<int, 4> x;
<joelteon> nah, i can't
<alexgordon> ?
<joelteon> it has to be an array of pointer to tile, i think
<alexgordon> Matrix<Tile*, 4> x;
<alexgordon> array default constructs, which for a pointer is 0
<alexgordon> (aka NULL)
<alexgordon> prophile: I feel ashamed to admit that it was actually softer than softcore
<prophile> was it pictures of hardcore ceramics
<alexgordon> that doesn't sound very soft...
<joelteon> oh yes
<joelteon> i'm so goddamn smart
<prophile> actually it makes a lovely sound
<prophile> they used to use plates for reverb you know
<alexgordon> prophile: are you taking advantage of my delirious state?
<alexgordon> by feeding me nonsense
<prophile> hah
<prophile> plate reverb is a real thing, but it was generally sheets of metal as i understand
* alexgordon hits prophile with a foam pyramid
* prophile is hit with a foam pyramid
<prophile> i'm going to go and play dota now
<prophile> and there is nothing you can do about it
* alexgordon has no idea what dota 2 actually is
<alexgordon> except that it's not half life 3
<alexgordon> or HL2E3
<alexgordon> or Portal 3
<alexgordon> or TF3
<alexgordon> or anything ending with 3
<devyn> prophile: I read that as "pilates for reverb"
<prophile> devyn: that too
<prophile> a rare but wonderful method
<alexgordon> so I just googled "foam cone" and got this http://cdn.shopify.com/s/files/1/0029/4912/files/foam-mag-camille-2.jpg?101840
<alexgordon> erm, foam pyramid
dnyy has left #elliottcable [#elliottcable]
<alexgordon> was too much for dnyy
<devyn> alexgordon: hawt
* alexgordon looks at it again... to make sure it's as hot as it was 30 seconds ago
<joelteon> alexgordon: can I write implicit conversions from builtin types to my types
<alexgordon> yep
<alexgordon> C++ does that automatically in constructors unless you tell it not to
<joelteon> what does the signature for const char* -> foo look like
<alexgordon> make a constructor for foo that accepts const char* or std::string
<joelteon> oh
<joelteon> ohhhhhhh
<joelteon> ok
<alexgordon> C++gasm
<joelteon> can I automatically convert T to T*
<joelteon> that might be a bad idea though
<joelteon> yeah it probably is
<alexgordon> sounds like a bad one
<alexgordon> other way around might be OK
<alexgordon> joelteon: remember that smart pointers exist
<alexgordon> and are almost always better than manually deleteing stuff
eligrey has quit [Ping timeout: 240 seconds]
<devyn> o_O
<devyn> okay, I hadn't seen the new Mac Pro
<devyn> what the fuck is this
<devyn> serious sci-fi shit
<devyn> wow
<alexgordon> devyn: have you been under a trash can for all these months?
<devyn> alexgordon: it looks like one ;)
<devyn> but yes pretty much
<devyn> lol
<purr> lol
<devyn> this is actually very impressive
<alexgordon> devyn: it's impressive until you realize it's ridiculously overpriced for what it is: a computer
<devyn> of course it is
<devyn> the design is still very pretty, if completely static
<devyn> and efficient
<joelteon> too bad C++ doesn't have prettyprint
<joelteon> that's the one thing I really like to have
<alexgordon> joelteon: yeah, though these things exist
<alexgordon> devyn: and they could have made one for $2000 with one normal GPU, a normal i5, normal non-ECC ram, etc
<alexgordon> heck they could have made one for $1250
<alexgordon> *that* I would have bought
<alexgordon> not gonna pay $4-10k for a computer, in any universe
<devyn> I don't know, I see this as a rather clever experiment
<devyn> no way would I buy it
<devyn> but it's still very neat
<alexgordon> yeah but meanwhile apple's top of the range computer *has a fucking screen attached* ...
<whitequark> alexgordon: ECC RAM is a necessity if you have that much
<alexgordon> whitequark: define "that much"
<joelteon> i really want a mac pro
<whitequark> alexgordon: mac pro has 64g
<alexgordon> always?
<whitequark> I think?
<devyn> no, just up to
<alexgordon> I only need 16 anyway
<devyn> the lowest config is 12 GB
<alexgordon> hell I have 3 on here
<alexgordon> 3 is pushing it
<alexgordon> 8 would probably do
<devyn> (yes, $3000 for a computer with 12 GB)
<devyn> >_>
<alexgordon> 16 gives enough room for future proofing
<alexgordon> 64 is just insane, for me a programmer
<devyn> I have 16 GB of RAM in my machine
<whitequark> alexgordon: then you don't need a fucking mac pro.
<devyn> it wasn't expensive at all
<alexgordon> whitequark: right but I DO need a mac
<whitequark> devyn: 64G isn't expensive either, especially non-ECC
<alexgordon> and if it's not a mac pro, then it's got to be an imac
<alexgordon> that's what I'm complaining about
<alexgordon> there's no headless imac
<whitequark> that'll be under $200 I think
<whitequark> maybe $400
<alexgordon> only mac mini and mac pro. the former is too constrained, the latter is too expensive
<devyn> and why the fuck is it 1866 MHz RAM
<whitequark> then you're fucked
<devyn> 12GB ECC DDR3-1866 (3 x 4GB)
<devyn> who needs that
<alexgordon> doesn't DDR4 exist yet?
<devyn> yes, but it's not A Thing™ you can buy yet
<whitequark> devyn: do you really bump into memory throughput?
<alexgordon> wow they've been working on it for 9 years o_O
<alexgordon> never knew these things had such lead times
<joelteon> 64 would be so goddamn cool
<devyn> whitequark: no, I don't, so I don't see why 1866 MHz is necessary
<whitequark> alexgordon: anything mass-produced in general does
<whitequark> devyn: oh, you're complaining it's too high.
<devyn> yes
<alexgordon> whitequark: talking about DDR4, btw
<whitequark> they seem to have developed it for hardcore multimedia tasks
<whitequark> editing 4K video
<whitequark> *that* may well use 64G of 1866MHz memory
<alexgordon> isn't that like, what avid does?
<whitequark> there's a reason it's called "pro"
<devyn> yes, I suppose so actually
<alexgordon> right but, since the product exists, surely there's no point apple making their own one
<whitequark> unlike most other products it's not just a marketing gimmick; you'll buy mac pro if:
<alexgordon> for such a limited market
<whitequark> 1) you actually need that much horsepower
<devyn> whitequark: though on Macs, a lot of multimedia related stuff is largely GPU-accelerated anyway
<whitequark> 2) you're a fucking hipster who has too much money
<whitequark> 3) there is no third group.
<devyn> whitequark: there's a reason Apple pushed OpenCL so much
<whitequark> devyn: now look at its three GPUs
<alexgordon> whitequark: I know which group ELLIOTTCABLE falls into
<whitequark> alexgordon: did he buy one?
<devyn> whitequark: it's a Xeon CPU so two GPUs
<alexgordon> whitequark: he bought one *years* ago
<alexgordon> ;)
<alexgordon> whitequark: few years ago, he bought a pretty much top of the line mac pro
<whitequark> well, can't blame him for that, even if he just wakes up and masturbates while looking at that trashcan
<alexgordon> I have no doubt he'll be upgrading
<whitequark> whatever makes you happy, etc
<devyn> lol
<purr> lol
<alexgordon> *joke about elliott's hand*
<devyn> it does have one hell of a lot of Thunderbolt 2 ports
<whitequark> hm
<alexgordon> lol yeah apple's wishful thinking
<whitequark> now that I look at the specs, they seem smaller
<alexgordon> "surely if we just add more thunderbolt ports, someone will make some devices. right guys?"
<whitequark> it's not actually impressive anymore...
<whitequark> alexgordon: funniest thing is, apple doesn't even follow their own spec
<devyn> alexgordon: idk, I'm hopeful... TB *is* impressive, and there are non-Apple thunderbolt motherboards you can buy
<alexgordon> it's impressive like FW800 was impressive
<whitequark> by that spec, thunderbolt maintenance must be all provided in ACPI
<devyn> heh
<whitequark> but it's NOT
<whitequark> so while linux implements complete thunderbolt spec, it can't use thunderbolt on macs
<alexgordon> heh
<whitequark> because they don't fucking follow their own fucking spec
<whitequark> (it can now, there's been a driver implemented, but still)
<whitequark> thunderbolt is... idk what is it good for
<whitequark> it has display and pci-e passthrough. display is, well, cool. everything else? for storage, usb3 is on par, because even with fastest SSDs, the drive is the slow element
<whitequark> and there's nothing else you can want pci-e passthrough for
eligrey has joined #elliottcable
<whitequark> ok, I wrote an allocator, hooray
<devyn> whitequark: PCI-e passthrough could be neat, especially with the new Mac Pro, since you can't add any PCI-e cards to it
<ELLIOTTCABLE> errr
<ELLIOTTCABLE> wrong
<ELLIOTTCABLE> can't stay here and talk
<ELLIOTTCABLE> but,
<ELLIOTTCABLE> drive isn't weak link,
<ELLIOTTCABLE> only if *single*
<ELLIOTTCABLE> majority of people who care about mac pros or thunderbolt or ANYTHING
<ELLIOTTCABLE> or hell, a *large* minorit if people who care aobut *macs*,
<ELLIOTTCABLE> have raids
<ELLIOTTCABLE> nowadays
<ELLIOTTCABLE> raid on the desk next to the charger and Thunderbolt screen for your laptop is Standard Practice nowadays, par for the course
<ELLIOTTCABLE> anybody who can afford a Mac nowadays can also afford a small, user-sized (non-poweruser) RAID
<ELLIOTTCABLE> and those can *definitely* keep up with Thunderbolt way moreso than USB3, etcetcetc
<ELLIOTTCABLE> sorry staying out of it, leaving
<ELLIOTTCABLE> <3 all been a while want to talk soon can we have sex? okay bye
<purr> Let it be known that ELLIOTTCABLE hearts all been a while want to talk soon can we have sex? okay bye.
<devyn> :)
<alexgordon> ELLIOTTCABLE: cock
alexgordon has quit [Quit: Textual IRC Client: www.textualapp.com]
<ELLIOTTCABLE> 9oo9
<whitequark> MOTHERFUCKING ALLOCATOR
<whitequark> I basically have a bunch of Rust code which misbehaves
<whitequark> and I know the result is a hard fault somewhere
<whitequark> and there is no debugging information
<whitequark> and assembly is fucking incomprehensible
<whitequark> and it uses a hand-written allocator which may or may not overwrite random memory
<whitequark> ARGH
<whitequark> at least I have an in-circut debugger.
<whitequark> oh and the compiler conveniently flattened everything into a completely unreadable blob of instructions
<joelteon> im a moran
<devyn> a morman? you don't say
<joelteon> I wrote this code in C++ and it's like 8x slower
<joelteon> probably because vectors are wrong
eligrey has quit [Quit: Leaving]
<devyn> mildly NSFW, but hilarious: https://www.youtube.com/watch?v=6ZRSqR6p3gI
<joelteon> That was weird
<devyn> yep
<devyn> I still have no idea why it exists
<whitequark> devyn: the fuck I just watched
<devyn> I have no clue
<whitequark> cuttle: hey cuttle
<whitequark> you are clever
<whitequark> I'm dumb, explain this to me: http://en.wikipedia.org/wiki/Benford's_law
<joelteon> man, honey clears your sinuses right out
<joelteon> is alexgordon gone
<joelteon> that facker
<joelteon> or fuker
<joelteon> fucker
<devyn> honey? why not just eat thai food
<joelteon> because we *have* honey
<joelteon> and we don't have thai food
<devyn> you don't have thai food?
<devyn> who doesn't have thai food?
<joelteon> you have thai food at hand?
<devyn> yes
<joelteon> what's wrong wiht you
<joelteon> with
<devyn> :D
<joelteon> i cant type
<joelteon> I should write an idling game using vty
<devyn> haha
<devyn> hahahah
<joelteon> heh, the 80s
<whitequark> those legwarmers are still A Thing
<whitequark> at least according to TMnI/TKnR...
<cuttle> hi whitequark
<cuttle> whitequark: the way I understand Benford's Law is that there are a lot of distributions out in nature where things are unbounded, but it's less and less likely as you get higher
<cuttle> maybe
<cuttle> whitequark: like, it won't work for IQs, since those are known to follow a normal distribution around 100 (in fact definitionally)
<cuttle> whitequark: but yeah, my understanding is kind of vague, but I think it always has to be since it's a vague law that applies to a lot of different distributions
<cuttle> whitequark: but like, imagine you have numbers that are usually in the several-hundreds, but the highest ones stretch into the 1000s
<cuttle> whitequark: then having 1 as the first digit includes several 100-sized chunks
<cuttle> whitequark: as well as a chunk below
<whitequark> hrm
<whitequark> it supposedly works regardless of units of measurement
<whitequark> explain that!
<cuttle> well, units of measurement are simply multiplying by a constant
<cuttle> if you scale everything in the distribution the same constant amount
<cuttle> it's still the same shape
<whitequark> but but, if the most frequent digit was 1 and I multiplied it by two
<whitequark> wouldn't that mean the new most frequent digit is 2 ?
<cuttle> no because all of the 5s and 6s and 7s and 8s and 9s of an order of magnitude less also became 1s
<whitequark> oooh.
<whitequark> that makes sense
<cuttle> so, 1-9 being equally probable doesn't work
<cuttle> because it's not scale-invariant
<cuttle> you want a distribution to be the same regardless of scaling, for it to apply to lots of numbers with lots of units
<cuttle> so, taking it as an assumption that it's scale-invariant, benford's law is the resulting solution
<cuttle> it has a great digram in there, where if you take something equally distributed from 1-9, the thing I said where 5-9 become 1 makes 1 *ten times* more likely
<cuttle> so in a way benford's law is sort of a fixpoint of scaling data sets
<whitequark> hmmm
<cuttle> nuck: haven't seen you in forever bb
yorick has joined #elliottcable
Sgeo has quit [Read error: Connection reset by peer]
alexgordon has joined #elliottcable
Determinist has quit [Ping timeout: 246 seconds]
Determinist has joined #elliottcable
fwg has quit [Ping timeout: 240 seconds]
fwg has joined #elliottcable
alexgordon has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
sharkbot has quit [Remote host closed the connection]
sharkbot has joined #elliottcable
alexgordon has joined #elliottcable
<yorick> whitequark: yeah isn't it weird
<whitequark> ec should make out with him
eligrey has joined #elliottcable
fwg has quit [Ping timeout: 246 seconds]
yorickpeterse has quit [Quit: The NSA took my baby]
yorickpeterse1 has joined #elliottcable
yorickpeterse1 has quit [Client Quit]
yorickpeterse has joined #elliottcable
eligrey_ has joined #elliottcable
eligrey has quit [Disconnected by services]
eligrey_ is now known as eligrey
<joelteon> ok
<joelteon> i need to make an ASCII pile of poo
<jesusabdullah> libcaca?
<nuck> cuttle: ohai >_>
<alexgordon> joelteon: ascii?
<joelteon> yes
<alexgordon> no unicode?
<alexgordon> joelteon:
<joelteon> a poodle
eligrey has quit [Quit: Leaving]
<alexgordon> joelteon: oh right you wanted a pile
<alexgordon>
<alexgordon>
<alexgordon>
gozala has quit [Ping timeout: 272 seconds]
gozala has joined #elliottcable
gozala has quit [Ping timeout: 245 seconds]
silentbicycle_ has quit [Ping timeout: 272 seconds]
ELLIOTTCABLE has quit [Ping timeout: 246 seconds]
upgrayeddd has quit [Ping timeout: 252 seconds]
ELLIOTTCABLE__ has joined #elliottcable
silentbicycle_ has joined #elliottcable
upgrayeddd has joined #elliottcable
gazoombo has quit [Ping timeout: 240 seconds]
upgrayeddd has quit [Ping timeout: 245 seconds]
upgrayeddd has joined #elliottcable
silentbicycle_ has quit [Ping timeout: 240 seconds]
silentbicycle_ has joined #elliottcable
ELLIOTTCABLE__ has quit [Ping timeout: 240 seconds]
ELLIOTTCABLE__ has joined #elliottcable
<joelteon> so I need to make an ascii pile of poo