00:00:02 <oerjan> actually why would that be necessary...
00:00:10 -!- Robdgreat has left.
00:01:07 <oerjan> ...there's not mascarpone on the wiki.
00:01:17 <Arc_Koen> http://catseye.tc/projects/mascarpone/README.markdown.html
00:02:35 <oerjan> "New operations can be defined as strings of symbols, and these symbols are given meaning by an interpreter that is "captured" in the definition, similar to the way that lexical variables are captured in closures in functional languages."
00:03:13 <elliott> Phantom_Hoover: yeah, gimme a few minutes
00:03:32 <oerjan> i think that means you can cheat by having it map to a function instead.
00:03:47 <oerjan> and i vaguely recall the implementation did that, or was it emmental...
00:04:07 <Arc_Koen> well implementing emmental was surprinsigly easy
00:04:19 <Arc_Koen> you just stored new symbols as a list of symbols
00:04:38 <Arc_Koen> and when you're looking for the definition of a symbol, you look for it in the list of definitions
00:05:10 <Arc_Koen> and when you've find it, you look for the definition of each symbol in the definition, in the tail of the list of definitions
00:05:21 <Arc_Koen> mascarpone is slightly more tricky
00:05:35 <Arc_Koen> I was planning to have some kind of functions yes
00:05:59 <oerjan> this reminds me of some of ais523's problems with Feather.
00:06:18 <oerjan> that is also ridiculously meta-circular.
00:07:39 <Arc_Koen> btw I have no idea what "the way that lexical variables are captured in closures in functional languages" is
00:08:31 <Arc_Koen> can you tell me more about feather?
00:09:17 <monqy> not even hackego can
00:09:46 <Phantom_Hoover> but seriously, feather's a vague concept wrapped in years of in-jokes
00:09:48 <Bike> i think the meaning there is that interpreters "close over" operations in that you can't get the operation "out" of the interpreter except in the way specified in that interpreter's semantics
00:09:57 -!- Phantom_Hoover has quit (Quit: Leaving).
00:10:03 -!- Phantom_Hoover has joined.
00:11:41 <Arc_Koen> but it's the interpreter that is captured in the operation
00:12:22 <Bike> oh, yeah, other way around
00:12:28 <Arc_Koen> the way I see it an interpreter is a (symbol * operation list) list
00:13:00 <Arc_Koen> and an operation is (symbol list) * interpreter
00:13:01 <oerjan> Arc_Koen: yeah that doesn't work for the bottom interpreter
00:13:30 <Arc_Koen> so interpreter would be something like Null | Initial | New of (symbol * operation list) list
00:13:39 <Arc_Koen> so technically it's potentially not infinite
00:13:51 <Arc_Koen> but I have a hard time explaining that to ocaml
00:14:16 <Arc_Koen> because there's no equivalent to the "let rec x = f(y) and y = g(x)" for types
00:14:44 <Arc_Koen> that is, if I define operations first, ocaml is gonna tell me it doesn't know what "interpreter" means
00:14:44 <oerjan> it is probably better to see an operation as some function performing whatever action the operation actually should perform.
00:14:56 <oerjan> and i would guess that's what the implementation does.
00:14:56 <Arc_Koen> yeah, ok, I'll try to do it that way
00:15:02 <Bike> hm, perhaps i'm misunderstanding the page, but it seems just like reified environments
00:15:15 <Arc_Koen> I don't want to look at chris' implementation before I do my own
00:17:04 <Bike> wait, how do you put strings on the stack?
00:17:22 <Bike> oh, wait, nevermind, it's obvious.
00:21:49 <Arc_Koen> btw the wiki seems to handle the /// page badly
00:22:16 <Arc_Koen> that is, "Slashes" is the redirect and /// the page, so /// is categorized, but links to /// directly don't work
00:23:30 <oerjan> @tell atriq <atriq> My son looks a bit like me, he can put away the plates after dinner now thanks to edwardk! <-- hum...
00:23:55 <oerjan> Arc_Koen: i sometimes nag elliott about that.
00:24:02 <kmc> shachaf: someone is porting Mosh to AIX
00:24:11 <oerjan> Phantom_Hoover: not in the main namespace of esolang
00:24:29 <elliott> It's about me being irresponsible.
00:24:41 <oerjan> as long as we agree on that!
00:24:47 <shachaf> kmc: How's the Android port doing?
00:24:48 <elliott> kmc: is mosh available for plan 9 yet
00:24:55 <elliott> actually mosh does something that plan 9 should never be made to do
00:25:46 <oerjan> Arc_Koen: it used to work at the time /// was a featured language, but somehow it broke again.
00:25:49 <kmc> shachaf: somebody has a fork of connectbot that does mosh
00:25:51 <kmc> elliott: which
00:26:14 <kmc> also did you see http://content.dell.com/us/en/enterprise/d/campaigns/xps-linux-laptop.aspx
00:26:27 <kmc> it's an ultrabook-ish laptop from dell that comes with ubuntu
00:26:48 <kmc> and is also like all cloud and shit
00:27:46 <oerjan> elliott: by upgrading, i assumed?
00:28:02 <shachaf> What's with the resolution?
00:28:24 <elliott> oerjan: I changed the config
00:29:25 <elliott> Arc_Koen: "I'm actually not sure what concatenative exactly mean". it means that syntactic concatenation denotes semantic composition
00:29:49 <elliott> broadly, it means that if you get two valid programs and append their sources, the meaning of the resulting program is the function composition of those two programs
00:30:02 <elliott> (in a traditional stack language, your semantics are a function from stacks to stacks)
00:30:02 <Bike> 'nothing more'? but that's cool
00:30:08 <Bike> something something monoids
00:30:23 <elliott> so if you have 3 3 + that's the composition of the program that takes a stack and pushes 3 to it, that same program again, and the program that takes two elements off the top of the stack, adds them, and pushes the result back
00:30:32 <Arc_Koen> yes yes that's the idea in carriage
00:30:34 <elliott> which gives you a program which pushes 6 to a stack
00:30:48 <elliott> Bike: I don't think you need an identity for this to be well-defined, actually
00:31:11 <Arc_Koen> but I mean, apart from possible nameclashes in more complex languages
00:31:29 <Arc_Koen> I think that many many languages are concatenative
00:31:59 <elliott> many actual real programming languages?
00:32:20 <elliott> I guess you can sort of argue that C is sort of concatenative kind of, but not really
00:32:57 <Arc_Koen> well the fact that you need a function called main is embarrassing
00:34:32 <Phantom_Hoover> elliott why do i get the sense you are not doing anything relating to dwarf fortress
00:35:20 <oerjan> i guess ocaml fits better than C
00:35:38 <elliott> Phantom_Hoover: a minute ok
00:35:47 <oerjan> hm wait what about imports
00:35:55 <oerjan> or can ocaml have those anywhere
00:36:23 <oerjan> haskell must have them at the top, alas
00:36:30 <oerjan> and also has that pesky main
00:36:40 <Arc_Koen> Ocaml is better than Haskell THERE YOU HAVE IT
00:37:16 <oerjan> i assumed that was so you could chase imports without reading the whole file
00:37:19 <elliott> i am glad haskell makes you organise imports reasonably
00:37:25 <elliott> oerjan: well haskell has to process modules to parse
00:37:48 <oerjan> well that's a separate last parsing step now
00:37:59 <Phantom_Hoover> so guys do you all know about alonzo church's son dating haskell curry's daughter
00:38:16 <oerjan> Phantom_Hoover: vaguely rings a bell
00:38:22 <monqy> I think I've heard of it before
00:38:43 <kmc> shachaf: you should be able to query the unicode character database via DNS
00:39:00 <Bike> where are the CS paparazzi when you need 'em
00:39:52 <elliott> Phantom_Hoover: that is the coolest thing ever
00:39:55 <shachaf> kmc: Any idea why UnicodeData.txt is incomplete?
00:39:57 <Arc_Koen> both superturing's parents are functional?
00:40:14 <kmc> what do yo umean
00:40:33 <shachaf> Doesn't contain a bunch of codepoints.
00:41:26 <shachaf> I'm referring to ftp://ftp.unicode.org/Public/UNIDATA/UnicodeData.txt
00:41:37 <shachaf> Where else should I get my up-to-date Unicode information?
00:44:14 <pikhq_> That is a weird-looking character.
00:44:50 <shachaf> OK, 人 is also not in there.
00:44:58 <pikhq_> (which is of course 人 + 中)
00:45:17 <shachaf> You'd think U+4EBA would be there, right?
00:45:28 <shachaf> It's missing all those what'sitcalled characters.
00:45:50 <pikhq_> Yeah. 人 was definitely in the first Unicode version with CJK.
00:46:00 <pikhq_> The typical kindergartner knows it.
00:46:08 <elliott> Only some of these characters display here. :(
00:46:09 <kmc> oh that's not actually a big box?
00:46:18 <kmc> i guess i got a double width replacement box :3
00:47:35 <shachaf> pikhq_: Maybe it's just not part of UnicodeData.Txt
00:48:12 <shachaf> ftp://ftp.unicode.org/Public/UNIDATA/CJKRadicals.txt seems to have it.
00:48:28 <pikhq_> Makes sense. It *is* a radical.
00:48:49 <shachaf> Can't I get one file with all the codepoints in a uniform format?
00:49:02 <shachaf> I thought UnicodeData.txt was supposed to be complete!
00:50:56 <elliott> how about some more minutes
00:51:20 <shachaf> zomg ftp://ftp.unicode.org/Public/UNIDATA/BidiMirroring.txt
00:51:33 <shachaf> I feel like I should've known about that.
00:52:47 <kmc> wtf "PARAPHRASE BRACKET"
00:52:50 <kmc> who uses that
00:56:09 <shachaf> ftp://ftp.unicode.org/Public/UNIDATA/NamesList.txt
00:56:12 <shachaf> Maybe that's the file I need.
00:56:23 <shachaf> Of course I'll need to write a more complicated program to process that. :-(
00:57:49 <Bike> I'm not sure I understand what @/expand is for in macarphone.
00:58:13 <kmc> i like the idea of a kid who loves Japan and always reads about Japan on the Internet (like you do) but doesn't have Japanese fonts
00:58:21 <kmc> and so he thinks that Japanese people use a writing system where you put four hex digits in a box
00:58:24 <kmc> and learns to read this
00:59:23 <elliott> kmc: that sounds like the plot to a really bad short story
00:59:47 <kmc> better as a minor character detail in a story
00:59:59 <kmc> are the han characters in unicode ordered in any particular way?
01:00:05 <kmc> by stroke or something
01:00:10 <shachaf> They're ordered by codepoint.
01:00:42 -!- Phantom_Hoover has quit (Remote host closed the connection).
01:01:03 <shachaf> 19F0KHMER SYMBOL TUTEYASAT * represents the second Ashadha in the lunar calendar during the Adhikameas leap year
01:01:15 <shachaf> Lots of details UnicodeData.txt doesn't have.
01:01:43 <shachaf> @+These form a set of four quine corners, for quincuncial arrangement. They are also used in upper and lower pairs in mathematic, or more rarely in editorial usage as alternatives to half brackets.
01:01:58 -!- jfischoff has quit (Quit: jfischoff).
01:02:16 <shachaf> These form a set of four corner brackets and are used editorially. They are distinguished from mathematical floor and ceiling characters. Occasionally quine corners are substituted for half brackets.
01:03:17 <shachaf> This style is sometimes known as open-face or blackboard-bold. Double-struck symbols already encoded in the Letterlike Symbols block and omitted here to avoid duplicate encoding. Considerable variation in font style is acceptable, as long as the glyphs retain the visual element of doubled strokes.
01:03:34 <shachaf> 1F374FORK AND KNIFE = restaurant, meal * glyph may show a fork and spoon * glyph may show a crossed fork and knife * glyph may show a fork and knife with a plate
01:04:26 <Bike> i like all the emoticons, myself
01:04:35 <Bike> iirc there are some ones for catgirlfaces
01:04:44 <elliott> monqy: do you want to be in the succession fort......
01:06:45 <oerjan> <elliott> kmc: that sounds like the plot to a really bad short story <-- reminds me of a part of the plot to a book series about swedish (i think?) immigrants to america. one guy got this swedish-english phrase book and learned it by heart. only when he tried to speak with americans did he discover that english isn't spoken as it is spelled.
01:07:30 <Lumpio-> But does the name list have 1F4A9 PILE OF POO?
01:08:54 <shachaf> 1F4A0 DIAMOND SHAPE WITH A DOT INSIDE = kawaii, cute * meaning of cuteness is based on association of glyph with shape of a flower
01:10:11 <kmc> oerjan: haha
01:16:06 <elliott> monqy: i want to start the fortress but it's impossible to play df properly without listening to its background music but i don't want to :{
01:16:24 <monqy> elliott: what if you get over your issue w/ its background music
01:17:19 <oerjan> those things take time, monqy
01:18:01 <oerjan> that explains why the dorfs don't all croak immediately
01:52:12 <oerjan> typical, i get at least 40 http://ompldr.org/vZ2kycw/iqtest.PNG
01:52:59 <shachaf> monqy: lens now allows you to implement unsafeCoerce
01:55:20 <oerjan> and i can find no reasonably canonical interpretation that gives one of the numbers listed.
01:55:35 -!- madbr has joined.
01:55:41 <madbr> I kinda wonder... are there any ARM systems-on-a-chip that can do über-low-latency sound processing?
01:56:06 <oerjan> (the question is "how many squares do you see", btw)
01:57:46 <kmc> madbr: there are SoC's with integrated DSP
01:58:41 <kmc> a combination of that and running a RTOS on the ARM part should suffice for many things
01:59:45 <madbr> isn't the main CPU a lot faster?
01:59:57 <kmc> depends what you are doing i guess
02:00:11 <kmc> i don't know really
02:01:56 <oerjan> elliott: some small electric pixies explained it to me.
02:04:15 <oerjan> i wonder if i may be a housecat http://www.sheldoncomics.com/archive/121129.html
02:04:18 <madbr> kmc: sound synthesis
02:04:55 <madbr> physical modelling algo with a delay line, bandpass filter, nonlinearity, exciter plus a bunch of effects
02:06:09 -!- c00kiemon5ter has joined.
02:06:45 <HackEgo> c00kiemon5ter: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: http://esolangs.org/wiki/Main_Page. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
02:07:13 <Sgeo__> elliott, monqy Fiora there was an update a bit earlier
02:07:57 <oerjan> it has occured to me that the itflabtijtslwi deadfish could be converted to use decimal (and so allow bigger numbers) if i could just find a way to do multiplication in a fixed number of substitution commands.
02:08:12 <madbr> kmc: tbh I've never worked with DSPs
02:08:25 <madbr> kmc : mostly because you never get access to them anyways
02:09:55 <oerjan> which i _think_ i can reduce to sorting a list of unary numbers
02:13:08 <oerjan> which i still don't know how to do.
02:21:26 <madbr> yeah on cell phones it's almost impossible to get low latency
02:22:00 <madbr> they are built for angry birds and youtube, not for music
02:22:17 <madbr> (well, making music that is... consuming music is a-ok of course)
02:22:29 <madbr> and even angry birds is a stretch
02:22:36 <pikhq_> And they're not that good at any of it.
02:22:39 <Gregor> Welp, got my new business cards.
02:22:43 <Gregor> Printed on the back of playing cards.
02:22:52 <pikhq_> Gregor: I saw. Those things are awesome.
02:23:32 <pikhq_> Shoulda made it pony themed.
02:23:59 <Gregor> I think “they're playing cards” is plenty eccentric.
02:24:11 <pikhq_> I mean "instead of playing cards".
02:24:34 <Bike> can you actually play cards with 'em?
02:24:43 <pikhq_> But whatever. There's many ways of being eccentric, and all of them pretty nice.
02:25:16 <pikhq_> Bike: Except that he won't have a deck for long.
02:25:28 <Bike> That's ok, he can just downgrade to Pinochle.
02:26:01 <Gregor> It is a real, complete deck of playing cards, yes, but I don't intend to use them as such since that would defeat the whole “business card” thing ;)
02:29:43 -!- sebbu2 has quit (Ping timeout: 245 seconds).
02:35:42 -!- kmc has set topic: Hagfish are usually not eaten owing to their repugnant looks, as well as their viscosity and unpleasant habits. | http://codu.org/logs/_esoteric/.
02:36:47 <kmc> The hagfish is kept alive and irritated by rattling its container with a stick, prompting it to produce slime in large quantities. This slime is used in a similar manner as egg whites in various forms of cookery in the region.
02:57:06 -!- Arc_Koen has quit (Quit: Comment on appelle un mec qui pilote un avion ?).
03:20:18 -!- jfischoff has joined.
03:28:09 -!- david_werecat has joined.
03:34:25 -!- sebbu has joined.
04:02:09 -!- oerjan has quit (Quit: Onght iodg).
04:18:20 -!- sebbu2 has joined.
04:21:23 -!- sebbu has quit (Ping timeout: 245 seconds).
04:31:45 <kmc> that is the most pretentious use of diæresis i have seen
04:31:47 <pikhq_> Well, if "noël", why not "poëtic"?
04:32:02 <kmc> i don't think even the new yorker would do that
04:32:18 <elliott> kmc: Do you know of a good XCompose configuration file?
04:32:21 <pikhq_> The excuse for "noël" is that it's spelled that way in French.
04:32:23 <elliott> ISTR the default one is really awfully terrible.
04:32:43 <kmc> http://mainisusuallyafunction.blogspot.com/2010/10/typing-mathematical-characters-in-x.html
04:33:03 <kmc> the default one is pretty good as far as actual letters, i've found
04:33:06 <shachaf> mathematical characters such as ꙮ
04:33:39 <kmc> but i want to use e.g. da do de di for ♬♩♫♪
04:33:53 <kmc> shachaf: i don't have that one
04:34:04 <kmc> i did add the unaccented greek alphabet though
04:36:04 <kmc> copy pasta
04:36:12 <madbr> elliott: you could use the spanish keyboard tbh
04:36:22 <pikhq_> elliott: I use the default.
04:36:26 <madbr> qwerty and has a lot of common accents
04:36:35 <pikhq_> elliott: I guess I'll have to ask you if you find a better one.
04:36:37 <elliott> that's not really my use-case
04:36:51 <elliott> pikhq_: i think i found one that redid the whole thing from scratch with the one problem that i didn't like the combos it used
04:36:59 <kmc> if you just want accented latin characters, the default xcompose is fine
04:37:35 <kmc> don't think it covers vietnamese, but most european languages using the latin alphabet
04:37:37 <pikhq_> Yeah. You can type "Erdős" just fine that way.
04:37:43 <elliott> I would kind of like LaTeX support :P
04:37:54 <pikhq_> Doesn't do Vietnamese, yeah.
04:37:56 <kmc> elliott: there's a gtk input mode
04:38:05 <kmc> or... whatever thingy gtk uses
04:38:07 <kmc> yeah probably
04:38:18 <kmc> insert gnome bashing here
04:38:22 <shachaf> mathematical characters like ☭
04:38:39 <pikhq_> 'Course, Vietnamese is a Roman alphabet rendering stress test.
04:38:44 <kmc> shachaf: you should write a c compiler named cccp
04:39:02 <kmc> oh it is already a research group name
04:39:16 <kmc> bet it's awkward when they get visiting postdocs from baltic states
04:39:18 <Bike> kmc: how do you tell what Multi_key is in your setup, do you know?
04:39:31 <kmc> well, it's just an X keysym
04:39:35 <kmc> you can bind it with xmodmap
04:39:48 <kmc> but i guess for the default you woud have to look in the keyboard layout definition files
04:39:59 <pikhq_> I suspect by default it's unbound.
04:40:17 <kmc> or just open xev and then mash all keys
04:40:25 <Bike> yeah that was my previous method
04:40:40 <Bike> i already have win as hyper. i think i'm running out of weird modifier keys to use.
04:40:52 <kmc> on thinkpads you can use Fn as Multi_key
04:40:54 <elliott> i wonder what to put compose on
04:40:57 <elliott> i use right-alt as my xmonad key
04:41:04 <elliott> I guess I could use right win
04:41:04 <kmc> because it sends a key event if you press and release it without pressing anything else
04:41:18 <Bike> oh, it does? that's news to me
04:41:18 <elliott> kind of awkward to press tho :(
04:41:30 <elliott> i should buy a happy hacking keyboard so all the keys i can possibly press are readily accessible
04:41:43 <pikhq_> Yeah, but composing glyphs is a relatively infrequent thing.
04:41:47 <elliott> and switch to colemak and become a hermit and retreat into the antarctic wastelands and kill myself
04:41:49 <Bike> i suppose i could use H-Menu instead of Menu for wm shit
04:42:33 <kmc> i have left ctrl for opening browser tabs, left win for opening native apps, left alt for alt, right alt for WM actions, right win for Compose, and right Ctrl for nothing because it's a fucking pain to hit
04:42:48 <kmc> i also have Fn send Esc if you press and release, but I don't use it much
04:42:51 <pikhq_> elliott: Alas, I don't have the option.
04:42:54 <kmc> there is a hack to use Caps Lock as both Ctrl and Esc
04:43:02 <kmc> should switch to that
04:43:08 <kmc> i mainly use Ctrl-[ for Esc these days
04:43:38 <elliott> pikhq_: well antarctica is a little inaccessible I admit, but suicide?
04:43:55 <pikhq_> elliott: I was referring to the entire list.
04:44:06 <pikhq_> Suicide is essentially always an option, of course.
04:44:14 <kmc> colemak is for people who think dvorak is too mainstream
04:44:35 <Bike> what a terrifying thought
04:44:38 <pikhq_> A profoundly suboptimal option, but nevertheless.
04:44:43 <elliott> well if I am going to switch to an obscure keyboard layout
04:44:53 <elliott> then I better switch to the one whose superiority to QWERTY I am most convinced of
04:45:22 <kmc> i think it's pretty clear that both dvorak and colemak are superior to qwerty
04:45:46 <Bike> oh, are there good studies now?
04:45:47 <kmc> maybe colemak is slightly better than dvorak, maybe not, buti don't think it matters
04:46:00 <kmc> and dvorak is much more standard
04:46:03 <elliott> "I'd better switch to the one that I am most convinced is best"
04:46:25 <elliott> well I don't use other people's keyboards very much but it's awkward enough to switch to Dvorak on those that I'd rather just remember how to type with QWERTY
04:46:34 <kmc> it is not that awkward
04:46:48 <kmc> if i have to use someone else's computer for more than 5 minutes i would usually want to switch it
04:46:56 <elliott> I mean I already use xmonad and customise the programs I use fairly heavily
04:46:58 <kmc> switching to dvorak takes about 30 seconds in most operating sysem
04:47:03 <elliott> so someone else's computer is always going to be awkward
04:47:18 <kmc> winter is always going to be cold so you might as well go outside naked
04:47:19 <elliott> some people restrict themselves to almost-stock settings and the like to avoid that painfulness and that's fine but those aren't my priorities
04:47:58 <elliott> kmc: I don't find the gain compelling enough for the (expected) loss in the vast majority of my computer-using time given that the scenario it helps with is already moderately painful
04:48:36 <kmc> when i went to the computer store to evaluate laptop keyboards, it was very useful that i could switch them to dvorak
04:49:16 <pikhq_> "i went to the [...] store" — kmc, noted jokester
04:49:17 -!- sebbu has joined.
04:49:37 -!- sebbu has quit (Changing host).
04:49:37 -!- sebbu has joined.
04:52:38 -!- sebbu2 has quit (Ping timeout: 245 seconds).
04:53:11 <elliott> kmc: yo I'm gonna practice being a keyboard nerd since I know quite a fair bit about keyboards and it'd be a shame not to turn this knowledge into smugness:
04:53:20 <elliott> kmc: you don't have to type on laptop keyboards to know they're all shit
04:53:40 <kmc> sure but i still want to know which one is least shit
04:54:00 <shachaf> kmc: You should get a Model M.
04:54:10 <elliott> but they're seriously overrated
04:54:15 <elliott> (I typed on a Model M for >a year)
04:54:58 <kmc> i used a real model m for a few years and a unicomp model m clone for a few more
04:55:23 <elliott> even ignoring practical constraints like it has a ton of useless keys and is fucking huge and makes a racket, the buckling springs are just way too heavy really
04:55:42 <elliott> unless you have mechanical fingers of steel or whatever I'm pretty sure another keyboard will be better for typing
04:56:11 <kmc> but it's the mythical keyboard of Real Hackers from before I was born so it must be good
04:56:24 <shachaf> You know what's overrated?
04:56:29 <shachaf> Not having → and End keys.
04:56:32 <Sgeo__> Calling things overrated?
04:56:36 <kmc> meh screw keyboards, you should become a gun nerd
04:56:52 <kmc> there you can argue the superiority of designs almost 100 years old
04:57:00 <elliott> kmc: it's probably easier to kill people with a model m than a gun
04:57:12 <kmc> call the mythbusters
04:57:23 <elliott> this seems like a good time to bring up how much I fucking loathe numpads
04:57:27 <elliott> and wish they would retroactively never have existed
04:58:15 <madbr> shachaf: oh got yeah I can't live without home/end
04:58:36 <madbr> I have to admit that numpads... suck
04:58:57 <pikhq_> elliott: They're there for those people that want to input lots of numbers but can't be bothered to learn to touch type.
04:59:16 <elliott> well for spreadsheets/accounting and stuff I assume they are more efficient than the numpad
04:59:21 <kmc> if you only input numbers all day, presumably it is less finger movement
04:59:22 <elliott> but it should really just be a separate device
04:59:23 <Bike> well, my laptop keyboard is numpad-free
04:59:24 <Bike> suck it haters.
04:59:33 <elliott> and indeed you can buy USB numpads!
04:59:39 <elliott> so all we have to do is bring USB back in time
04:59:50 <kmc> indeed you can find USB numpads for free in basements
05:00:42 <madbr> bike: it's when they start removing other stuff like home, end, pgup, pgdown, F keys, delete, insert...
05:00:57 <Sgeo__> Without a numpad it's harder to play NetHack!
05:01:03 <Sgeo__> For people not used to vi-style keys
05:01:09 <Bike> hjkl is the future, sgeo.
05:01:13 <shachaf> Those aren't people, Sgeo__.
05:01:28 <quintopia> irc is the messaging system of the future
05:01:44 <pikhq_> Insufficiently XJSONML
05:01:45 <Bike> irc navigated via hjkl.
05:02:01 <pikhq_> ... I'm sorry for inventing that
05:02:05 <madbr> bike: you can't do that, hjkl types letters
05:02:16 <madbr> has to be reverse T shaped
05:02:18 <quintopia> madbr: you hit i before you want letters
05:02:22 <Bike> no, see, you just get a virtual keyboard, which you move around on with hjkl
05:02:46 <madbr> IJKL would be acceptable
05:02:54 <madbr> HJKL is the wrong shape
05:02:57 <elliott> colemak tries to keep zxcv in place
05:03:01 <elliott> if i was making a keyboard layout
05:03:06 <elliott> i would try to keep hjklyubn in the same place
05:03:16 <kmc> and you would squander half of the home row
05:03:34 <pikhq_> Because he'd make the left side chorded.
05:03:35 <madbr> hjkl belongs on VT terminals
05:03:46 <kmc> 'j' is one of the least used letters in English and it has the best possible spot in QWERTY
05:04:20 <Bike> but what about words like "jalopy"
05:04:35 <pikhq_> My name starts with the most important QWERTY letter!
05:04:35 <kmc> the only less used letters are x, q, and z
05:04:42 <kmc> you should feel special then
05:05:05 <pikhq_> Josiah "pikhq" Worcester does.
05:05:36 <shachaf> kmc: "p. sure the most important qwerty letter is h"
05:06:48 <elliott> for once i agree with quintopia
05:06:51 <pikhq_> I did stick it in my nick.
05:06:52 <elliott> seriously q is the best fucking letter in the alphabet
05:06:56 -!- jfischoff has quit (Quit: jfischoff).
05:07:40 <madbr> elliott: which is why no language has that sound?
05:07:42 <kmc> elliott: how do you type that upside down 'd'
05:07:46 <quintopia> agreement? this is too much. i may have to recant in order to not agree with elliott
05:08:00 <madbr> even jews have given up on q
05:08:03 <quintopia> but it's hard to recant a plain fact
05:08:33 <elliott> they can be as wrong as they want
05:08:36 <elliott> q will still be the greatest
05:08:42 <elliott> it doesn't even matter how it's pronounced
05:09:23 <pikhq_> I vote we replace "k" with "q".
05:10:35 <pikhq_> minnna, qe- wo qiȳu- tè qaqou!
05:11:23 -!- epicmonkey has joined.
05:14:34 -!- jfischoff has joined.
05:19:20 -!- jfischoff has quit (Client Quit).
05:22:14 -!- Frooxius_ has joined.
05:22:24 -!- Frooxius has quit (Read error: Connection reset by peer).
05:22:27 -!- Frooxius_ has changed nick to Frooxius.
05:31:11 -!- epicmonkey has quit (Read error: Operation timed out).
05:38:10 <Sgeo__> Why can't I su nobody?
05:38:17 -!- madbr has quit (Quit: Radiateur).
05:57:34 <kmc> jerk elliott
05:59:49 -!- sirdancealot8 has quit (Ping timeout: 248 seconds).
06:10:39 -!- Frooxius has quit (Ping timeout: 244 seconds).
06:20:13 <elliott> kmc: quelliott is like qualia but all the qualia are horrible.
06:20:55 <elliott> kmc: Also how come the incorrect URL highlights I was talking about stay *even when irssi scrolls*?
06:21:03 <elliott> Like, they move up with the scrolling.
06:22:50 -!- Frooxius has joined.
06:24:31 <elliott> kmc: How do you do diæreses with the default compose setup?
06:24:39 <elliott> compose " letter doesn't seem to -- oh
06:28:17 -!- FreeFull has quit.
06:30:36 <Sgeo__> Oh: I talked to that kid again today
06:31:05 <Sgeo__> The one who thinks prepared queries don't fully protect against SQL Injection
06:31:31 <Sgeo__> His new argument: The null byte breaks stuff, and according to him, PHP and most other languages except for Pascal use C-style strings
06:31:55 <Bike> it must be sad to be trapped in the seventies
06:32:10 <elliott> the null byte "breaks stuff"
06:32:18 <Sgeo__> elliott, not his words, sorry
06:32:18 <elliott> because somehow your input string gets a null byte in the middle of it
06:32:28 <elliott> despite your strings being null-terminated
06:32:36 <Sgeo__> elliott, assumed to be because of malicious user
06:33:00 <elliott> i'm saying it is impossible to have a nul in the middle of a null-terminated string
06:34:03 <Sgeo__> I checked out and noted to him that the SQLite and MySQL prepared query C-level APIs, for the parts where you give the API a string, it asks for a length (SQLite makes it optional either give length or assume nul-terminated string)
06:34:30 <Sgeo__> He said he definitely did attack a PHP script that used MySQL prepared queries with a null byte
06:36:25 <elliott> not sure why you are bothering to argue with this idiot
06:38:26 <Sgeo__> Oh, also with the null-terminated thing: He talked about using the null byte to do weird file system stuff. I said that was an OS-level issue, he said no because it works on both Windows and Linux
06:38:46 <Sgeo__> Found a page in the PHP documentation that mentions it, he just sort of said that it was wrong
06:38:56 <elliott> not sure why you are bothering to argue with this idiot
06:39:17 <pikhq_> Sgeo__: "Weird file system stuff"?
06:39:27 <Sgeo__> pikhq_, not his words, sorry
06:39:45 <pikhq_> Like... Terminating a path early?
06:40:21 <elliott> not sure why you are bothering to argue with this idiot
06:46:50 -!- Frooxius has quit (Ping timeout: 244 seconds).
06:51:25 -!- david_werecat has quit (Ping timeout: 260 seconds).
06:54:06 -!- Bike has quit (Ping timeout: 264 seconds).
06:54:56 -!- Bike has joined.
06:58:10 -!- rapido has joined.
06:58:34 -!- rapido has quit (Client Quit).
07:01:21 -!- Tod-Autojoined has joined.
07:01:21 -!- TodPunk has quit (Read error: Connection reset by peer).
07:22:30 <elliott> kmc: it would be cool if mosh provided fully local line editing falicities controlled by special control codes!!! i anticipate the chance of this feature getting added to be 0
07:22:34 <elliott> wow that sentence is awkward
07:47:22 -!- augur has quit (Remote host closed the connection).
07:55:50 -!- epicmonkey has joined.
08:12:10 -!- epicmonkey has quit (Ping timeout: 246 seconds).
08:24:37 -!- oklopol has joined.
09:15:14 -!- sebbu2 has joined.
09:15:14 -!- sebbu2 has quit (Changing host).
09:15:14 -!- sebbu2 has joined.
09:17:41 -!- sebbu has quit (Ping timeout: 244 seconds).
09:18:46 -!- Bike has quit (Ping timeout: 256 seconds).
09:26:19 -!- epicmonkey has joined.
09:35:44 -!- Phantom_Hoover has joined.
09:40:58 -!- Phantom_Hoover has quit (Read error: Operation timed out).
09:48:25 -!- augur has joined.
09:56:21 -!- Phantom_Hoover has joined.
10:49:26 <elliott> Phantom_Hoover: guess what i didn't do
11:08:40 <Sgeo__> I found a definition of ? that tricked me for a short while
11:09:03 <Sgeo__> It's a bit cheaty, but it doesn't even work anyway
11:10:01 <Sgeo__> If I did the math right (I didn't write it down), a ? a = a + 2 and a ? a ? a = a + 3
11:10:13 <Sgeo__> Unfortunately a ? a ? a ? a != 4
11:11:08 <Sgeo__> monqy, a vague notion of some operation, such that addition is sort of like repeated ?
11:11:35 <Sgeo__> I think it's kind of fun that that given operation works for a?a and a?a?a but not a?a?a?a
11:11:48 <monqy> so what are the properties said ? has to satisfy
11:12:59 <Sgeo__> Ideally, repeating it should result in addition, and it shouldn't be too trivial
11:13:14 <monqy> in what sense / in what sense
11:13:49 <Phantom_Hoover> i think Sgeo__ is unhappy with the successor function and wants to invent his own
11:13:51 <Sgeo__> Hard to pin down what I think is too trivial, but as far as the first in what sense: a?a = a+2; a?a?a = a+3; a?a?a?a = a+4, etc.
11:14:14 <monqy> ummm the successor function is unary Phantom_Hoover
11:14:54 <elliott> you have to give things up monqy .............. to achieve ur dreams
11:15:27 <monqy> heck it's!! not even associative!!!
11:15:32 <Sgeo__> I don't think I'm too worried about commutativity. But that definition is definitely too trivial. (Also wrong, but hey... although, come to think of it)
11:15:54 <monqy> is not associative what you mean by wrong
11:16:10 <Sgeo__> What I mean by wrong is a?a = a+1 when I was looking for a?a = a+2
11:16:42 <Sgeo__> Although, I guess looking for an operator where a?a = a+1 might be more fruitful... and again ignoring trivial things >.>
11:16:47 <elliott> a ? b = Sa unless this is the first ? and then it's a instead
11:16:58 <elliott> definition of ?: a ? a = a + 1
11:17:09 <monqy> a ? ... ? a = n where n is the number of a appearing in there
11:17:27 <elliott> 'a promising start' -people
11:17:48 <elliott> Sgeo__: ps what associativity is ?
11:18:20 <Sgeo__> Doesn't matter, I guess
11:18:45 <Sgeo__> Also, I think f(a,b) = ln(e^a + e^b) is a fun operator
11:19:07 <Sgeo__> I was sort of hoping that that or a variation of it could be used for this, but it's fun in and of itself, I think
11:19:35 <monqy> how about the "cool associativity" where a ? a ? a ? a = (a ? a) ? (a ? a)
11:20:00 <elliott> the Double Cool Associativity whereby a ? a ? a = a + 3
11:20:21 <elliott> the real joke: people who make a distionctioneoi rjowrjeoi wjeroi jweo fjowfj iowjf iowjef owjf kfwpfl [wlf lwf[p lw] r;f]3[r; ]f/; f ew,f ;lwker lp[ro[ 3- 2o=- =`= -=` 0-3i -4 k1o
11:20:21 <monqy> associativity where my dreams come true??? yes please
11:20:30 <elliott> monqy: do you ever give up
11:20:37 <elliott> that was me giving up on writing that sentence
11:20:56 <monqy> what was the sentence...............................
11:22:09 <fizzie> If you have the properties that a ? a = a + 2 and a ? a ? a = a + 3, then (a ? a ? a) ? (a ? a ? a) = (a + 3) ? (a + 3) = a + 3 + 2 = a + 5 -- but you also wanted a ? a ? a ? a ? a ? a in general to be a + 6, which is perhaps a bit of a problem.
11:22:54 <monqy> i don't think sgeo cares about associativity though, which is weird given the way he was writing his stuff!!!
11:23:04 <fizzie> Also going backwards from a?a?a = a+3, a?a = a+2, I suppose then a = a+1.
11:23:05 <monqy> "it's almost like he hadn't thought it through"
11:23:32 <elliott> fizzie: imo Double Cool Associativity solves these problems? and more?
11:23:58 <fizzie> Double Cool probably solves anything, it's so cool.
11:24:23 <elliott> fizzie: other solutions: infinity number system where every number is infinity
11:25:24 <Sgeo__> Well, is it really a big deal to say that ((a ? a) ? a) ? a not always = (a ? a) ? (a ? a)
11:25:37 <elliott> sgeo i think the a = a + 1 thing might be more of a problem
11:27:01 <monqy> it's unfair to say that a = a + 1 when he hasn't even explained his notation
11:27:12 <monqy> maybe a = a + 2!!!!!
11:27:43 <elliott> if we say that + is defined in the "usual" manner
11:28:01 <elliott> a + Sb = a ? (a + b) -- adjust for associativity
11:28:30 <elliott> i dont quite know why you wouldnt just go with a ? a = a + 1 tho
11:28:40 <elliott> not like 1 is privileged as a base-case for repeaty thirng rj jwraw klsf aew; ';.
11:28:44 * Sgeo__ can live with a ? a = a + 1
11:28:52 <monqy> then whats a ? (a + 1)!!!!!!!
11:29:43 <Sgeo__> Reason I was thinking a ? a = a + 2 is because a + a = a * 2
11:31:17 <elliott> i'll let fizzie figure it out
11:38:51 -!- augur has quit (Read error: Connection reset by peer).
11:39:05 -!- augur_ has joined.
11:46:02 <oklofok> "<Sgeo__> Well, is it really a big deal to say that ((a ? a) ? a) ? a not always = (a ? a) ? (a ? a)" if you don't have associativity, computations become harder usually.
11:57:42 <Sgeo__> Then (a+1)?(a+1) = a+1+1=a+2
11:58:10 <Sgeo__> But, ideally, (a?a)?a = a+2 = (a+1)?a
11:58:23 <Sgeo__> So (a+1)?a = (a+1)?(a+1)
12:00:38 <oklopol> oh was the idea to change the definition so that a?a = a+1 for all a but you still get some nontrivial operation?
12:00:44 <oklopol> so that a?b may not be a+1
12:01:12 <oklopol> then why do you say a+2 = (a+1)?a
12:01:37 <elliott> oklopol: because the point is to have addition be "repeated ?"
12:01:45 <elliott> so you have to define addition inductively by ?
12:02:24 <oklopol> in any case is there an associative operation on N such that a?a = a+1 for all a?
12:02:36 <Deewiant> And the end result will probably be the successor operation, ignoring the second argument of ? :-P
12:02:52 <Sgeo__> oklofok, there's an operation
12:03:08 <Sgeo__> Not sure if it's associative, but
12:03:18 <Sgeo__> log_2(2^a + 2^b) = a?b
12:03:22 <oklopol> if you want to have addition be repeated ?, then i'm pretty sure it has to be successor . takeleftarg
12:04:22 <Sgeo__> a?a = log_2(2*2^a) = log_2 2 + log_2 2^a = 1 + a
12:04:29 <elliott> Deewiant: that's ruled out by Sgeo__'s undefined nebulous triviality rule
12:05:41 <oklopol> then (a?b)?c = log_2(2^log_2(2^a + 2^b) + 2^c) = associative
12:06:07 <elliott> but is (a?a)?a = a+2?????????????? the mysteries
12:06:30 <oklopol> oh i though he checked that
12:06:56 <Sgeo__> No. At least, I'm fairly sure the answer is no. I came up with the general thing against the whole ? idea when wondering about this particular operation
12:06:57 <elliott> god (W|A) doesn't seem to think so either
12:07:33 <fizzie> You would have to have that (a+1)?a = a+2 thing if you wanted (a?a)?a = a+2.
12:07:34 <oklopol> so you have an operation that's associative such that a?a = a+1
12:09:46 <oklopol> in general, (a+n)?a = a+n+1 for all a with that idea
12:09:56 <oklopol> i mean what fizzie just said
12:10:29 <oklopol> and we require nothing for a?b where b is less than a
12:11:15 <oklopol> so that's the characterization right, any binary function such that a?b = a+1 when b \leq a
12:11:40 <elliott> Phantom_Hoover: ye but im tired right now so do you really want me to set up the fortress in that state
12:11:48 <elliott> consider how horrifying awake elliott fortresses are
12:12:06 <oklopol> in particular, you can make it commutative right
12:13:00 <oklopol> so, this is the unique commutative binary operation f such that addition is repeated f i guess
12:13:30 <oklopol> we could publish this in the journal of ridiculous theorems
12:14:52 <Deewiant> oklopol: If it's repeated with both arguments being the result of the previous, there are others
12:16:02 <oklopol> so my requirement was a + b = ((((a?a)?a)?a)?...)?a where a appears b+1 times on the right
12:16:45 <elliott> p. sure Sgeo__ meant what oklopol meant
12:16:47 <oklopol> and then Sgeo__'s is a commutative associative operation
12:16:57 <oklopol> or pretty much anything else.
12:17:07 <Deewiant> I just wasn't sure what you meant with "addition is repeated f"
12:17:21 -!- myndzi has quit (Ping timeout: 246 seconds).
12:17:30 <oklopol> but Sgeo__ already solved it for that def
12:17:43 -!- myndzi has joined.
12:17:58 <elliott> Deewiant: that was just a random string of symbols
12:18:30 <oklopol> elliott's just a very lucky guy
12:22:15 -!- Phantom_Hoover has quit (Read error: Operation timed out).
12:36:11 -!- Phantom_Hoover has joined.
12:38:36 <elliott> fizzie: do you know of a thing to get outgoing messages automatically split when they're too big w/ irssi
12:38:46 <elliott> i found this ancient perl script to do it but idk if it's any good or whatever
12:38:56 <fizzie> scripts.irssi.org splitlong.pl is what I use.
12:40:09 <fizzie> Except I use it with a /set splitlong_max_length 400 which is overly cautious, but the default -- which is calculate the exact limit programmatically -- sometimes miscalculates stuff due to things like freenode cloaks.
12:40:45 <fizzie> (The "fis@iris.zem.fi" it believes will be in the message is shorter than the real "fis@unaffiliated/fizzie".)
12:41:06 <elliott> what i found was http://scripts.irssi.org/html/autowrap.pl.html
12:41:39 <elliott> fizzie: solution: get rid of your cloak since you have a cool hostname??
12:42:12 <fizzie> Well http://scripts.irssi.org/scripts/autowrap.pl and http://scripts.irssi.org/scripts/splitlong.pl they're quite similar, except autowrap always uses that limit of 400, it seems.
12:42:21 <fizzie> Slightly curious that I hit the same number independently.
12:42:27 <elliott> wow, Debian has literally a kajillion identds
12:43:15 <fizzie> And it binds to event_send_text vs. sig_command_msg so it might affect some cases where the other would not, but it's split everything I've typed just fine.
12:43:36 <fizzie> The cool hostname unworks if I connect to a non-IPv6 server these days. :/
12:44:08 <elliott> fizzie: I connect with this fancy SSL thing. And then authenticate with this fancy SASL thing.
12:44:12 <elliott> It feels like the future times a billion.
12:44:17 -!- sebbu has joined.
12:44:17 -!- sebbu has quit (Changing host).
12:44:17 -!- sebbu has joined.
12:44:21 <elliott> my server is the one doing the connection
12:44:28 <elliott> i could make it connect with ipv6
12:44:36 <elliott> i don't know if i can handle that much future
12:45:31 -!- sebbu3 has joined.
12:45:51 <elliott> oklopol: http://www.worldipv6launch.org/measurements/
12:46:43 <elliott> fizzie: do you use a specific identd
12:46:47 -!- sebbu2 has quit (Ping timeout: 260 seconds).
12:46:51 <elliott> I wrote one once but I can't be bothered to use something that isn't apt-get!!
12:47:05 <elliott> Phantom_Hoover: the future briefly ceased being forever
12:47:50 <fizzie> elliott: I seem to be using oidentd, but I don't have a very special reason for it.
12:48:59 -!- sebbu has quit (Ping timeout: 260 seconds).
12:49:46 <fizzie> http://outside.hut.fi/2_days.png GRAPH
12:49:52 -!- sebbu has joined.
12:49:57 -!- sebbu3 has quit (Ping timeout: 256 seconds).
12:50:18 <Deewiant> Why do I get an IP address instead of a hostname
12:51:03 <ion> elliott: I use oidentd.
12:51:59 <fizzie> Deewiant: Perhaps Freenode does the bidirectional check?
12:52:03 <elliott> It's cold in the other H town too but probably less cold.
12:52:04 <fizzie> 190.188.75.109.in-addr.arpa domain name pointer 109-75-188-190.reverse.cust.as47215.net.
12:52:07 <fizzie> $ host 109-75-188-190.reverse.cust.as47215.net
12:52:07 <fizzie> Host 109-75-188-190.reverse.cust.as47215.net not found: 3(NXDOMAIN)
12:52:22 <Deewiant> Why do I have a hostname that doesn't work
12:52:22 <elliott> I see fizzie isn't using a secure connection.
12:52:42 <fizzie> That should go back to 109.75.188.190 so that everyone would accept the reverse as genuine.
12:52:47 <elliott> (How do I get irssi to prefer IPv6 servers for the irc.freenode.net thingy?)
12:52:56 -!- monqy has quit (Quit: hello).
12:53:19 <fizzie> You can say -6 to /connect and /server.
12:53:37 <ion> I tried SSL/TLS with IRC, but WeeChat’s upgrade without dropping connections (as well as Irssi’s for that matter) doesn’t work with encrypted connections.
12:54:23 <elliott> ion: For irssi that's not a problem since it doesn't get new releases.
12:54:35 <elliott> fizzie: is there a way to -6ify my existing server entry without retyping my settings :(
12:54:38 <Deewiant> Oh noes, you'll have to disconnect for a few seconds! YOU MIGHT MISS SOMETHING IMPORTANT
12:54:51 <Deewiant> elliott: /save and edit the config file?
12:54:59 <elliott> Deewiant: like a commoner!
12:55:12 <fizzie> Maybe if you add the server with the same name, it only edits the explicitly specified settings? Then again, maybe NOT.
12:55:15 <ion> deewiant: It’s not about missing messages, it’s about unnecessary quit/join spam.
12:55:24 <elliott> Oh, I should set up the verify-server-SSL-key thing too.
12:55:27 <elliott> What if someone IMPERSONATES FREENODE?
12:55:33 <ion> Especially if you use nightly builds and upgrade daily.
12:55:48 <fizzie> I did think about maybe SSLing, I think bip does that. But it doesn't do the SASL auth.
12:55:49 <Deewiant> I think that's a reasonable source of quit/join spam
12:55:50 <elliott> (it is so about missing messages)
12:55:59 <ion> elliott: What if YOU HAVE CHATTED ON A FAKE NETWORK ALL THIS TIME?
12:56:08 <fizzie> What if I'm a fake fizzie too.
12:56:17 <fizzie> You should "verify my key", if you know what I mean.
12:56:21 <fizzie> (I don't know what I mean.)
12:56:59 * ion signs fizzie’s public key, if you catch my drift
12:58:50 <elliott> I think if I mosh into my dtached irssi which connects with IPv6 ~verified~ SSL and authenticates with SASL using an identd that might just be too many things and I would explode.
12:58:59 <fizzie> My key has 18 signatures. And apparently expired in 2006.
12:59:06 <fizzie> (I don't get much PGP'd mail.)
12:59:09 <elliott> Oh hey, /server -6 irc.freenode.net DTRT.
12:59:15 <elliott> It just didn't show it as DoingTRT in /server list.
12:59:22 <fizzie> Or maybe that's the creation time.
13:00:28 <elliott> Oh, oidentd looks Totally Bloated.
13:00:33 <elliott> Hierarchical configuration files and everything!
13:01:35 <fizzie> elliott: That's weird; "/server blah" is supposed to just connect to blah, not edit the server list.
13:02:01 <fizzie> ("/SERVER disconnects the server in active window and connects to the new one. It will take the same arguments as /CONNECT.")
13:02:44 <Deewiant> I never liked that /server command in irssi. I was always worried that I'd accidentally disconnect.
13:02:49 <elliott> fizzie: Or was it /server add?
13:02:55 <elliott> Apparently it was: /server add -6 irc.freenode.net
13:02:56 <fizzie> That would make more sense.
13:03:05 <fizzie> It's nice that it did the right thing.
13:03:22 <elliott> Nicer than I've grown to expect of this client.
13:03:23 -!- sebbu3 has joined.
13:03:42 -!- sebbu3 has quit (Changing host).
13:03:42 -!- sebbu3 has joined.
13:04:31 <ion> elliott: The right thing to do with Irssi is to add a network (/network add Freenode IIRC), add a server to that (/server add -auto -network Freenode -6 irc.freenode.net IIRC) and connect using the network name (/connect Freenode IIRC).
13:05:08 <ion> That’ll let you add alternative servers in case one is down and it will also connect correctly on startup.
13:05:14 <fizzie> I believe this was already done, and the point was just to modify the settings of the server.
13:05:26 <ion> And you’ll be able to add channels to be joined automatically to networks.
13:05:49 -!- sebbu has quit (Ping timeout: 246 seconds).
13:05:59 <elliott> ion: My network has one (1) server.
13:06:02 <elliott> I never anticipate it having any more.
13:06:10 <fizzie> My freenode server has the host name "freenode", and my ircnet server is called "ircnet". (The bouncer listens to all connections in the same port and differentiates networks by the password, but irssi refuses to accept multiple identical host:port servers, so I have to use dummy names and /set proxy_address 127.0.0.1, which is stupid.)
13:06:28 <elliott> fizzie: Do people actually use IRCNet?
13:06:48 <ion> IRCnet is probably the most popular network among Finns.
13:07:43 <Deewiant> And now, a languages joke: http://i.imgur.com/fIB5o.png
13:07:49 <fizzie> http://irc.netsplit.de/tmp/networks/top10_2012u.png -- it's still quite high up there.
13:08:18 <fizzie> Beats Undernet and EFnet in users quite clearly.
13:08:48 <fizzie> Also Freenode has an interestingly strong week-periodic thing, stronger than the other networks.
13:09:03 <fizzie> I suppose it's because SERIOUS BUSINESSMEN use it for SERIOUS BUSINESS.
13:09:36 <oklopol> i've used quakenet, freenode, ircnet, ef
13:09:41 <Deewiant> I thought EFnet was at least bigger than Undernet.
13:09:52 <fizzie> I've used those same networks as o.
13:10:12 <Deewiant> I'm currently on QuakeNet, FreeNode, IRCnet, EFnet, and OFTC, and I've also used Undernet and maybe Rizon.
13:10:31 <fizzie> I'm no longer in anywhere else than Freenode and IRCnet. :/
13:11:27 <Deewiant> And I've been on Ustream and other specific things with non-IRC-client programs.
13:11:41 <oklopol> i'm on free and quake because those are the only places where people get sad if i don't show up on channels, and i can't manage to make irc autoconnect when my computer explodes and i have to restart.
13:12:09 <elliott> fizzie: Can you write an identd for me?
13:12:48 <elliott> What it should do is if you have an ~/.identrc it runs that passing the server that's asking for the identity, and uses its output as the identity.
13:12:56 <Deewiant> oklopol: http://webchat.xs4all.nl/ I think
13:12:57 <fizzie> I don't think I can be bothered.
13:13:07 <elliott> fizzie: Okay, but how about: reconsider?
13:13:13 <oklopol> so how many shoes do you people have, i apparently have like 10 pairs
13:13:17 <fizzie> Deewiant: It's listed as a network, though.
13:13:24 <fizzie> Deewiant: That one connects to real networks.
13:13:36 <Deewiant> fizzie: Oh, darn, missed that.
13:13:42 <Deewiant> I only read most of the left column.
13:13:58 <fizzie> It seems to mean irc.webchat.org.
13:14:03 <oklopol> most of them don't even have a sole
13:14:14 <fizzie> Which is some kind of a thing.
13:14:17 <Deewiant> And there's also www.webchat.org, which is its homepage.
13:14:25 <fizzie> Yes, but I don't understand it.
13:14:39 <elliott> Deewiant: Do YOU use an identd??? Wait, maybe you answered already.
13:14:54 <Deewiant> elliott: (Can't you tell with /whois?)
13:15:06 <elliott> Deewiant: That would be... cheating.
13:15:06 <fizzie> I've used the XS4ALL webchat from some HTTP-only airports.
13:15:46 <Deewiant> oklopol: Not enough: I need to buy new shoes ASAP because there is lots of snow everywhere and all my current shoes result in wet socks in such an environment.
13:16:18 <fizzie> Deewiant: Speaking of whois, what's this as47215.net all about? It seems to be some kind of a German webhost/colo place.
13:17:04 <Deewiant> fizzie: Yes, I have a virtual server from DE.Punkt.
13:17:14 <elliott> maybe you should live somewhere it doesn't snow
13:17:16 <oklopol> and i actually prolly have better suited shoes but changing shoes is such an emotional rollercoaster.
13:17:33 <Deewiant> fizzie: I (think I) changed the RDNS now via the Web UI, but if it worked it hasn't propagated yet.
13:18:25 <elliott> Why does freenode even bother to make ident requests?
13:18:26 <Deewiant> Well, I can query their nameservers directly, of course; it did work.
13:19:37 <elliott> I guess I'll install oidentd since nullidentd requires inetd. But still, ew.
13:21:09 <fizzie> oklopol: I have something like five or six pairs of shoes maybe, if you count everything? But only one pair of borderline sensible winter ones.
13:21:30 <fizzie> All the shoes are kind of different type of shoes, I don't really have redundant shoes.
13:21:31 <elliott> I have one (1) pair of shoes I actually wear.
13:22:07 <fizzie> My wife takes care of the redundant shoe ownership tasks in our household, which is I think a reasonably common setup.
13:22:15 <Deewiant> I have three pairs of outdoor shoes currently, two of which are redundant in the sense that I use them only when it's warm, and interchangeably.
13:22:35 <elliott> fizzie: If alt+<direction> produces "OD" and "OC" for left and right respectively, what's misconfigured?
13:22:38 <elliott> I'm so sick of this stuff not working.
13:22:44 <elliott> It never seems to get better.
13:23:44 <fizzie> elliott: I don't know how you managed to make the alt key not work, to be honest.
13:23:52 -!- sebbu has joined.
13:23:52 -!- sebbu has quit (Changing host).
13:23:52 -!- sebbu has joined.
13:24:13 <elliott> fizzie: Come on, it's like, not worked out of the box in Ubuntu for years and stuff.
13:24:26 <elliott> This is one of the most persistently broken things.
13:24:41 <elliott> I think it might be mosh's fault.
13:24:57 <Deewiant> They seem to produce [D and [C for me in zsh, and do some kind of cursor movement in bash.
13:25:41 -!- sebbu3 has quit (Ping timeout: 245 seconds).
13:25:42 <fizzie> Alt in general, including keys in irssi, have worked out of the box for me pretty much always. Though in bash alt-arrows seem to insert [D and [C for me.
13:25:52 <fizzie> I don't know what they should do there.
13:26:09 <elliott> http://www.freenode.net/certfp/certfp-irssi.shtml guys what is this about
13:26:20 <elliott> it seems to be a totally different thing to the things i have been setting up
13:26:22 -!- Arc_Koen has joined.
13:26:27 <elliott> it's... setting yourself up a certificate so freenode can verify you??? what is this
13:26:44 <fizzie> Yes, it seems to be a client certificate.
13:26:47 <elliott> that i have a script handle for me
13:26:50 <fizzie> I did not know they even did client certificates.
13:27:04 <elliott> it looks sort of like you can get passwordless logins by doing this
13:27:12 <elliott> but i have no idea how it differs from sasl in terms of goodness or whatever
13:27:23 <fizzie> It's X.509, it's like X-X-Xcellent.
13:27:55 <fizzie> "CertFP allows clients connected via SSL with a client SSL certificate to authenticate to services using the SHA1 fingerprint of their client SSL certificate. One must have registered with services."
13:28:02 <fizzie> I wonder if this is some kind of a new dealie.
13:28:33 <elliott> this seems possibly fancier than sasl??
13:28:55 <fizzie> It seems vaguely sensible in the sense that if you're doing SSL already, it's sort of in the same rigamarole.
13:29:10 <fizzie> I wonder if they've managed to not screw up the cloak thing when authenticating over that thing, though.
13:29:29 <fizzie> Since the usual nickserv-pass-in-PASS thing they did screw up.
13:29:33 <elliott> Cloaks work for me with SASL if that helps?
13:29:40 <elliott> And I've used NickServ pass in server password for forever.
13:30:24 <fizzie> SASL is I think the only guaranteed thing.
13:30:47 <fizzie> With server-pass it sometimes doesn't manage to enable the cloak before joining, if services are being laggy.
13:31:02 <elliott> I hear SASL is sometimes non-guaranteed i.e. times out or something.
13:31:19 <fizzie> I guess that might happen. But when it works, it always happens before anything else can happen.
13:32:07 <elliott> Now, if you want to verify the Freenode server SSL certificate against a certificate authority (CA), then you’ll need to download the CA certificate from the authority that signed the server certificate. In this case, its Gandi.net, and their CA certificate file can be found here: http://crt.gandi.net/GandiStandardSSLCA.crt. However, using the file in its native DER format for Irssi wasn’t working for me. So, using openssl, I converted the bina
13:32:14 <elliott> be found here: http://crt.gandi.net/GandiStandardSSLCA.crt. However,
13:32:17 <elliott> using the file in its native DER format for Irssi wasn’t working for me. So, using openssl, I converted the binary
13:32:20 <elliott> DER data file to PEM format, at which the Freenode certificate would properly verify:
13:32:23 <elliott> Sometimes I just don't trust technology, you know?
13:32:24 <elliott> (TODO: Get that wrapping thing being a thing.)
13:33:56 <fizzie> Mhm, bip can do client SSL certificates, maybe I should also set up that Freenode thing.
13:34:12 -!- elliott has quit (Quit: Reconnecting).
13:34:21 -!- elliott has joined.
13:34:31 <fizzie> I wonder what is the difference between "local minimum found" and "local minimum possible" from MATLAB fminunc.
13:35:32 <fizzie> You are at least identd'd.
13:35:39 <elliott> And I set ssl_verify, so I assume that's... working.
13:36:15 <elliott> 13:34:21 -!- hostname : li278-81.members.linode.com 178.79.159.81
13:36:18 <elliott> That doesn't look very IPv6.
13:36:26 <elliott> (Is there a way to tell whether I'm using an IPv6 connection?)
13:36:38 <elliott> 13:34:15 -!- Your host is morgan.freenode.net[64.32.24.176/6697], running version ircd-seven-1.1.3
13:36:47 <fizzie> That doesn't look very six.
13:36:48 <elliott> I guess that isn't terribly reassuring.
13:37:04 <elliott> Is there no SSL IPv6 service or something?
13:37:17 <Deewiant> Why are you using a server in the USA?
13:37:18 <fizzie> That doesn't sound like it'd make sense.
13:37:31 <elliott> Deewiant: I'm just using irc.freenode.net, man.
13:37:34 <elliott> fizzie: Wait, what doesn't?
13:37:49 <elliott> Also did you know: a whole two (2) people have edited the Esolang wiki anonymously using an IPv6 address.
13:37:57 <fizzie> Not having SSL for the IPv6.
13:38:02 <Deewiant> elliott: Yeah, and I question the sensibility of that.
13:38:20 <elliott> Deewiant: Well, what would you rather I do? List a bunch of servers manually and keep that up to date?
13:38:27 <fizzie> I'm using chat.eu.freenode.net; it's a compromise.
13:38:38 <elliott> This server is in London anyway, that's basically the US.
13:39:00 <Deewiant> I think I'm also using chat.eu.freenode.net, possibly with one manually listed server before that.
13:39:17 <shachaf> elliott: did you know someone could get the nick ehird right now if they wanted to
13:39:19 <elliott> "Shame they don’t allow auto-identification using client-side SSL certificates, like OFTC does." "@bloogle Yeah, but the SASL auth is every bit as good, if not better if using DH-BLOWFISH"
13:39:21 <fizzie> Okay, I also have lindbohm manually before it, but it's all cyclic so it's not really "prioritized".
13:39:39 <fizzie> http://sprunge.us/NXbZ -- see which continent is more like future.
13:39:41 <elliott> shachaf: I don't expect anyone except maybe shachaf t otry that.
13:40:17 <elliott> Well, you just looked up the last time ehird was logged into, so you must have some interest in the name.
13:40:33 <shachaf> elliott: By the way, is there an equivalent of Bazaar for a Setter?
13:41:12 <shachaf> So that Setter s t a b = s -> Foo a b t, of course.
13:42:24 <fizzie> Hey, there's a rajaniemi.freenode.net that's in some "TKK Comnet/FUNET" block, in Finland.
13:42:26 <elliott> Sounds like you just wrote it?
13:43:13 <elliott> fizzie: I bet you can't change just the server field of an irssi server. :(
13:43:19 <shachaf> elliott: Is there a name for it? Does it have interesting features?
13:44:01 <fizzie> elliott: I bet so too, since it'd identify the server to change by the host:port pair.
13:44:35 <elliott> I set ssl_verify without ssl_cert.
13:45:32 <Deewiant> elliott: It sounds like it'll then try to verify the server's certificate against who-knows-what certificates. Ideally the ones in /usr/share.
13:46:59 <elliott> Deewiant: I set ssl_capath to /etc/ssl/certs.
13:47:11 <elliott> So is that, like, the right thing?
13:47:57 <Deewiant> It has the standard set of trusted certs, at least.
13:48:42 <elliott> /server add -6 -ssl -ssl_verify -ssl_capath /etc/ssl/certs -auto -network freenode chat.eu.freenode.net 6697
13:49:02 -!- elliott has quit (Quit: Reconnecting).
13:49:10 -!- elliott has joined.
13:49:43 <elliott> I wonder what on earth is wrong.
13:50:58 -!- jix has quit (Ping timeout: 246 seconds).
13:51:23 -!- fizzie has quit (Quit: jumpin' jumpin').
13:52:23 <oklopol> someone should add ithkuil to the joke Deewiant linked
13:52:48 <elliott> I think instead Deewiant should help me get IPv6 working.
13:54:09 <Deewiant> I don't have IPv6 connectivity on any machine I own. I could set it up in my LAN but I haven't bothered.
13:54:17 <Deewiant> I don't think we have it at work either.
13:55:02 <shachaf> elliott: IPv6 is the devil.
13:55:23 <shachaf> By having fewer IP addresses than humans, we keep the resource valuable.
13:55:38 <shachaf> "ipv4 addresses: earth's greatest export?"
13:56:43 -!- fizziew has joined.
13:56:46 <fizziew> But what about THE INTERNET OF THINGS?
13:57:14 <fizziew> Also what is this stupid delay in reconnecting, I have the stupidest bouncer, it's stupid.
13:57:42 <fizziew> I'm MISSING MESSAGES here. Not that it's about that.
13:58:56 -!- fizzie has joined.
13:59:20 -!- fizziew has quit (Client Quit).
13:59:26 <elliott> fizzie: OK, so you get to help me instead!!
13:59:33 <elliott> To wit: WHY ISN'T IT WORKING FIX IT.
13:59:35 <fizzie> (Shocking news: it was myself who was the stupid.)
14:00:28 <fizzie> I don't know why it is not working, but why don't you just "nc -6 chat.eu.freenode.net 6667" to see if that goes anywhere?
14:00:41 <fizzie> Also I have a CLIENT CERTIFICATE now.
14:00:51 <fizzie> I am also using a SECURE CONNECTION.
14:01:19 <Deewiant> You are also using a Lithuanian server instead of rajaniemi.
14:02:15 <elliott> Deewiant: I use chat.eu.freenode.net and you still complain?
14:02:46 <elliott> 13:49:04 -!- - Welcome to barjavel.freenode.net in Paris, FR, EU.
14:02:53 <elliott> I... don't think that's Lithuania, Deewiant.
14:03:00 <fizzie> It was about me, I think.
14:03:23 <fizzie> Deewiant: I put {rajaniemi, lindbohm, chat.eu} to the list, and apparently the cycling thing was at chat.eu at the time when I fixed the stupid bit.
14:04:52 <elliott> Deewiant: OK but I actually am fizzie.
14:05:14 <fizzie> Oh no, am I elliott then? I... I'm not sure I want to. It sounds difficult.
14:05:31 <shachaf> fizzie: I think elliott isn't telling the truth.
14:05:41 <elliott> fizzie: Is this client certificate stuff easy to set up?
14:05:42 <shachaf> Because Finland ≠ Hexham, right?
14:06:12 * elliott wonders which of 70 flavours of netcat this is.
14:06:55 <fizzie> elliott: It was reasonably easy for bip, I don't know about irssi. Probably it's just one setting, since Freenode doesn't actually verify the client certificate, it just uses the fingerprint from it.
14:07:15 <Deewiant> Ithinks because recently openbsd-netcat started conflicting with GNU netcat in Arch and I went with GNU and it says the same thing.
14:07:33 <elliott> Deewiant: Well, *I*thinks no GNU software would say "nc -h for help".
14:07:42 <elliott> And not even accept --help.
14:07:52 <Deewiant> elliott: Well, you didn't say that bit.
14:08:19 <elliott> Deewiant: Re the client setting: I was thinking more of the making a certificate part.
14:08:21 <fizzie> (The one I tried with was the netcat-openbsd of Ubuntu something-at-work, and it has a -6.)
14:08:33 -!- imelliott has joined.
14:09:39 <elliott> So um this thing works fine, I am tempted to: blame the whole SSL thing?
14:10:10 <Deewiant> Turn off SSL and see if it works in irssi?
14:11:53 <fizzie> I got a v6'd SSL connection just fine, though. But maybe it's worth a try.
14:12:02 <elliott> Deewiant: What if people spied on me?
14:12:13 <Deewiant> If you're only here, you're publically logged anyway.
14:12:50 <fizzie> You could also try /connect -6 or something. Though the -6 setting *should* end up in there for listed servers too, I think.
14:13:05 <Deewiant> elliott: Don't be elsewhere while you're testing.
14:13:14 -!- imelliott has quit (Ping timeout: 246 seconds).
14:14:04 <fizzie> Deewiant: Incidentally, since you found out about Lithuania... does that client certificate fingerprint also show up in /whois to others?
14:14:37 <Deewiant> fizzie: No, only that and "is using a secure connection" apart from the usual.
14:16:00 <shachaf> elliott: Is it also a comonad?
14:16:05 <shachaf> elliott: When a ~ b, that is.
14:17:14 <Deewiant> I think that it is once again blizzard time. -->
14:17:30 <fizzie> It's like bullet time, except you dodge snowflakes.
14:26:02 -!- yhoje has joined.
14:27:26 -!- yhoje has left.
14:27:56 -!- yhoje has joined.
14:36:12 -!- yhoje has quit (Ping timeout: 264 seconds).
14:40:04 -!- yhoje has joined.
14:46:19 -!- yhoje has quit (Ping timeout: 264 seconds).
14:52:57 -!- yhoje has joined.
14:58:35 -!- atriq has joined.
14:58:57 -!- yhoje has quit (Ping timeout: 252 seconds).
15:00:37 -!- nooodl has joined.
15:05:25 <atriq> I like today's Girl Genius
15:05:25 <lambdabot> atriq: You have 1 new message. '/msg lambdabot @messages' to read it.
15:06:18 <atriq> @tell oerjan it was something about Plated letting us traverse self-similar descendants
15:07:37 <shachaf> atriq: Did you hear: upon was renamed to onceUpon, and uponTheDeep was renamed to upon?
15:09:24 <atriq> This makes it somewhat harder to use lens to sing Adele songs in Haskell!
15:11:59 <atriq> Although slightly easier to recite Grimms' Fairy Tails
15:12:46 <quintopia> fairy tails are tastier than gator tails
15:17:27 <atriq> I'll trust you on that, having never tasted either
15:23:00 -!- Frooxius has joined.
15:25:39 -!- boily has joined.
15:43:42 <Arc_Koen> I can't escape ' in my terminal any longer
15:43:53 <Arc_Koen> for instance echo '\' prints a backslash
15:44:31 <nortti> it should do that, I think
15:45:25 <Arc_Koen> yeah, but the " quotes usually cause a lot of problems
15:45:41 <Arc_Koen> for instance if the string I want can be interpreted as bash commands or something
15:45:51 -!- Phantom_Hoover has quit (Read error: Operation timed out).
15:49:54 -!- Phantom_Hoover has joined.
16:01:22 -!- jfischoff has joined.
16:05:53 <Phantom_Hoover> elliott, OK so seriously, are you ever going to start that fort.
16:07:08 <fizzie> Deewiant: Blizzard time indeed; I just waited 35 minutes on the Konemies bus stop, from 17:25 to 18:00, because the 17:35 510 failed to apparate at all. (Unless it's this one and really late.)
16:08:24 <fizzie> Arc_Koen: I tend to use 'blabla'\''bleble' if I want the string blabla'bleble as an argument.
16:08:49 <fizzie> Admittedly the '\'' sequence looks quite crummy, but it works.
16:09:25 <elliott> Phantom_Hoover: i've been busy ok
16:09:47 <elliott> fizzie: hey can you @tell elliott get ipv6/client certificates working
16:10:15 <fizzie> @tell ellitt get ipv6/client certificated workdjsjud
16:11:03 <elliott> lambdabot doesn't let you self-@tell!!
16:11:13 <fizzie> Hey my battery outruns.
16:11:26 <quintopia> elliott: can't you /nick and self-tell
16:11:28 <fizzie> @tell elliott hhhey get ipv6/client certificated workdjsjud
16:12:10 <elliott> fizzie: that's about as good as my typing, so it's good enough
16:12:10 <lambdabot> elliott: You have 1 new message. '/msg lambdabot @messages' to read it.
16:12:17 <elliott> quintopia: ye but that is like spammy & inelegant & stuf
16:13:11 <quintopia> so is the method you used, but whatev
16:13:16 <shachaf> @ty ?spammy & ?inelegant & ?stuf
16:13:17 <lambdabot> (?spammy::a1, ?inelegant::a1 -> a, ?stuf::a -> b) => b
16:15:28 -!- ogrom has joined.
16:24:45 -!- atehwa has quit (Ping timeout: 246 seconds).
16:34:30 -!- AnotherTest has joined.
16:51:20 -!- nooodl_ has joined.
16:52:58 <Deewiant> fizzie: I only had to wait 10 minutes, missing one 103 and the second coming late; fortunately these 10[23] buses apparate quite frequently around that time of day.
16:54:23 -!- nooodl has quit (Ping timeout: 246 seconds).
17:02:12 -!- myndzl has joined.
17:03:23 <fizzie> 510 is not the most reliable bus line ever.
17:04:31 <Deewiant> By "missing one" I mean that one didn't come at all, or then it was at least 3 minutes early which I think is exceedingly unlikely in this weather.
17:04:55 <Deewiant> So there was a bit of unreliability all around.
17:07:22 -!- epicmonkey has quit (Ping timeout: 265 seconds).
17:07:59 -!- myndzi has quit (Ping timeout: 246 seconds).
17:08:00 -!- lightquake has quit (Ping timeout: 246 seconds).
17:08:00 -!- lightquake has joined.
17:16:37 -!- FreeFull has joined.
17:18:41 -!- oklopol has quit.
17:47:00 -!- ogrom has quit (Ping timeout: 255 seconds).
17:49:09 -!- jix has joined.
17:51:32 -!- atehwa has joined.
17:51:45 -!- Bike has joined.
17:54:48 -!- david_werecat has joined.
17:55:06 -!- david_werecat has quit (Client Quit).
17:55:06 -!- sebbu3 has joined.
17:55:27 -!- sebbu3 has quit (Changing host).
17:55:27 -!- sebbu3 has joined.
17:58:36 -!- sebbu has quit (Ping timeout: 265 seconds).
18:08:14 -!- atehwa has quit (Remote host closed the connection).
18:08:21 -!- atehwa has joined.
18:10:36 -!- jfischoff has quit (Quit: jfischoff).
18:16:49 -!- atehwa has quit (Ping timeout: 260 seconds).
18:17:53 -!- sebbu3 has changed nick to sebbu.
18:18:29 -!- AnotherTest has quit (Quit: Leaving.).
18:23:38 -!- atehwa has joined.
18:27:57 -!- atehwa has quit (Ping timeout: 244 seconds).
18:29:10 -!- epicmonkey has joined.
18:36:42 <atriq> Aaaaaaaaaaaaaaaaaah!
18:36:58 -!- atriq has quit (Quit: aoiufhewpiufhaeeiufhweiupf).
18:38:31 <shachaf> -- | An slow, iterated SHA512 hash function to make dictionary attacks more
18:38:32 <shachaf> defaultHash a = (iterate hash a) !! 512
18:39:20 -!- ogrom has joined.
18:39:32 -!- ogrom has quit (Client Quit).
18:39:35 -!- david_werecat has joined.
18:40:26 -!- atehwa has joined.
18:41:45 -!- soundnfury has joined.
18:44:37 -!- atehwa has quit (Ping timeout: 246 seconds).
18:50:42 -!- sebbu3 has joined.
18:50:48 <olsner> that makes it look like (iterate hash a) !! n = SHA-n
18:51:02 -!- sebbu3 has quit (Changing host).
18:51:02 -!- sebbu3 has joined.
18:52:04 -!- jfischoff has joined.
18:53:41 -!- sebbu has quit (Disconnected by services).
18:53:48 -!- sebbu3 has changed nick to sebbu.
18:56:24 -!- Bike has quit (*.net *.split).
18:56:24 -!- lightquake has quit (*.net *.split).
18:56:25 -!- fungot has quit (*.net *.split).
18:56:59 <olsner> funny how shachaf also posted the exact same four lines in #haskell-blah
18:57:21 <shachaf> olsner: What's wrong with that?
18:58:47 <olsner> it's confusing because it resulted in slightly different conversations, only one of which contained what I thought I replied with
18:58:55 -!- Frooxius_ has joined.
18:59:09 <olsner> also, duplication annoys me
18:59:46 <ion> I don’t really see a problem with IRC crossposting when both recipients are likely to find the message interesting.
19:00:09 <shachaf> olsner: Not my fault that you subscribe to both channels. Most inhabitants of one channel don't subscribe to the other.
19:00:54 <olsner> there are like several people who are in both channels
19:01:24 <ion> I am on both channels and i didn’t get annoyed by that at all.
19:01:54 -!- Bike has joined.
19:01:54 -!- lightquake has joined.
19:01:54 -!- fungot has joined.
19:03:20 * shachaf resists the temptation to cross-post to every channel ion is in.
19:04:14 <Gregor> “Android is upgrading… optimizing app <n> of 118”
19:04:39 <Gregor> What do you bet that “optimizing” in this context involves absolutely nothing that anyone would reasonably consider “optimizing”
19:05:42 <olsner> iirc, there is something optimization-like (or perhaps just code generation) it does to the bytecode that gets cached
19:05:45 -!- Frooxius has quit (Read error: No buffer space available).
19:08:36 <FreeFull> shachaf: What function is hash?
19:09:02 <FreeFull> And it just happens to get ran 512 times too
19:10:05 <kmc> shachaf: what code is that from?
19:12:10 <kmc> they should probably use PBKDF2 instead
19:12:18 <kmc> i don't know how much it matters though
19:14:36 <shachaf> They should probably be using a number bigger than 512.
19:18:03 -!- zzo38 has joined.
19:18:57 <fizzie> But you can't use anything else than 512 iterations with SHA-512. It would be unseemly.
19:19:41 <kmc> shachaf: that too
19:20:40 <kmc> Django 1.4 defaults to 10,000 iterations of SHA256
19:21:12 <shachaf> They also don't use PBKDF2 or anything like that?
19:21:31 <kmc> no i mean PBKDF2 with 10,000 iterations
19:22:33 <shachaf> Er, actually, I have no idea how fast or slow PBKDF2 iterations are.
19:24:51 <kmc> yeah it is kind of low
19:26:36 <kmc> django's implementation takes 75 ms on my AMD Phenom(tm) II X6 1075T Processor
19:26:48 <kmc> is claimed to be "four times slower than openssl's implementation"
19:29:19 <kmc> i wonder if it's increased in the soon to be released Django 1.5
19:31:25 <kmc> hm bcrypt is explicitly designed to foil GPU-based attacks
19:31:48 <kmc> and small ASICs, but not FPGAs with lots of RAM
19:31:49 <kmc> http://security.stackexchange.com/questions/4781/do-any-security-experts-recommend-bcrypt-for-password-storage
19:33:08 <kmc> obviously i should do 100,000 iterations of SHA512 and then bcrypt that
19:33:11 <kmc> that's defense in depth right
19:33:26 <Bike> the maginot line of encryption
19:33:53 <kmc> information security has a lot of maginot line situations
19:34:16 <kmc> people work really hard to secure one area and then get owned by something trivial and completely different
19:42:02 <pikhq_> Gregor: It's running a device-specific optimizer pass on the bytecode, actually.
19:42:47 <pikhq_> Gregor: Swapping endianness, some constant propagation, etc.
19:42:53 -!- monqy has joined.
19:56:17 <zzo38> Do you like the symmetric variants I made which is related to some of asymmetric chess variants?
19:56:22 <shachaf> Gregor: Where should I find burritos in San Francisco?
19:56:55 <Gregor> shachaf: Literally anywhere.
19:57:25 <Gregor> Walk outside and look for a Mexican restaurant that isn't a chain. Probability is very high that you will find one within two blocks that will be good.
19:57:25 -!- oerjan has joined.
19:58:19 <oerjan> happy australian mailman list reminder day, everyone!
19:58:19 <lambdabot> oerjan: You have 1 new message. '/msg lambdabot @messages' to read it.
19:58:24 <lambdabot> atriq said 4h 52m 5s ago: it was something about Plated letting us traverse self-similar descendants
19:59:32 <kmc> is that true for all of SF
19:59:42 <oerjan> 04:31:30: <kmc> "poëtic"
19:59:42 <oerjan> 04:31:45: <kmc> that is the most pretentious use of diæresis i have seen
19:59:42 <kmc> i want to see a taqueria density heatmap
20:01:02 <oerjan> i recall poëta being used as an example of something needing a diæresis in latin
20:01:12 <kmc> shachaf: have you seen http://www.jwz.org/images/5112789301_da50dca4bd_b.jpg
20:03:32 <oerjan> <shachaf> mathematical characters like ☭ <-- don't you know that dialectic materialism is _better_ than mathematics, duh
20:05:15 <oerjan> kmc: is the joke that tenderloin is the most crime-ridden district in san francisco, or something?
20:05:21 <kmc> that's one of the jokes
20:05:40 <kmc> i don't get them all
20:05:50 <kmc> mission is a hipster fixie
20:05:59 <kmc> noe valley is full of kids i guess
20:06:18 <kmc> castro is full of gay men who work out all the time
20:06:22 <kmc> haight is weird
20:06:46 <kmc> though i think going down those hills on a unicycle would be terrifying
20:06:52 <Bike> what is the point of fixed-gear bikes, by the way
20:07:16 <kmc> FiDi is a folding bike for people who commute by train
20:07:54 <kmc> well, having only a single gear gives you a simpler and more reliable bicycle
20:08:08 <kmc> lacking freewheeling seems less useful
20:09:12 <olsner> if they have a bike, shouldn't they just commute by bike?
20:09:38 <kmc> not 30 miles from palo alto
20:09:41 <kmc> though people do it
20:09:52 <kmc> http://sf2g.com/
20:11:30 -!- david_werecat has quit (Ping timeout: 264 seconds).
20:11:48 <kmc> i guess that is the opposite direction
20:11:55 <kmc> people who live in SF and bike to google every day
20:12:08 <oerjan> `addquote <kmc> colemak is for people who think dvorak is too mainstream
20:12:16 <olsner> hmm, I thought google was already in SF?
20:12:19 <HackEgo> 858) <kmc> colemak is for people who think dvorak is too mainstream
20:12:36 <kmc> the main google office is in Mountain View
20:13:00 <kmc> which is part of the "San Francisco Bay Area" but about 40 miles from downtown SF
20:13:09 <HackEgo> 758) <kmc> haters gonna make som valid points
20:13:10 <kmc> and completely different as far as what it's like to live there
20:13:14 <Phantom_Hoover> olsner, this is america, they don't like putting things next to other things
20:13:26 <kmc> well SF itself is a proper dense city
20:13:38 <kmc> but like every city (in Europe as well) it has sprawling suburbs around it
20:13:47 <HackEgo> 365) [on petrol] <ais523> oklofok: it's actually poisonous, so I advise against drinking it <Phantom_Hoover> ais523, also contains benzene, my carcinogen of choice.
20:13:50 <HackEgo> 227) <oerjan> (the former is a very deep theorem, i'd have had to read the whole book to understand it, so i didn't.)
20:13:52 <HackEgo> 851) <Phantom_Hoover> unfortunately df is not yet able to simulate norway
20:13:52 <HackEgo> 297) <crystal-cola> anyway I've stopped ``trolling'' <crystal-cola> since apparently you guys don't like me claiming obviously false and absurd claims
20:14:44 <kmc> the proper dense city part of SF is approximately a square 10 km on a side
20:17:36 <Phantom_Hoover> huh, apparently there are people who prefer their bikes not to have brakes
20:18:48 <olsner> hmm, so these chips are "fine cheese" flavored
20:21:10 <fizzie> The accomplishment of today: googled out why the VLC logo is a traffic cone.
20:21:18 <fizzie> "One day, people from the VIA association (VIA is a students' network association with many clubs ... amongst those is VideoLAN. More info here : http://www.via.ecp.fr ) came back drunk with a cone. They then began a cone collection (which is now quite impressive i must say). Some time later, the VideoLAN project began and they decided to use the cone as their logo."
20:21:26 <fizzie> Such a meaningful story.
20:22:29 <olsner> ah, I collected a cone once ... they're bigger and heavier than you think and end up being not quite as fun as you expect
20:23:23 <kmc> apparently you can write GCC plugins in Python now O_O http://gcc-python-plugin.readthedocs.org/en/latest/
20:25:44 <zzo38> Does it mean you do not have to recompile GCC if you want it to do something else?
20:29:14 <oerjan> @tell Sgeo__ You realize ? must have the property a ? (a+k) = a+(k+1) for all natural numbers k? (also you have to decide whether it should work left or right associatively, like tetration.)
20:29:46 <kmc> zzo38: that is already the case, gcc can load plugins from .so's
20:30:25 <oerjan> @tell Sgeo__ (tetration from exponentiation)
20:32:16 <oerjan> @tell Sgeo__ Sorry, k >= 2.
20:32:21 -!- HackEgo has quit (Read error: Connection reset by peer).
20:32:48 * oerjan swats lambdabot -----###
20:33:03 -!- HackEgo has joined.
20:33:25 * oerjan also swats Sgeo__ for having so many underlines he's worried he'll not see the message.
20:37:38 -!- lambdabot has quit (Ping timeout: 256 seconds).
20:38:11 * oerjan swats everyone else for not mocking his swatting. -----###
20:42:00 -!- carado has joined.
20:42:29 -!- lambdabot has joined.
20:43:53 -!- sivoais has quit (Remote host closed the connection).
20:44:34 -!- sivoais has joined.
20:46:04 -!- sirdancealot8 has joined.
20:48:42 <oerjan> ok oklofok seems to have pretty much solved the problem.
20:49:19 <fizzie> That's what always happens.
20:49:39 -!- epicmonkey has quit (Ping timeout: 260 seconds).
20:51:15 -!- atehwa has joined.
20:52:30 <Sgeo__> curses, wrote (a+b)(c+d) in a different form not requiring parentheses againn
20:52:30 <lambdabot> Sgeo__: You have 3 new messages. '/msg lambdabot @messages' to read them.
20:53:19 <Sgeo__> Going to go back to repeatedly hitting snooze
21:02:27 <shachaf> kmc: Expr + doctest = the future
21:02:52 <shachaf> > scanl1 f [a, b, c] :: [Expr]
21:04:16 -!- EgoBot has quit (Ping timeout: 256 seconds).
21:04:41 -!- EgoBot has joined.
21:09:45 -!- Gregor has set topic: Phantom Hoovers are usually not eaten owing to their repugnant looks, as well as their viscosity and unpleasant habits. | http://codu.org/logs/_esoteric/.
21:30:08 <shachaf> The number of 7-character git hashes you need to get a 50% chance of collision.
21:31:01 * oerjan may have forgotten some of his powers of 2
21:34:46 <lambdabot> The operator `Control.Lens.Getter.&' [infixl 1] of a section
21:44:17 -!- boily has quit (Quit: Poulet!).
21:51:12 -!- monqy has quit (Ping timeout: 250 seconds).
21:51:22 -!- monqy has joined.
21:56:13 -!- oklopol has joined.
21:57:07 -!- oklofok has quit (Ping timeout: 245 seconds).
22:01:00 <fizzie> The operator Control.Lens.Getter.& is a must-have.
22:06:46 <lambdabot> The operator `GHC.Classes.&&' [infixr 3] of a section
22:06:51 <lambdabot> The operator `Control.Arrow.&&&' [infixr 3] of a section
22:12:47 <oerjan> > False&(True&(&&)&&&(&&)&uncurry(&&&))
22:13:38 <fizzie> They are all must-haves, I see.
22:16:15 <oerjan> i cannot see any way of chaining them in order.
22:16:34 <oerjan> :t (?a & ?b && ?c &&& ?d)
22:16:36 <lambdabot> Couldn't match expected type `a0 -> b0' with actual type `Bool'
22:16:36 <lambdabot> In the second argument of `(&)', namely `?b && ?c &&& ?d'
22:16:36 <lambdabot> In the expression: (?a & ?b && ?c &&& ?d)
22:16:46 <oerjan> :t (?a &&& ?b && ?c & ?d)
22:16:47 <lambdabot> Couldn't match expected type `a0 b0 c'0' with actual type `Bool'
22:16:47 <lambdabot> In the second argument of `(&&&)', namely `?b && ?c'
22:16:47 <lambdabot> In the first argument of `(&)', namely `?a &&& ?b && ?c'
22:18:13 <olsner> :t (?a & ?b (&& ?c &&& ?d))
22:18:14 <lambdabot> Couldn't match expected type `Bool'
22:18:14 <lambdabot> with actual type `a0 b0 (c0, c'0)'
22:18:14 <lambdabot> In the second argument of `(&&)', namely `?c &&& ?d'
22:20:12 <olsner> :t ((?a & ?b &&) ?c &&& ?d))
22:20:15 <olsner> :t ((?a & ?b &&) ?c &&& ?d)
22:20:16 <lambdabot> The operator `&&' [infixr 3] of a section
22:20:16 <lambdabot> must have lower precedence than that of the operand,
22:22:47 <Sgeo__> elliott, Phantom_Hoover Fiora monqy
22:28:57 -!- sebbu2 has joined.
22:28:57 -!- sebbu2 has quit (Changing host).
22:28:57 -!- sebbu2 has joined.
22:31:09 -!- sebbu has quit (Ping timeout: 260 seconds).
22:41:49 <zzo38> Is there the program to play MOD/S3M/IT/XM samples/instruments in Csound?
22:49:13 -!- Frooxius_ has changed nick to Frooxius.
22:54:06 -!- FireFly has joined.
22:59:57 -!- Bike has quit (Quit: Lost terminal).
23:00:00 -!- NihilistDandy has joined.
23:00:16 -!- Bike_ has joined.
23:02:36 -!- NihilistDandy has quit (Client Quit).
23:02:53 -!- NihilistDandy has joined.
23:03:42 -!- NihilistDandy has quit (Client Quit).
23:04:00 -!- NihilistDandy has joined.
23:04:08 -!- NihilistDandy has quit (Client Quit).
23:07:54 -!- david_werecat has joined.
23:10:08 -!- carado has quit (Ping timeout: 246 seconds).
23:26:58 -!- sebbu2 has changed nick to sebbu.
23:29:35 -!- monqy has quit (Quit: hello).
23:33:39 -!- Bike_ has changed nick to Bike.