00:02:46 -!- derdon has quit (Remote host closed the connection).
00:10:12 <monqy> 17:09:15 < Delcan> Crawl loves... it's just misunderstood
00:10:12 <monqy> 17:09:20 < Perryman> hate is love.
00:10:12 <monqy> 17:09:25 < Perryman> thats what i tell my girl anyway
00:10:12 <monqy> 17:09:30 < Perryman> did i say that out loud
00:10:12 <monqy> 17:09:33 < monqy> hi
00:10:19 <monqy> ##crawl quality. ##crawlity.
00:11:31 <ais523> monqy: was that hi as punishment?
00:12:11 <monqy> as punishment as it can be without anyone understanding it
00:12:36 <ais523> I mean, us in #esoteric understand that you mean hi as a threat
00:12:39 <ais523> does ##crawl generally know that?
00:13:39 <ais523> it's the sort of thing you threaten
00:13:49 <elliott> i know i'd feel ashamed if monqy "hi"d me :(
00:14:09 <ais523> -I could keep this up for ages-
00:14:15 <elliott> -It doesn't work if you don't punctuate.-
00:14:25 <ais523> -it doesn't work as well at least-
00:14:31 <ais523> -but punctuating is rare on IRC anyway-
00:14:41 <elliott> im getting bored of brogue im going to play crawl instead
00:15:47 <elliott> after playing a fixed-viewport game for a while crawl's moving viewport makes me dizzy again oops
00:20:37 <monqy> maybe triy a different stairs down
00:20:49 <elliott> viewport i needed help with
00:21:12 <elliott> it's not different enough to nethack for me to play it more yet
00:22:17 <elliott> coppro: but if ais523 listens to me maybe i will ;)
00:23:40 <elliott> _You sense Yredelemnul urging you to kill the trapped human.
00:25:09 <elliott> idk i was expecting yred to
00:25:31 <elliott> wait, undead can worship yred?
00:25:43 <monqy> yred hates nonliving though
00:25:56 <elliott> you can worship it but it'll hate you
00:26:06 <monqy> nonliving is distinct from undead in crawl
00:26:11 <monqy> (they're technical terms)
00:26:43 <monqy> the crawl holinesses are natural, plant, undead, nonliving, holy
00:26:56 <monqy> (holiness is also a crawl term)
00:27:25 <elliott> should i worship yred (no)
00:27:38 <monqy> yaey, undead friends. yaey, alley managemtn
00:28:20 <elliott> do you want to hear a joke
00:29:47 <monqy> wow after playing monqys-crawl crawl levels look so big and empty
00:29:55 <elliott> i want to see monqys-crawl :'(
00:30:07 <monqy> it's not ready for the public eye!!!
00:30:35 <monqy> i was almost going to fix the segfaulting today but then i remembered i had homework
00:30:43 <elliott> segfaulting is part of the monqys-crawl experience!!!
00:30:43 <monqy> "man i sure suck at not having that"
00:31:11 <elliott> monqys-crawl doesn't actually exist you've just put a segfault in some branch of standard crawl code and lied about the rest "my prediction"
00:31:51 <monqy> there are lots of segfaults !
00:32:20 <elliott> ill go berserk over monqy's LIES
00:32:34 <quintopia> elliott: what's the command to find out what my internet interfaces are
00:32:37 <ion> Hmm. Someone should make a roguelike where you have to try to win by avoiding actual segfaults.
00:32:56 <monqy> maybe i'll termcast monqys-crawl once I get some of the segfaultiness ironed out
00:33:17 <elliott> monqy: are you telling me that segfaults wouldn't make a termcast at least 67.4% more amusing
00:33:29 <elliott> im sorry but i just cant condone your heretical views
00:33:30 <quintopia> monqy: what is the betterness of monqys-crawl?
00:33:34 <monqy> elliott: segfault likelihood increases quickly as you descend
00:33:44 <elliott> monqy: that's even funnier!!!
00:33:47 <monqy> quintopia: it's not really a serious branch
00:33:58 <elliott> you know what's unserious?
00:34:10 <elliott> im going on crawl strike until i see monqys-crawl "taking a stand for freedom"
00:34:36 <monqy> let's see i if i remember how to do this
00:34:42 <elliott> i can tell you if you want
00:34:49 <elliott> ais523 what's the termcast oneliner
00:35:04 <monqy> i just forget how to type
00:35:06 <elliott> ais523: i don't need to know any more
00:35:11 <elliott> monqy: this happens to me also
00:35:12 <ais523> it'll be hilarious if this works ;)
00:35:54 <ais523> elliott: it may have just been the HackEgo startup bug
00:35:56 <elliott> its so funny i cant survive help :(
00:36:10 <HackEgo> 2010-08-29.txt:22:22:02: <ais523> script -f >( cat ./ratry_login - | nc -q5 noway.ratry.ru 31337 > /dev/null ) "$@"
00:36:39 * ais523 goes back to feeling smug
00:37:10 <ais523> yay, it brought you back from the ded
00:38:11 <ion> Hmm. xw on CDO seems to be running a bot of some kind, too.
00:38:18 <ion> And now he died.
00:38:41 <ion> And now he died again.
00:39:35 <monqy> xw is n7's autorobin fork
00:39:44 <elliott> this thing is better than monqy's bot
00:40:08 <monqy> i got termcaste runeing
00:40:29 <ion> monqy: On what server?
00:42:09 <monqy> what should i play
00:42:38 <monqy> i don't know if my modifications touched ranged combat, so it might megasuck
00:42:39 <elliott> monqy: play something that won't die instantly or i guess nothing will since everything got nerfed
00:42:45 <elliott> just play what you've been playing or w/e
00:42:52 <ais523> let's pick a combo that's bad but not horrendously bad
00:42:57 <ais523> unless they removed AM altogether
00:43:04 <monqy> AM got removed and then unremoved
00:43:24 <ais523> that's not so much bad as completely missing the point :)
00:43:28 <elliott> monqy: just play (a) whatever you played when testing or (b) muck or something
00:43:44 <monqy> but i've only been playing op things
00:44:03 <elliott> ~see more of the monqys-crawl experience~
00:44:17 <monqy> ok i'll cycle through everything you suggest that isn't just "whatever you tried"
00:44:26 <ais523> what about just viable random?
00:44:29 <monqy> so first, a feck that takes chei first chance it gets
00:44:50 <elliott> at least xom's teleportation isn't very scary when the maps are this small
00:44:59 <ais523> hmm, the felid is a Unicode character that isn't in dejavu sans mono
00:45:00 <monqy> oh no is it not working
00:45:07 <elliott> it works apart from dark grey that turns out black because
00:45:08 <ais523> gnome-terminal can render it fine
00:45:23 <monqy> ais523: it's not the normal fe glyph
00:45:29 <elliott> monqy plays as a flashing rainbow square
00:45:30 <ion> elliott: The politically correct term is African-American.
00:45:34 <ais523> oh, it's /meant/ to look like that
00:45:50 <ais523> elliott: what about looking at it through screen?
00:46:04 <ais523> what's with the messages, too?
00:46:09 <elliott> ais523: I know for a fact it works if you set TERM=xterm, so why bother?
00:46:30 <elliott> octopode abyssal knight of chei
00:46:32 <monqy> oh right it was uh seam wasn't it
00:46:48 <ais523> trying to remember where I remember that combo from, it might have been a nemelex choice
00:47:17 <monqy> what weapon, i mean
00:47:21 <monqy> i dunno if it was neme choice
00:47:26 <ais523> whatever you have the best apt for, I guess
00:48:12 <ais523> low hitpoint warning already?
00:48:13 <elliott> go with something op already
00:48:23 <ais523> you're still in sight of all three of the entry squares :)
00:48:46 <monqy> i tested with a LOT OF THINGS
00:48:53 <monqy> and yes im totally ashamed because of how op it is
00:49:16 <elliott> we'll accept this convenient fiction and ignore the fact that you obviously always play that combo when offline
00:49:33 <elliott> does monqys-crawl have the randomised languages and colours and turncount-type-changing built in
00:49:41 <monqy> no that's my .crawlrc
00:50:19 <ais523> elliott: he dropped to five, then back to six the next turn
00:50:19 <ion> Someone shoulde make a Britishe localisation for Crawle.
00:50:28 <ais523> so I guess it has some sort of scaling, even more extreme than Sprint's
00:50:36 <elliott> this looks more fun than crawl already
00:50:37 <ais523> ion: Crawl is in British English already, actually
00:50:39 <ais523> well, Australian English
00:50:45 <ais523> or was it New Zealandish?
00:50:46 <ion> ais523: But not Britishe Englishe
00:50:47 <elliott> monqy: you realise you're going to die so hard if you remove upstairs
00:50:51 <ais523> but much closer to British than American
00:51:10 <ais523> monqycrawl reminds me of Sprint a bit
00:51:11 <ion> Wait. Why do you have so much ********?
00:51:14 <monqy> elliott: i forgot i had a conduct where i didnt take upstairs
00:51:20 <ais523> ion: it's obviously being scaled
00:51:21 <elliott> monqy: tell us about why you started with lots of piety :'(
00:51:23 <monqy> ais523: it's kind of like zigsprint
00:51:33 <elliott> monqy: also that's a bad conduct sheeshe!!!
00:51:36 <monqy> elliott: nope, but i have sprint piety gainsenabled
00:51:40 <ais523> monqy: actually, that seems like a good game mode
00:51:57 <ais523> start the player in a zig, put Sprint-like exp gains
00:52:10 <ais523> increase the amount of loot somewhat
00:52:15 <monqy> ais523: zig sprint is a thing, it's not really a zig but
00:52:28 <ais523> oh, they made a Sprint map that's like a zig?
00:52:46 <ais523> hmm, perhaps I'll play that, I prefer Sprint to regular Crawl quite a bit
00:52:57 <elliott> monqy: hmm, are you sure just making maps the full size of the viewport was too small??
00:53:12 <elliott> it looks like they're too thin and too tall
00:53:21 <monqy> elliott: these are square
00:53:29 <monqy> elliott: it's the shape of your font that makes it seem like that
00:53:51 <elliott> i think the piety scaling might be a bit extreme
00:54:06 <monqy> getting rid of corpses means getting rid of corpse sacrifices
00:54:21 <monqy> i'll see if it's extreme after then
00:54:23 <elliott> i think all the monsters are
00:54:27 <monqy> and i turned on the sprint stuff just for fun
00:54:34 <elliott> the kind of packs you get when going downstairs
00:54:34 <monqy> some of them are megabuff
00:54:39 <monqy> they just don't show up yet
00:54:41 <elliott> will probably annihilate a low-level player with th enerfing
00:54:53 <monqy> it doesn't help that I have 10ac. ac is op in monqys-crawl
00:55:52 <elliott> this actually looks way better than crawl
00:57:27 <elliott> monqy: you went upstairs!!!
00:57:35 <monqy> i forgot about hte conduct
00:57:39 <monqy> i also forgot about the no-resting conduct
00:57:47 <monqy> it's just wired into me from so much crawl play
00:57:52 <elliott> monqy: no-resting is a bit unreasonable if you get poisoned or slow or whatever
00:58:02 <elliott> you're only removing HP regeneration, right?
00:58:06 <monqy> i have infinite piety
00:58:13 <elliott> magic/pois/etc. regeneration seem reasonable to keep
01:00:36 <elliott> monqy: yuore skills are default
01:01:04 <monqy> oops i forgot to go upstairs
01:02:20 <elliott> do something about that hp
01:02:54 <monqy> i get careless when its too easy and then there's a huge difficulty spike
01:02:57 <monqy> monqys-crawl is like that
01:03:06 <ion> crawl is like that
01:03:08 <monqy> now what to play!!
01:03:18 <elliott> monqy: octopode chaos knight of chei
01:03:30 <elliott> recipe for humour: <bad race> chaos knight of chei
01:03:39 <monqy> op can't armoure so it might be a bit "difficult"
01:03:51 <elliott> its ok armour is monqys-crawl's version of mephitic cloud
01:04:00 <elliott> "dont use it" - monqys-crawl's version of monqy
01:05:33 <elliott> ais523: does this look fun to you i cant tell if im just hallucinating
01:05:46 <ais523> elliott: I'm playing zigsprint instead
01:06:13 <elliott> ais523: bah, how can i choose between you playing crawl and monqy playing monqys-crawl?
01:06:53 <elliott> monqy: you forgot to remove the monster colour-coding
01:07:06 <monqy> it's low on the priority list
01:07:54 <elliott> monqy: see that's what i meant by
01:07:57 <monqy> opck "decidedly bad"
01:07:58 <elliott> dying when going downstairs post-un-nerfing
01:08:05 <elliott> (the no armour simulates the un-nerfing)
01:08:59 <elliott> monqy: do a same letter combo
01:09:47 <ion> Err, i mean #
01:09:55 <elliott> actually asks for combination?
01:10:21 <elliott> i think it's pronounced like the first word of Beowulf
01:10:46 <ion> It’s the name of a Chinese network hardware manufacturer.
01:11:22 <elliott> it sounds like it accidentally picked you an ok combination
01:11:59 <ion> Oh, i had totally forgotten DDs sense surroundinges.
01:12:53 <ais523> "the two-headed ogre shouts in stereo!"
01:12:59 <elliott> monqy: im dibs on setting up the first monqys-crawl server
01:13:21 <ion> p eliot is a jerk's ghost
01:13:34 -!- elliott has changed nick to jerks_ghost.
01:13:42 -!- jerks_ghost has changed nick to jerks-ghost.
01:13:56 <jerks-ghost> monqy: will you implement monsters following you downstairs please
01:14:31 <monqy> "forgot to updsatatairs"
01:14:59 <jerks-ghost> does that also mean you'll remove things teleporting downstairs when they should by rights take multiple turns to get down just because they're right next to you
01:16:05 <monqy> everyone got the same changes applid
01:16:23 <jerks-ghost> ais523: btw you can reply to my messages with : hope this helps
01:16:38 <ais523> just didn't feel like replying earlier
01:17:02 <ais523> bleh, didn't quite last long enough
01:17:08 <ais523> but it turned out not to matter
01:17:39 <jerks-ghost> monqy: at least the monsters shield you from the centaur
01:19:26 -!- lexande has left ("Leaving").
01:19:39 -!- azaq23 has quit (Quit: Leaving.).
01:20:01 <jerks-ghost> monqy: did you asnwer my regen question i forget
01:20:05 <jerks-ghost> ais523: you're missing out on monqys-crawl!!!
01:20:38 <jerks-ghost> monqy: you're removing hp gen; i'm assuming you'll leave magic/pois/slow/exh/etc. regen?
01:20:59 <monqy> yeah, except for magic
01:21:12 <monqy> hp/magic regen will be handled in similar ways
01:21:37 <jerks-ghost> you could make food regen hp for everyone but "woops no corpsen"
01:21:55 <jerks-ghost> (maybe leave corpsen and just make them regen hp instead of the hunger mechanic???)
01:23:05 <monqy> would rather just atuomatically just regain hp from kile
01:23:24 <jerks-ghost> that doesn't help if you need hp regenerating later on after kille
01:23:44 <monqy> dd of makh sure doesn't mind
01:23:56 <jerks-ghost> so that killing increases max hp instead of hp if you're at max
01:24:24 <jerks-ghost> dunno if that would have any downsides or whatever "i d k"??? it's "just an idea"
01:30:20 <monqy> in monqys-crawl, levels take a long time to generate because of how small they are
01:30:49 <monqy> could be stuck vetoing everyyyyyyything
01:31:05 <monqy> or it may have actually hung itself rip
01:31:16 <monqy> i'll put it out of its misery and try again
01:31:54 <jerks-ghost> 02:23 <jerks-ghost> you could do hp on kill + remove max hp
01:31:54 <jerks-ghost> 02:23 <monqy> dd of makh sure doesn't mind
01:31:54 <jerks-ghost> 02:24 <jerks-ghost> so that killing increases max hp instead of hp if you're at max
01:32:00 <jerks-ghost> 02:24 <jerks-ghost> (or maybe just max hp???)
01:32:02 <jerks-ghost> 02:24 <jerks-ghost> dunno if that would have any downsides or whatever "i d k"??? it's "just an idea"
01:32:44 <monqy> dunno how that would work out
01:32:49 <ais523> well, that explains why people hate torment, at least
01:33:20 <ais523> tormented repeatedly and then poked to death
01:33:25 <ais523> I couldn't out-regen them
01:33:34 <ais523> and they also killed the berserking trolls really quickly
01:35:20 <jerks-ghost> monqy: well the idea would be: remove increasing max hp on level up
01:35:42 -!- quintopia has quit (Remote host closed the connection).
01:35:45 <monqy> oh nuts slime critter
01:37:08 <monqy> that one got to enourmouse, ie 88dam
01:37:15 <monqy> with maxdam that's uh
01:37:22 <monqy> but then i subtract ac so
01:37:24 <monqy> less than that but
01:37:27 <monqy> "you get the idea"
01:37:33 <ais523> jerks-ghost: they're designed to punish fighting in corridors
01:37:46 <ais523> and interpret everywhere as a corridor in a level as confined as a monqys-crawl level
01:38:04 <jerks-ghost> monqys-crawl must be gr8 for characters that are claustrophobic
01:40:42 <monqy> they affect everything
01:40:49 <monqy> probably make branch endings crash
01:42:43 <monqy> ok fine i'll go to the termplel
01:43:34 <monqy> i'll continue playing my testing gam
01:45:18 <monqy> oh right in this game I have ignite blood and a vampiric battleaxe
01:46:30 <monqy> you gain levels more often
01:46:57 <monqy> i told you it's not presentable yet :|
01:48:46 <jerks-ghost> monqy: have you not been using wizmode for testing?
01:48:54 <monqy> wizmode is for cheaters
01:53:57 <monqy> did you see the action
01:54:14 <monqy> every square had a monster
01:54:28 <jerks-ghost> maybe you should make there only be one downstair wit
01:54:33 <monqy> i spawned next to a giant spore, which killed everything except me (barely), a hill giant(which it confused) and a centaur (which was outside the blast radious)
01:54:42 <monqy> jerks-ghost: that's on the todo list
01:55:47 <monqy> i should get back to my
01:58:04 <jerks-ghost> haven't you ever heard of "commit early commit often"!!!!!!!!
01:58:09 <monqy> should i commit what i have, segfaults and all
02:00:09 <monqy> normally i try to commit early and often!! monqys-crawl is just an exception
02:00:25 <monqy> wherein i just wanted to play around
02:00:50 <jerks-ghost> do a bunch of changes and then use git commit -p
02:00:55 <jerks-ghost> sometimes you have to stash half-way through
02:01:14 <jerks-ghost> once i had a commit session that lasted a few hours on mchost
02:01:25 <monqy> so how do i do this
02:01:45 <monqy> just one big commit for now?
02:01:58 <jerks-ghost> that's OK; it's when you have more complex requirements that it gets
02:02:03 <jerks-ghost> like if you have changes that both touched the same place
02:02:13 <jerks-ghost> then it's stash and restash and apply and modify and uugh time
02:04:58 <monqy> dind't you see it already before? did i accidentally make another
02:05:26 <kmc> i use git add -p
02:05:51 <kmc> life's a bitch
02:07:04 <jerks-ghost> monqy: alternatively: just do it all in one big commit if you would instead give up
02:07:45 <jerks-ghost> (shouldn't you be doing it in a branch though) oh wait there's no commits
02:07:53 <jerks-ghost> (don't do it in another branch its boreing)
02:08:10 <monqy> ok i made the comites
02:08:27 <monqy> doing what in a branch
02:08:53 <jerks-ghost> "technically" git "accepted wisdom" would have it in a branch if you had any pretenses about merging it back into crawl itself
02:09:13 <monqy> i don't think they'd accept my changes
02:09:27 <jerks-ghost> also uhhhh does it put stuff in my home directory or whatever
02:09:35 <monqy> i don't know how to push!!!
02:11:59 <monqy> come on you can do better than that, ion
02:12:48 <ion> “that” –she
02:17:13 <monqy> "one waty to foidn out"
02:17:25 <monqy> yeah there it goes
02:17:44 <monqy> oh no i just remembered i forgot to make the code look presentable too
02:17:58 <jerks-ghost> http://gitorious.org/~monqy/crawl/monqys-crawl
02:18:23 <jerks-ghost> monqy: your commit messages are lowercase :'(
02:18:34 <monqy> was i suposed to upercase them
02:18:37 <monqy> did i do soemthing wrong
02:18:57 <jerks-ghost> the crawle repo uses "Blah blah blah blah." (but "Blah blah blah blah" is the git standard (but crawl adds .s too))
02:19:08 <jerks-ghost> also wouldn't it have been simpler to just
02:19:10 <monqy> i never should have comited
02:19:22 <monqy> it was my first try
02:19:32 <monqy> it tries to load sprint maps
02:19:49 <monqy> it doesnt have the sprint map thingy set so
02:20:10 <jerks-ghost> defender->heal(1 + random2avg(7, 2), true); // heh heh
02:20:10 <jerks-ghost> 2159 defender->heal(1 + 7, true); // heh heh
02:20:21 <ion> Transmogrify the tribbles (title without a dot in the end)
02:20:23 <ion> (empty line)
02:20:31 <monqy> i expect the optomizer to catch that
02:20:40 <ion> An optional longer explanation for the commit, potentially multiple lines/paragraphs.
02:20:43 <monqy> i think that's where distortion heals blink frogs
02:21:16 <monqy> what did i do there i forget
02:21:21 <monqy> im sure i had a good reason
02:22:12 <jerks-ghost> if i get it working maybe i'll remove identification or something "getting my toes wet in the water that is the crawl codebase"
02:22:21 <monqy> oh right i remember why i did that thing
02:22:48 <monqy> in crawl when you lrd certain monsters it either instakills them or adds 2 damage
02:23:05 <monqy> i got rid of the instakill and changed add 2 to multipl;y by 2
02:23:09 <monqy> i think what i did was fair!!
02:23:25 <jerks-ghost> unrandomisde parts that made every monster either terrible or too good
02:23:49 <jerks-ghost> (maybe you should have split the lrd part into another commit!!!)
02:27:06 <monqy> im not going to remove unrandomisation but i'll probably twaeak it and also other monster-related things
02:27:08 <jerks-ghost> ls: /Developer/SDKs: No such file or directory
02:27:08 <jerks-ghost> Makefile:412: *** You do not seem to have any Mac OS X SDKs installed! This build is doomed to fail. Stop.
02:27:16 <jerks-ghost> monqy: what do you have against random :'(
02:27:27 <jerks-ghost> its kind of boring if fights are predetermined!
02:27:28 <monqy> random in crawl is pretty annoying
02:28:20 <jerks-ghost> monqy: will you accept patches to make it work with xcode 4.something
02:29:09 <monqy> but really those should be going to the normal crawl gus
02:30:07 <jerks-ghost> /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs
02:31:59 <jerks-ghost> Makefile:447: *** The Mac OS X /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs: SDK seems missing. Stop.
02:35:01 <jerks-ghost> monqy: how does it decide whether to build tiles or not
02:35:12 <monqy> if you say make TILES=y it tileses
02:35:21 <monqy> maybe make tiles makes tiles too
02:35:24 <monqy> idk ive never made tiles
02:35:30 <jerks-ghost> Cloning into 'crawl-ref/source/contrib/pcre'...
02:35:30 <jerks-ghost> fatal: The remote end hung up unexpectedly
02:35:30 <jerks-ghost> Clone of 'git://gitorious.org/crawl/crawl-pcre.git' into submodule path 'crawl-ref/source/contrib/pcre' failed
02:38:53 * ion realized a brain/computer interface could be called a silicon implant.
02:41:14 <monqy> you know what's fun
02:41:22 <monqy> and then everything recompiles
02:43:09 <shachaf> is this "ur permenant nikc"
02:44:27 <jerks-ghost> ./dat/des/altar/altar.des:279: Map too large - 20x7 (max 16x16)
02:44:34 <monqy> that goes in source/something/somethingesel i think
02:44:40 <monqy> ar you termckasting
02:44:50 <jerks-ghost> ./dat/des/altar/altar.des:279: Map too large - 20x7 (max 16x16)
02:45:25 <monqy> it doesnt say so!!
02:46:58 <monqy> it works for me yes
02:47:40 <jerks-ghost> yeah make clean && make and maybe then it'll break? because
02:47:52 <monqy> ok i made clean and am mak
02:52:36 <jerks-ghost> you should add a new god called Maktakstim
02:55:41 <monqy> ??? ? ? ? ? ? ? ? ? ??
02:58:34 <jerks-ghost> monqy: i think it might have built with tiles??
03:01:00 <jerks-ghost> tool/tile.cc:9:25: warning: SDL_image.h: No such file or directory
03:01:00 <jerks-ghost> tool/tile.cc: In member function ‘bool tile::load(const std::string&)’:
03:01:00 <jerks-ghost> tool/tile.cc:381: error: ‘IMG_Load’ was not declared in this scope
03:01:10 <jerks-ghost> make -C rltiles all ARCH=x86_64-apple-darwin11.3.0 TILES=n
03:02:03 <monqy> did oyu do the git submodule --init thing but really you shouldn--it's somewhere--'t need to build rltiles if you're not tiles ????? maybe crawls build just dum
03:03:04 <monqy> http://crawl.develz.org/wordpress/downloads try this i think it has some documentat
03:03:35 <monqy> oh i forget where that is
03:03:46 <monqy> maybe it's in like
03:03:48 <jerks-ghost> # TILES -- set to anything to enable tiles build
03:04:17 <jerks-ghost> @printf " * If you experience any problems building Crawl, please take a second look\\n"\
03:04:17 <jerks-ghost> " * at INSTALL.txt: the solution to your problem just might be in there!\\n"
03:05:25 <jerks-ghost> Mac builds use the Unix build process described above, but require you
03:05:25 <jerks-ghost> to add 'APPLE_GCC=y' to the 'make' command-line. In addition, to build the
03:05:25 <jerks-ghost> graphical version of Crawl, you must add 'NO_PKGCONFIG=y' and
03:06:34 <monqy> im make from the fresh clone now
03:10:47 <monqy> i dont want to support it can i unfork it
03:12:06 <shachaf> jerks-ghost: why are you called "jerks-ghost", jerks-ghost
03:14:36 <monqy> ok FINALLY it break
03:15:38 <jerks-ghost> diff the two directories but i'd do that first
03:18:20 <jerks-ghost> monqy: if diff says theres no differences ill
03:19:47 -!- TeruFSX2 has changed nick to TeruFSX.
03:20:58 <monqy> the onlllllly thing i can find that diffffers that i think might make adifference is
03:21:35 <monqy> sure enough i remvo and it break
03:21:51 <monqy> but now what went wrong !!
03:21:55 <jerks-ghost> 03:45 <jerks-ghost> ./dat/des/altar/altar.des:279: Map too large - 20x7 (max 16x16)
03:22:05 <ion> I love assumptions like this in math questions: “The person is entirely made of water, which weighs 1kg/L.”
03:22:07 <jerks-ghost> like you divided by two once too many or something!!!
03:22:22 <jerks-ghost> at least there's not many places it can be
03:23:25 <jerks-ghost> http://gitorious.org/~monqy/crawl/monqys-crawl/commit/934ebc7df7933d82e37bf959b5e62c749a2efffc
03:23:29 <jerks-ghost> or http://gitorious.org/~monqy/crawl/monqys-crawl/commit/fc9667dff88582d60a78de5b4bc36074c69dcb98 (<-- less likely)
03:26:06 <monqy> i'll think about it when i have less homwork
03:26:16 <shachaf> monqy: hi does elliott have me on /ignore
03:27:05 <shachaf> what is that thing at the end of your sentence
03:27:27 <monqy> oh i have to go "oops"
03:38:47 -!- oklopol has joined.
03:44:34 -!- quintopia has joined.
04:01:23 -!- quintopia has quit (Quit: leaving).
04:01:36 -!- jerks-ghost has quit (Remote host closed the connection).
04:03:23 -!- quintopia has joined.
04:16:03 -!- zzo38 has joined.
04:19:53 -!- asiekierka has joined.
04:29:02 -!- madbr has joined.
04:32:42 -!- zzo38 has set topic: It looks like a pipe, but it is not. Anemometers don't fire bullets. Not even the explosive ones! Or I don't think they do ... if they do, meteorologists are more hardcore than I thought possible. | http://codu.org/logs/_esoteric/.
04:58:08 -!- calamari has left ("Leaving").
05:08:46 -!- TodPunk has quit (Quit: This is me, signing off. Probably rebooting or something.).
06:01:12 -!- pikhq_ has joined.
06:01:28 <pikhq_> Ceci n'est pas une "ceci n'est pas une pipe".
06:02:33 <pikhq_> The French musta thought English should be more like French.
06:03:46 -!- pikhq has quit (Ping timeout: 276 seconds).
06:04:38 -!- asiekierka has quit (Quit: Wychodzi).
06:12:00 <madbr> latin meets german did it
06:15:13 <zzo38> Who is Arthur Mometer?
06:15:58 -!- zzo38 has set topic: It looks like a pipe, but it is not. Astronomers don't fire bullets. Not even the explosive ones! Or I don't think they do ... if they do, meteors are even more hardcore than I thought possible. | http://codu.org/logs/_esoteric/.
06:26:21 <shachaf> @tell elliott By the way, the "unordered tuple" thing is just n-choose-k.
06:36:34 <shachaf> @tell elliott OK, maybe not quite.
06:38:27 <monqy> unordered tuple? multiset?
06:40:38 <shachaf> @tell elliott (n + k - 1) choose (k - 1)?
06:40:47 <shachaf> monqy: Of a particular size.
06:41:25 <monqy> I've heard that (n + k - 1) choose (k - 1) thing called "multichoose" too
06:42:19 <kmc> shachaf, do you know why -fstack-protector would have a much higher relative performance impact on an Atom N270 than a Core 2 Duo?
06:42:28 <shachaf> monqy: I haven't heard that.
06:43:12 <kmc> that's about all i have
06:43:40 <kmc> also i used GCC 4.4 and 4.6 respectively, so that might be it
06:43:51 <kmc> shachaf: ran a Mosh benchmark program a few times
06:44:16 <shachaf> You should try profiling it.
06:44:23 <shachaf> No idea whether that'll yield any insights.
06:44:42 <kmc> function-level profiling probably will not
06:44:50 <kmc> since each function has its own copy of the stack-check code
06:45:16 <kmc> ...that could be why, actually.
06:45:21 <kmc> i-cache blowout
06:45:34 <shachaf> Right, instruction-level profiling.
06:45:50 <kmc> i don't know how to do that, and i don't know if the Atom has the right CPU features
06:45:58 <shachaf> That seems like a reasonable theory. How much code does it actually generate?
06:46:19 <shachaf> The Linux program "perf" makes at least some kinds of instruction-level profiling pretty easy, I think.
06:46:40 <kmc> i also don't have the Atom machine handy anymore :/
06:47:01 <shachaf> You should use perf anyway!
06:48:05 <kmc> i was also thinking it could be bad branch prediction
06:48:12 <kmc> basically i can only speculate
06:49:02 <shachaf> kmc: apt-get install linux-base # maybe?; perf record ./blah; perf annotate
06:55:01 <kmc> shachaf: on amd64 it adds at least 25 bytes to each function
06:55:39 <shachaf> -fstack-protector is for people who make mistakes. Just don't make mistakes.
07:01:35 <itidus21> i write all my source code compressed to save keystrokes
07:03:21 <Deewiant> You might as well write the binary directly then because making changes to any part of that code will entail rewriting most of it
07:04:39 <itidus21> hmm.. i write the binary compressed..
07:07:37 -!- Phantom_Hoover has joined.
07:09:12 <zzo38> If you write the binary directly, it makes the program less portable and less adjustable, I would think. Still, for sufficiently small and simple programs it may be useful.
07:09:35 <itidus21> it requires a lot of messing around with the Alt key though
07:09:45 <zzo38> itidus21: Is this in ZIP format?
07:10:17 <itidus21> its not quite genuine though.. i faked some parts of it
07:10:40 <itidus21> yes.. for all those people who know that zip files start with a PK when viewed as ascii :P
07:10:47 <zzo38> The PK followed by CTRL+C and CTRL+D is what made me realize; these are the first four bytes of any ZIP file.
07:11:11 <itidus21> the rest of the code is faked with spaces etc
07:11:58 <itidus21> humm.. im just being silly of course
07:13:50 <itidus21> i was inspired by shachaf's comment about not making mistakes
07:14:30 <kmc> itidus21: you're a Real Hacker
07:16:15 <kmc> shachaf: the code gets about 12% bigger with stack-smashing protection
07:16:18 <kmc> (on 32-bit now)
07:16:21 <itidus21> i'm not trying to hide the fact i took a zipfile(instead of traditionally open compression format) and opened it in Windows wordpad and pasted the first few chars
07:17:11 <itidus21> or the fact that producing source code in compressed format would be computationally more expensive
07:17:55 <shachaf> kmc: That's a bigger increase than I thought.
07:18:00 <kmc> i guess that... could explain a 20% performance hit?
07:18:10 <shachaf> You should profile it. :-)
07:18:19 <kmc> i don't know if I care enough to dig deep on this
07:18:24 <kmc> but it would be cool to learn how to do that
07:18:30 <kmc> but i have so many other things i could do
07:18:40 <kmc> i'm going to blog post it and hope someone else tells me the answer
07:18:41 <itidus21> but i think it would be cool in a sci fi kind of way to be able to type data in compressed format
07:19:09 <kmc> http://en.wikipedia.org/wiki/Shorthand
07:19:16 <shachaf> kmc: I would measure it but I don't have a program or an Atom CPU handy.
07:19:24 <kmc> http://en.wikipedia.org/wiki/Stenotype
07:19:36 <shachaf> But perf etc. can measure icache misses pretty easily, I think.
07:19:39 <kmc> court stenographers are the true cyberpunk heros
07:19:56 <kmc> the program is in Mosh; i can send you a link
07:20:07 <shachaf> I heard of that program once.
07:20:46 <kmc> hm oh I do have an Atom machine I forgot about
07:20:53 <kmc> but it's running a weird Linux distro
07:21:10 <shachaf> Doesn't make a big difference, I think.
07:21:11 <kmc> also it's not the same one I did the original measurements on
07:21:26 <shachaf> If you can't reproduce the original measurements on, even better!
07:21:28 <kmc> and it's not turned on
07:21:35 <kmc> so I'd have to walk all the way upstairs
07:21:45 <kmc> it's not even plugged in, man!
07:22:22 <kmc> the forces of curiosity and laziness are doing battle in my head
07:23:03 <itidus21> unpossible = simpsons reference 1. maybe. airconditioner fighting the heater in the rental car is the second.
07:23:34 <shachaf> kmc: You should "learn" to do simple instruction-level profiling.
07:23:43 <shachaf> And perf has this fancy TUI.
07:24:42 <kmc> ok fine i'll at least see if perf can get cache misses on atom
07:25:06 <kmc> i'll use the same Atom machine I used before (but not the same OS install)
07:25:12 <kmc> which is not ideal because it's under load
07:25:59 <kmc> oh, wait, no I don't have that machine at all
07:26:04 <kmc> it's in new york city
07:26:16 <kmc> too many netbooks
07:26:40 <kmc> the one i have has a 900 MHz Celeron M, not an Atom
07:29:58 <kmc> let the internet sort this one out
07:34:08 <kmc> shachaf: it's 12% bigger because we stack protect *every* function
07:34:18 <shachaf> Just protect the ones with bugs in them.
07:34:22 <kmc> even the ones with no buffers
07:34:34 <kmc> shachaf: i might have to upgrade gcc to get that feature
07:35:21 <shachaf> kmc: Did you know Either (Nat,Nat) Nat is an even type, but (Nat,Nat) isn't?
07:35:47 <pikhq_> shachaf: Better: annotate the ones with bugs in them.
07:36:00 <pikhq_> Then, you can stick every single statement in its own function!
07:36:37 <shachaf> @tell elliott Did you know Either (Nat,Nat) Nat is an even type, but (Nat,Nat) isn't? Becuase of the extra diagonal.
07:36:52 <kmc> what's an even type?
07:37:14 <shachaf> @tell elliott Also it's actually just (n + k - 1) choose k, for the sensible value of k
07:37:21 <shachaf> kmc: A type you can divide by 2.
07:37:31 <shachaf> For example Bool is even but Maybe Bool isn't.
07:37:45 <kmc> i didn't read that paper on quotient types or whatever
07:38:07 <shachaf> Just #haskell-blah gossip with dolio.
07:38:27 <shachaf> kmc: It makes some sense for (n^2 + n) to be even for any value of n.
07:38:35 <shachaf> 00:33 < dolio> The reason is, I guess, that Left (n, n) is the same as Right n, and Left (m, n) is the same as Left (n, m), so there are two of every thing you want to make into one element.
07:38:38 <shachaf> 00:34 < dolio> You can't just divide Nat * Nat by 2, though, because there is only one of each of the diagonal.
07:38:42 <shachaf> 00:34 < dolio> You have to add an extra diagonal to it.
07:41:06 <shachaf> I know what even finite types are but even infinite types are a bit weirder.
07:41:43 -!- madbr has quit (Quit: Radiateur).
07:41:58 <shachaf> I'm not sure either makes sense.
07:46:32 <zzo38> Tell me what half of (Either (Nat,Nat) Nat) is?
07:47:12 <shachaf> An unordered pair of Nats.
07:48:31 <shachaf> Well, in general for a finite type x, (x^2+x)/2 is an unordered pair of xs.
07:48:47 <shachaf> zzo38: See dolio's thing above for a mapping.
07:49:06 <shachaf> I'm not sure that it's valid for infinite types, but take it for arbitrary finite types instead. :-)
07:55:59 -!- impomatic has quit (Ping timeout: 245 seconds).
08:02:06 <cheater_> doesn't even type just mean there's a cycle of length two through which you can create a quotient
08:02:55 <cheater_> because if yes then it's fairly obvious anything with a binary choice in series with other type features will be even
08:04:53 <cheater_> i really like the fact that this channel is talking about haskell, even though it's not really an esolang
08:07:01 <kmc> OR IS IT??????
08:07:12 <kmc> esolang is in the eye of the beholder
08:07:27 <kmc> what do you use it for?
08:07:35 <kmc> did you tell #haskell? did they cream their pants?
08:07:41 <cheater_> i told you the last thing i used it for
08:07:55 <kmc> i don't remember
08:08:01 <cheater_> except that was a lie because your pants were probably off already
08:08:09 <kmc> there's really no reason to wear pants
08:08:59 <kmc> yeah when i have guests over, i have to put pants on
08:09:02 <kmc> sometimes they come off though
08:09:35 <cheater_> like when the mormons come over for a bible talk?
08:10:21 <cheater_> highlight of yesterday's visit to hi-fi tradeshow: audiophile breakers
08:15:24 <kmc> circuit breakers?
08:15:43 <cheater_> the stuff that's in your breaker box
08:15:43 <zzo38> data Zero; data UnordNatTag :: * -> * where { Z :: UnordNatTag Zero; S :: UnordNatTag x -> UnordNatTag (Maybe x); }; type UnordNatPair = DSum UnordNatTag; Is this a correct way of an unordered pair of natural numbers?
08:15:58 <kmc> that's a hell of an industry
08:16:02 <kmc> i should get into it
08:16:07 <kmc> do they have special food for audiophiles yet?
08:16:11 <kmc> they have special food for gamers
08:16:22 -!- ais523 has quit (Ping timeout: 276 seconds).
08:16:25 <cheater_> i've seen special food for geeks
08:16:52 <cheater_> http://en.wikipedia.org/wiki/Ubuntu_Cola
08:16:56 <kmc> http://www.gamergrub.com/
08:17:17 <kmc> looks like it's not really related to the OS?
08:17:28 <cheater_> it's still an african scam though
08:18:04 <kmc> 419 scam westerners every day
08:18:06 <pikhq_> You can do much better than audiophile stuff.
08:18:09 <cheater_> "boost your game with gamer grub"
08:18:25 <cheater_> pikhq_: oh yeah, that's really good too
08:18:33 <pikhq_> Actually, combine them.
08:18:39 <pikhq_> Homeopathic audiophile equipment.
08:19:07 <cheater_> this western electric stuff, i've seen a recreation of it
08:19:15 <cheater_> just one tweeter cost 9000 euro
08:19:19 <pikhq_> "Each box contains a 50X solution of Sennheiser headphones, making your audio experience the best it can be!"
08:19:29 <cheater_> and it's like an electromagnet and a small horn and that's it
08:19:43 <cheater_> nothing you couldn't do on a rotary mill
08:21:28 <zzo38> Do you know answer for my question above?
08:23:13 <cheater_> not i, perhaps try to explain your reasoning?
08:24:10 -!- Phantom_Hoover has quit (Ping timeout: 276 seconds).
08:24:55 <zzo38> I am using the DSum type from "dependent-sum" package; it is defined as follows: data DSum tag = forall a. !(tag a) :=> a
08:28:12 <zzo38> This way may be better: data UnordNatTag :: * -> * where { Z :: UnordNatTag (Maybe Zero); S :: UnordNatTag (Maybe x) -> UnordNatTag (Maybe (Maybe x)); };
08:30:09 <zzo38> It does seem to work; it is not really unordered but the larger number must come first so it acts like just two natural numbers put in one box where you don't know the order
08:36:53 -!- Phantom_Hoover has joined.
08:38:33 -!- Taneb has joined.
08:38:57 <zzo38> getUNP :: UnordNatPair -> (Integer, Integer); getUNP (Z :=> Nothing) = (0, 0); getUNP (S x :=> Nothing) = (\(a, b) -> (succ a, b)) $ getUNP (x :=> Nothing); getUNP (S x :=> Just y) = (\(a, b) -> (succ a, succ b)) $ getUNP (x :=> y);
08:39:47 <zzo38> putUNP :: (Integer, Integer) -> UnordNatPair; putUNP (0, 0) = (Z :=> Nothing); putUNP (x, y) | x < 0 || y < 0 = error "Cannot use negative numbers here"; putUNP (x, y) | x < y = putUNP (y, x); putUNP (x, 0) = case (putUNP (pred x, 0)) of { (Z :=> Nothing) -> (S Z :=> Nothing); (S a :=> Nothing) -> (S (S a) :=> Nothing); }; putUNP (x, y) = case (putUNP (pred x, pred y)) of { (Z :=> b) -> (S Z :=> Just b); (S a :=> b) -> (S (S a) :=> Just b); };
08:49:27 -!- Taneb has quit (Read error: Connection reset by peer).
08:49:53 -!- Taneb has joined.
08:49:57 -!- monqy has quit (Quit: hello).
08:58:44 -!- shubshub has joined.
09:01:19 <zzo38> My brother called mahjong players that called sequences a lot, a "chii"ter
09:01:39 -!- Taneb has quit (Quit: DOING OTHER THINGS).
09:03:47 -!- azaq23 has joined.
09:03:58 -!- azaq23 has quit (Max SendQ exceeded).
09:08:39 <kmc> what's that
09:09:08 <kmc> is it delicious
09:13:40 <shubshub> Ihave a huge fejoa tree in my yard
09:25:18 -!- zzo38 has quit (Quit: That is what happens when you burn your finger on an ash tray).
09:44:21 <pikhq_> Why am I good at being awake at unreasonable hours?
09:45:16 <pikhq_> Least desirable skill.
09:45:27 <pikhq_> Am I turning into elliott?
09:45:52 <pikhq_> I've got the "awake at stupid hours and complaining about it" bit down, at least.
09:46:03 * shubshub should start playing dwarf fortress
09:46:25 <pikhq_> Also the hubris. But that just comes with being a programmer.
09:46:38 <shubshub> How is that useful :P and whats hubris?
09:47:49 <shubshub> Phantom_Hoover: how is playing dwarf fortress a usefull skill?
09:48:14 <Phantom_Hoover> suppose you find yourself stuck in the wilderness with 6 other dwarves and a wagon full of supplies
09:48:48 -!- Ngevd has joined.
09:48:51 <pikhq_> Phantom_Hoover: I suffer from miasma.
09:49:39 <Phantom_Hoover> Ngevd, do you feel you could survive indefinitely given only a square of lightless dirt, one plump helmet seed and a rock.
09:50:08 <Ngevd> Phantom_Hoover, probably
09:50:28 <Ngevd> Assuming I had some space next to the dirt
09:50:48 -!- ais523 has joined.
09:50:48 <pikhq_> Phantom_Hoover: Can't farm that seed without a hoe and a water block.
09:51:05 <Ngevd> pikhq_, it's dirt, you can use your hands
09:51:22 <Ngevd> pikhq_, and Plump Helmets are pretty hardy
09:51:34 <Phantom_Hoover> You only need water for bare cave floors, and moving it to the farm should always be a colossal undertaking.
09:51:57 <pikhq_> The only dwarves I like are in Boatmurdered.
09:52:00 <Ngevd> My advice is don't try to make an underground fishing plaza
09:53:50 <shubshub> aaaaah that Peaceful Dragon Quest 9 Music ATLAST :D
09:54:53 <shubshub> Maybe DQH Rocket slimes is better then :P
09:55:27 <shubshub> Dragon Quest Heroes Rocket Slimes FTW
10:00:15 <itidus21> http://www.youtube.com/watch?v=cGA7kGPEVPE "Drunk teacher"
10:01:13 <Ngevd> All of my fortresses that have ended for reasons other than boredom or goblins, have ended due to underground fishing plazas
10:06:59 -!- cheater__ has joined.
10:10:46 -!- cheater_ has quit (Ping timeout: 276 seconds).
10:22:58 -!- MoALTz has joined.
10:31:36 -!- cswords has joined.
10:34:10 -!- cswords__ has quit (Ping timeout: 276 seconds).
10:40:54 -!- shubshub has quit (Quit: Its Illegal).
10:41:36 -!- MoALTz_ has joined.
10:42:29 -!- MoALTz has quit (Ping timeout: 248 seconds).
10:46:19 -!- Ngevd has quit (Read error: Connection reset by peer).
10:50:26 -!- Ngevd has joined.
10:53:37 -!- MoALTz__ has joined.
10:55:37 -!- MoALTz__ has quit (Client Quit).
10:56:18 -!- MoALTz_ has quit (Ping timeout: 244 seconds).
11:14:38 -!- Ngevd has quit (Ping timeout: 240 seconds).
11:17:26 <kmc> shachaf: i'm playing with perf now
11:21:01 <kmc> kind of insane to run a short test program and see "98,469,751,820 instructions"
11:21:08 <kmc> i'm still not really used to how fast computers are
11:29:07 -!- nortti has joined.
11:30:22 -!- derdon has joined.
11:39:07 <nortti> sasl still requires a bit of tweaking (you can't use autoconnect) but I have moustly solved my problems like irssi crashing
11:39:29 -!- Ngevd has joined.
11:53:55 -!- ion has quit (Ping timeout: 252 seconds).
11:56:43 -!- Ngevd has quit (Ping timeout: 276 seconds).
12:06:46 -!- itidus20 has joined.
12:08:29 -!- ion has joined.
12:10:10 -!- itidus21 has quit (Ping timeout: 256 seconds).
12:11:22 -!- Ngevd has joined.
12:13:37 <Ngevd> I'm in a Taneb mood
12:13:39 -!- Ngevd has changed nick to Taneb.
12:14:07 <HackEgo> :mt"4}*.J<,C.).Y?;H.sܭ~~Fv..$:..|]&.}Dt7C9.Sl-O9.Ӝ.O-g[XPwU{%.|,.tvڛ3q.ZeQ.ie=hߑ..e2w8M2ND/ZWj/.'a.*{t..6.؋2f7tZayj.%5d$ʠKK..PyEqnhS*f"4..uI.pvqI.fe. \ HÚNO.K1u.5M&g.
12:16:33 -!- MoALTz has joined.
12:17:04 <Taneb> `learn Taneb is not elliott, no matter who you ask.
12:18:51 <HackEgo> elliott wrote this learn DB, and wrote or improved many of the other commands in this bot. He probably has done other things?
12:26:12 -!- MoALTz has quit (Ping timeout: 244 seconds).
12:51:02 -!- nortti has quit (Quit: Lost terminal).
12:58:11 -!- MoALTz has joined.
12:59:29 <kmc> shachaf: i'm unable to reproduce the performance hit on my other atom machine :/
12:59:52 <kmc> and stack protector does not make a large difference in the number of branches or branch mispredictions
13:34:18 -!- Taneb has quit (Quit: Goodbye).
13:36:23 -!- itidus22 has joined.
13:40:37 -!- itidus20 has quit (Ping timeout: 248 seconds).
13:53:50 <cheater__> kmc: maybe one of your machines is compiled for atom's funny machine code and not for x86?
13:54:22 -!- Phantom_Hoover has quit (Ping timeout: 276 seconds).
13:56:36 <kmc> what's that cheater__?
13:56:42 <kmc> do you mean LPIA?
13:57:11 <cheater__> no idea, i remember atom has its own set of instructions that it translatex x86 to
13:57:57 <kmc> i think you're thinking of lpia
13:58:15 <kmc> which isn't really a different instruction set, just a different set of optimization options
13:58:21 <kmc> as i understand it
13:58:44 <kmc> was supported by ubuntu for a little while, but it never took off and got desupported
13:58:47 <cheater__> i thought atom was actually some sort of risced thing
13:59:00 <cheater__> yeah, i heard they shut it down when it brought no improvement at all
13:59:36 <kmc> i mean all (?) x86 implementations use microcode
13:59:58 <kmc> you can sort of consider that 'translating to RISC' but Atom is not special in this regard, afaik
14:00:20 <cheater__> yeah but you can't normally feed the cpu its microcode
14:00:35 <kmc> sometimes you can upload new microcode
14:00:41 <kmc> to fix bugs with particular instructions, or add new ones
14:00:55 <cheater__> can you do that on any desktop intels
14:01:06 <kmc> but that's more like redefining a single instruction; you can't upload an arbitrarily long stream of microcode and say "execute this" afaik
14:01:37 <cheater__> is that because of the pentium floating point mult fiasco
14:01:55 <kmc> hm I doubt that would be fixable in microcode anyway
14:01:56 <kmc> but i don't know
14:03:24 <cheater__> so what are you trying to do on atom?
14:03:40 <cheater__> sounds like something being done for work
14:05:04 <kmc> Transmeta implemented x86 with a software JIT sort of thing and a chip with their own VLIW architecture
14:05:24 <cheater__> or is it a super duper secret ksploracle project you're working on
14:05:43 <kmc> and of course you can JIT x86 to RISC architectures not explicitly designed for it
14:05:46 <kmc> using qemu
14:06:17 <cheater__> i like how qemu does every thing you can imagine
14:06:30 <kmc> and the Loongson 3 variant of MIPS has a bunch of extra instructions to make that JITted code faster
14:06:36 <kmc> specifically for x86
14:06:43 <kmc> i don't work for ksploracle
14:07:06 <kmc> lately i have been working on http://mosh.mit.edu/
14:07:11 <kmc> as a volunteer / open source thing
14:07:41 <cheater__> the only thing i've been doing lately is working out
14:07:47 <kmc> one thing i did was add automatic binary hardening -- stack protector, position independent code, etc.
14:07:51 <cheater__> i should start doing like, geeky stuff again
14:07:59 <kmc> and so I wanted to see how much performance this cost
14:08:05 <kmc> and one of my test machines was this atom netbook
14:08:35 <kmc> and it got a surprisingly large hit from stack protector
14:08:42 <kmc> and I'm not sure why
14:08:46 <kmc> so that's what I'm trying to figure out
14:10:04 <kmc> turning on all the hardening cost like 29% on the netbook
14:10:36 <cheater__> how much was it on a normal pc thing
14:10:55 <kmc> 17% on another 32-bit Atom system
14:11:09 <kmc> 4% on my Core 2 Duo laptop
14:11:20 <kmc> 2% on my Phenom II desktop
14:11:36 <kmc> it's expected that position-independent code has a 5-10% performance penalty on 32-bit x86
14:11:50 <cheater__> have you thought that maybe it's just a constant amount of performance being eaten
14:12:04 <cheater__> and since an atom has overall much less performance that constant amount is a higher percentage
14:12:24 <kmc> but I'm not sure why stack protector should have another 20% on top of that, for the one atom only
14:12:27 <kmc> yeah, that might be
14:16:46 -!- asiekierka has joined.
14:17:34 <kmc> comparing the speeds of the two atom machines just about accounts for 29% vs 17% hit
14:17:45 <kmc> but on the former, like 80% of the performance hit goes away if you disable stack protector
14:17:55 <kmc> and on the latter only like 40%
14:19:00 <kmc> but it's all a bit iffy
14:19:06 <kmc> there are various other differences, like GCC version
14:23:23 <kmc> i'm not trying to do rigorous science here
14:23:33 <kmc> just to get an idea of how much it costs us
14:23:49 <kmc> really that's already settled and I'm just writing up a blog post for other people who want to add hardening to their autoconf projects
14:27:27 <kmc> cheater__: mosh supports 256 colors, if the outer terminal does
14:28:46 -!- Patashu has quit (Ping timeout: 272 seconds).
14:36:31 <kmc> i mainly use it for irssi
14:36:58 <cheater__> i tire of chatting in fixed fonts quickly
14:37:22 <kmc> whereas i swiched from xchat to irssi so i could have a persistent session over mosh :)
14:37:33 <cheater__> http://www.dangerousmusic.com/ass.html
14:37:35 <kmc> gotta do the same for aim / gchat
14:37:51 <cheater__> yeah i kinda don't give a fuck about that
14:38:12 <cheater__> it's mostly, look on irc, see if you can have fun for n minutes, leave
14:38:26 <cheater__> having scrollback is useful but only marginally
14:38:55 <kmc> i've been on connections which are too shit to connect to irc at all
14:38:59 <kmc> but mosh works ok
14:39:15 <kmc> also if you're in the middle of a conversation and need to switch from wifi to cell or such
14:39:18 <kmc> then the roaming is nice
14:39:28 <cheater__> that it works better than irc that is
14:39:46 <kmc> yeah, IRC is TCP-based
14:39:56 <kmc> TCP does not do well with non-congestive packet loss
14:40:39 <cheater__> kmc: would you put these in your house http://hifiheroin.blogspot.de/2012/04/munich-high-end-2012-preview.html
14:40:45 <kmc> and it will drop entirely if you go too long without an ack
14:41:04 <kmc> whereas a mosh session will survive arbitrarily long and will come back within 3 seconds if the network comes up
14:41:32 <kmc> i can take my laptop on the subway and use my mosh sessions for the one bit where it comes out of the ground to go over the bridge between Boston and Cambridge
14:41:43 <cheater__> is 3 seconds a built in delay or something?
14:41:43 <kmc> that's the heartbeat packet interval
14:48:41 <Sgeo> kmc, the mobile shell, or something else?
14:48:51 <Sgeo> (See? I did Google!)
14:48:53 <kmc> cheater__: speaking of qemu have you used ubuntu's magical qemulated chroots?
14:49:17 <kmc> Sgeo: i'm talking about http://mosh.mit.edu/
14:49:26 <Sgeo> kmc, ah, so yes
14:51:51 <Sgeo> No Windows client :(
14:52:54 -!- augur has quit (Remote host closed the connection).
14:55:01 <Sgeo> " The heartbeats allow Mosh to inform the user when it hasn't heard from the server in a while (unlike SSH, where users may be unaware of a dropped connection until they try to type)."
14:55:03 <Sgeo> I love it already
14:55:08 <kmc> mosh-1.2 works in Cygwin, maybe
14:55:17 <kmc> some people say it does, some say it doesn't
14:55:36 <kmc> the core mosh team doesn't use Windows
14:56:08 * ais523 tries to explain the may bank holiday to non-british people
14:56:32 <ais523> (elliott and ngevd, when they logread this, will understand why that's amusing)
14:56:45 <ais523> (or maybe not, if they assume that every country works like that)
14:56:53 <kmc> what's the amusing part?
14:57:33 <ais523> kmc: it's a holiday that doesn't commemorate anything
14:57:42 <ais523> most countries have too many holidays to celebrate one thing or another
14:58:00 <ais523> the UK doesn't have enough, so it adds a couple at convenient times of year just because people need a holiday
14:58:12 <kmc> so it's not a May Day labour thing?
14:58:25 <ais523> well, hmm, perhaps it was originally, I'll check wikipedia
14:59:23 <Sgeo> :Q: Does mosh work with Amazon EC2?
14:59:23 <Sgeo> Yes, it works great, but please remember to open up UDP ports 6000061000 on the EC2 firewall.:
14:59:39 <ais523> hmm, seems it was originally created by merging 8¼ saint's days (on average) and the day after whitsun
14:59:43 <Sgeo> Did I seriously use : as a quote character?
14:59:50 <Sgeo> And why does mosh need so many ports?
14:59:54 <cheater__> kmc: what are magical qemulated chroots good for?
15:00:08 <ais523> in an attempt to cut down on the excessive number of saint's days
15:00:08 <kmc> US labor day is pretty decoupled from actual labor stuff for most people
15:00:15 <kmc> but this may reflect my white-collar upbringing
15:00:23 <kmc> Sgeo: it doesn't need them all at once, just one per session
15:00:39 <ais523> oh, wait, no, that's late spring bank holiday
15:00:44 <kmc> mosh-server picks the first available port in that range
15:00:53 <ais523> first monday in may is a purely arbitrary one indeed
15:01:09 <kmc> so if you only need 5 concurrent sessions, you only need open 5 ports
15:01:25 <kmc> ais523: cool
15:01:33 <kmc> i think japan has some holidays like that too
15:01:41 <ais523> and was only added in 1978; surprisingly recent
15:02:05 <cheater__> kmc: why doesn't it reuse the port
15:02:39 <kmc> there is no persistent mosh daemon
15:02:54 <kmc> each session launches its own mosh-server process
15:03:10 <kmc> they could be different users running different versions of mosh-server they compiled separately
15:03:20 <kmc> and so the easiest way to get packets to the right place is for each session to use its own port
15:04:31 <kmc> http://en.wikipedia.org/wiki/Public_holidays_in_Japan
15:05:33 <kmc> Massachusetts has one extra official public holiday compared to the US federal govt
15:05:40 <kmc> this year it fell on Tax Day
15:17:24 <ais523> kmc: holiday clashes are impossible in the UK
15:17:35 <ais523> this sometimes leads to Christmas being moved to dec 26, if it would naturally fall on a sunday
15:17:41 <ais523> (and thus boxing day moved to dec 27)
15:17:48 <ais523> although everyone celebrates it on dec 25 anyway
15:18:54 <shachaf> I'm not sure what *that* means.
15:19:02 <ais523> I didn't think you would be
15:19:07 <ais523> but it means hi, sort of
15:19:13 <kmc> shachaf: fun enough
15:19:19 <ais523> probably only elliott could get it, and that doesn't mean he /would/
15:19:34 <ais523> as it involves cross-referencing multiple inside mems
15:19:36 <cheater__> kmc: that's not the simplest, it's just the laziest
15:19:54 <kmc> oh yeah cheater__?
15:20:00 <cheater__> imo simplest is for mosh to use a shared resource
15:20:26 <kmc> a shared resource installed by a sysadmin? that's a non-starter for a lot of our users
15:20:56 <cheater__> no, you can have the process instantiate one locally if it's not found system-wide
15:21:00 <kmc> a big selling point of mosh is that you can compile it in your homedir and use it right away without any administrative intervention
15:21:08 <kmc> cheater__: local for that user only?
15:21:18 <kmc> so you still have the problem of multiple ports
15:21:39 <kmc> and you've added significant complexity to make this agent relay packets around
15:21:39 <cheater__> so one user ends up using one port at most
15:21:40 <cheater__> and at the least, all users just use one port
15:21:43 <kmc> for what gain?
15:21:58 <cheater__> you only have the problem in a fairly stupid situation
15:22:27 <cheater__> one which happens but does not happen to often
15:22:33 <cheater__> so kmc, you never told me what those qemu'ed chroots were good for
15:22:48 <kmc> what stupid situation?
15:22:49 <kmc> yeah cheater__
15:22:54 <cheater__> or a 32 bit chroot on a 64 bit system?
15:23:07 <kmc> you run a simple command or two and you get a chroot full of ARM binaries on an x86 Linux machine
15:23:11 <kmc> which automatically execute through qemu
15:23:12 <cheater__> the stupid situation is that there's no system-wide support for mosh
15:23:24 <cheater__> which only ever happens on multiuser systems
15:23:39 <cheater__> and let's be honest here, it's not like people do this sort of thing these days
15:23:39 <kmc> but they talk to your real kernel, not some emulated ARM machine
15:23:52 <cheater__> it's not like people run their BitchX from grex.org anymore
15:24:19 <cheater__> or try to hack the modified bsd kernel of cyberspace.org to be able to run 'drops
15:24:31 <kmc> as a mosh developer i'm really glad that Mosh doesn't have a persistent daemon or anything setuid / setgid or anything that runs as root
15:24:44 <cheater__> well you don't need to have a persistent daemon
15:24:46 <kmc> i'm reluctant to throw that away for the cause of saving a few UDP ports
15:24:53 <kmc> since there's 2^16 of them
15:24:56 <cheater__> you just have the first one that starts up handle everything
15:25:02 <cheater__> and persist until everything quits
15:25:10 <kmc> anyway I still don't see what problem this solves really
15:25:14 <kmc> it's certainly not the simplest solution
15:25:34 <cheater__> ideally you just want a single port
15:26:25 <kmc> schroot itself is also pretty slick
15:26:37 <kmc> it's used for the debian package build infrastructure, among other things
15:26:49 <kmc> a nice way to enter chroots and also have ephemeral copy-on-write clones of existing chroots
15:27:26 -!- augur has joined.
15:28:49 <kmc> cheater__ i don't see why you said "your solution is not simplest, just laziest" and then proceeded to describe a solution which is emphatically less simple, though it has certain advantages
15:29:03 <ais523> main advantage to a single port is that it's easier to firewall
15:29:10 <kmc> i guess in general the way engineers argue is they pick different definitions of the word "simple"
15:29:41 <cheater__> kmc: it's the simplest from the user's point of view
15:29:53 <cheater__> it's not the simplest from the developer's point of view
15:30:17 <kmc> it's better from the user's view, but arguably more complicated
15:30:27 <kmc> because you still need to explain about port ranges, in the multi-user case
15:30:39 <kmc> and you also need to explain about this extra daemon that stays running
15:30:49 <kmc> and about the unix socket or whatever it uses to talk to the mosh-servers
15:32:36 <kmc> i know corporate IT security doesn't like to open too many ports
15:33:02 <kmc> but is identifying a mosh session by a UDP port number really any worse than identifying it with a 16-bit number immediately inside the UDP packet?
15:33:18 <ais523> kmc: not beyond the firewall
15:33:33 <ais523> but configuring it correctly is awkward because many firewalls don't have obvious support for port ranges
15:33:57 <kmc> arguably the port number is better for security; it's easier to correlate port numbers with users and mosh-server processes
15:34:34 <kmc> i guess that's so
15:35:10 <kmc> my firewall config is just an iptables shell script and so i used a shell loop :)
15:35:16 <kmc> (though i think iptables also has range support)
15:43:12 -!- elliott has joined.
15:43:20 -!- elliott has quit (Changing host).
15:43:20 -!- elliott has joined.
15:46:12 <cheater__> multiple ports is just very impractical
15:46:21 <cheater__> and it's not like explaining my scheme is that difficult
15:47:39 <cheater__> "mosh will use a single port for all sessions if it can, but it might have to use more ports due to permissions. You can read about the exact scheme in [7]."
15:47:53 <cheater__> 99.999999999% of all people using it will not need to read [7]
15:48:42 <kmc> even if every person on earth is a Mosh user, only 7% of one person will need to read that link
15:50:08 -!- elliott has quit (Remote host closed the connection).
15:55:29 -!- itidus21 has joined.
15:57:56 -!- itidus22 has quit (Ping timeout: 255 seconds).
15:59:05 <kmc> anyway it's not impossible we will implement some kind of multiplexer
15:59:47 <kmc> i think what's more likely is that we'll use hole punching techniques (as do Skype and many other applications) to make both directions of UDP traffic look "outgoing" to firewalls
16:05:06 <cheater__> kmc: do you think there's an actual real reason to use quemu chroots?
16:05:07 <cheater__> i'm thinking, the one thing i am missing with linuxes right now is the ability to migrate a 32 bit system to 64 bit
16:05:07 <cheater__> it really really sucks there isn't anything to do it
16:05:08 <cheater__> so maybe this sort of thing is necessary
16:07:03 <kmc> i used it for testing and debugging the ARM build of Mosh
16:07:08 <kmc> on my amd64 desktop
16:07:20 <kmc> this was much more pleasant than using a whole system emulated arm system
16:07:23 <kmc> which i have also done
16:07:49 <kmc> qemu is not necessary for running 32-bit binaries on 64-bit kernel of the same architecture family, usually
16:08:29 <kmc> disabling 32-bit compat on a multi-user system is a savvy security move
16:08:34 <kmc> but i don't think many people do it
16:11:19 -!- AnotherTest has joined.
16:12:28 <cheater__> yeah but still, you can't migrate 32 bit to 64 bit
16:13:01 <kmc> what do you mean by "migrate" exactly
16:14:09 -!- cheater__ has quit (Quit: Ex-Chat).
16:16:06 <qfr> http://d24w6bsrhbeh9d.cloudfront.net/photo/3981354_460s.jpg oh really?
16:17:20 -!- cheater has joined.
16:17:48 <kmc> cheater: what do you mean by "migrate" exactly
16:18:14 <cheater> i have a computer with an OS which i installed with a 32 bit ubuntu dvd
16:18:26 <cheater> i want it to start using 64 bit versions of everything
16:35:56 <Sgeo> I'm a Cadaeic! I'm a Cadaeic! I'm a real Cadaeic!
16:37:29 <kmc> cheater: ok
16:37:34 <kmc> that's a package manager problem i guess
16:37:44 <kmc> i wonder if debian's new multiarch support helps
16:38:07 <kmc> i don't know if it has a way to "pivot" which binary is actually stored at /bin/ls or whatever
16:39:28 -!- cheater has quit (Ping timeout: 276 seconds).
16:49:13 -!- pikhq_ has quit (Ping timeout: 276 seconds).
16:50:14 -!- cheater has joined.
17:10:20 -!- cheater has quit (Ping timeout: 260 seconds).
17:12:07 -!- cheater has joined.
17:17:49 -!- cheater has quit (Ping timeout: 276 seconds).
17:31:12 -!- ais523 has quit.
17:39:45 -!- cheater has joined.
18:01:17 -!- pikhq has joined.
18:04:12 -!- nortti has joined.
18:05:56 <nortti> Power plug of my computer fell off when I was updating slitaz. Now it doesn't want to boor
18:06:43 <olsner> you may need to plug in the power again
18:09:47 <nortti> olsner: i have done that
18:10:37 <nortti> olsner: it boots to grub and starts loading and then it panics
18:11:30 <olsner> maybe you can ease it up against a wall, that's what you do with panicking horses apparently
18:13:45 -!- MoALTz has quit (Quit: brb).
18:15:40 -!- cheater has quit (Ping timeout: 276 seconds).
18:19:21 -!- AnotherTest has left.
18:19:43 -!- cheater has joined.
18:23:59 -!- asiekierka has quit (Remote host closed the connection).
18:27:45 <Gregor> I almost want to `addquote <nortti> Power plug of my computer fell off when I was updating slitaz. Now it doesn't want to boot <olsner> you may need to plug in the power again
18:28:40 <Gregor> It's too obvious. It's not an especially #esoteric joke.
18:45:48 -!- elliott has joined.
18:46:15 -!- zzo38 has joined.
18:53:50 <lambdabot> elliott: You have 6 new messages. '/msg lambdabot @messages' to read them.
18:56:34 <elliott> ten minutes and many hours
18:57:10 <Gregor> If we assume PM, you were in fact only ten minutes off for me.
18:58:50 -!- cheater has quit (Ping timeout: 260 seconds).
19:06:55 -!- pikhq_ has joined.
19:06:57 -!- pikhq has quit (Ping timeout: 252 seconds).
19:28:01 <olsner> hmm, my ability to differentiate the words terrific and terrible is deteriorating
19:28:58 -!- Ngevd has joined.
19:30:05 <elliott> Ngevd: How's your roguelike?
19:30:34 <Ngevd> I haven't worked out how to simulateneously render something and input characters!
19:30:48 <elliott> You render the screen, and then you wait for character input.
19:31:13 <olsner> use threads, and you can also simultaneously crash terrifibly unreliably
19:31:15 <elliott> (You probably want to use a terminal library like one of the curses bindings or vty if you're not.)
19:31:26 <Ngevd> Yeah, I'll look into that
19:31:32 <Ngevd> I DO MANY THINGS NOW
19:31:44 -!- cheater has joined.
19:31:52 <elliott> (Note that some of the curses bindings are awful.)
19:31:53 <Ngevd> Including what is now the second-best chess game for the 0x10c CPU
19:32:12 <elliott> (vty might be the best option, even if it is kind of gross.)
19:33:42 <oklopol> i had this girlfriend who lived in kuopio and was going to move to turku
19:33:52 <oklopol> and now i have this girlfriend who lived in turku and moved to kuopio
19:33:58 <oklopol> and they are fucking neighbors
19:34:23 <oklopol> they don't know each other, and have no links.
19:34:28 <Ngevd> Are you the neighbors?
19:34:34 <olsner> oklopol: you are a link
19:34:42 <olsner> oklopol: introduce them
19:35:02 <oklopol> "hello ex, this is why i left you"
19:35:33 <oklopol> i just wanted to drive 800km to show you.
19:36:03 <olsner> then they will knife-fight to the death (as all finns do) and you will take the winner
19:36:20 <elliott> olsner: Are you sure you weren't born Finnish?
19:36:31 <olsner> you don't have to drive all the way though, you can just tell them "hey, she's your neighbor now!"
19:36:32 <Lumpio-> Nothing wrong with some finnish
19:37:03 <elliott> Finns use ridiculous smilies like (￣ー￣).
19:37:33 <oklopol> olsner: i'm going there this weekend. i'm sure this will be fun.
19:37:47 <nortti> elliott: that is aproximaltely how out face looks like
19:38:31 <Ngevd> elliott, you're like that person in that book I'm reading by that person who Phantom_Hoover is disappointed by his appearance due to it not being recursive enough
19:39:02 <elliott> Ngevd: Is that the book I think it is?
19:39:32 <Ngevd> elliott, that would depend on which book you think it is.
19:39:48 <elliott> I will just leave http://esolangs.org/wiki/User:Chris_Pressey#G.C3.B6del.2C_Escher.2C_Bach:_An_Eternal_Golden_Braid here.
19:39:52 <Ngevd> Are you thinking of The Stainless Steel Rat's Revenge?
19:40:40 <Ngevd> One, you live in my local area
19:40:46 <Ngevd> Two, you have EYES EVERYWHERE
19:40:57 <Ngevd> Three, I am currently reading The Stainless Steel Rat's Revenge
19:41:08 <elliott> I'm going to have to move away from Ngevd. :'(
19:41:27 * elliott tries to figure out where the recursion comes in.
19:42:28 <olsner> next to the review of wolfram's book, I just can't figure out what level of sarcasm/irony/parody that GEB review is written with
19:43:07 <elliott> Has anyone ever ended a serious statement with "!!1!"?
19:43:32 <olsner> yes, I'm sure someone has
19:44:08 <elliott> Ah yes, that's what Chris thought of the Dragon Book: "The classic, borderline-incomprehensible book on compiler construction".
19:44:18 <olsner> Has anyone ever ended a retorical question with?
19:44:43 -!- azaq23 has joined.
19:44:46 -!- Ngevd has quit (Read error: Connection reset by peer).
19:44:51 -!- azaq23 has quit (Max SendQ exceeded).
19:44:54 <elliott> coppro: Who's Noé Rubinstein?
19:45:05 -!- Ngevd has joined.
19:45:13 <elliott> Never mind, Gmail answered my question first.
19:45:30 -!- azaq23 has joined.
19:47:05 <Ngevd> "The robot stood as stolid as a statue"
19:47:54 <Ngevd> Harry Harrison really needs a better editor
19:48:49 <olsner> Ngevd: stolid is a word that can be used in a sentence just like that
19:49:13 <olsner> it even makes sense, although statues aren't necessarily stolid
19:49:21 <olsner> good statues probably aren't
19:49:50 <olsner> terrifibly stillily stolid
19:50:33 <elliott> Ngevd: I'm bored. You should play Crawl so I can watch and laugh.
19:51:06 <Ngevd> elliott, I'm currently using Windows for stilly and friviltous treasons
19:51:29 <Ngevd> Can you stream with PuTTY?
19:51:34 <Ngevd> I thought it could only recieve
19:51:37 <elliott> Yes, you play on the server.
19:51:47 <elliott> Which then streams it for you (and also lets people send messages).
19:52:06 <Ngevd> Okay, running PuTTY
19:52:13 <elliott> crawl.develz.org, port 345
19:54:43 <olsner> basically you telnet to it and it appears in your terminal? sort of like how telnet works?
19:54:45 <elliott> It's sometimes a bit laggy. But then it isn't.
19:55:21 <elliott> olsner: It lets other people watch.
19:55:34 <elliott> Ngevd: Go for Minotaur or Troll or something.
19:55:35 -!- Ngevd has quit (Read error: Connection reset by peer).
19:55:43 <elliott> If you want something overpowered enough to let you survive on pure luck. :p
19:56:00 -!- Ngevd has joined.
19:56:07 <elliott> Ngevd: Go for Minotaur or Troll or something.
19:56:10 <elliott> If you want something overpowered enough to let you survive on pure luck. :p
19:56:38 <elliott> Fighter or Berserker would complete the overpowered-melee-fighter combo.
19:57:02 <elliott> Ngevd: You know about vikeys, right?
19:57:15 <elliott> Don't use the arrow keys, you'll die, because you can't do diagonals.
19:57:22 <elliott> If you have a number pad you can use that.
19:57:26 <elliott> Otherwise it's hjkl/yubn to move.
19:57:36 <elliott> Also, the two more useful keys are o and tab.
19:58:13 <elliott> Ngevd: You should chop that corpse.
19:58:55 -!- oerjan has joined.
20:00:12 <elliott> (You might want to adjust the skills you're training by pressing m.)
20:00:24 <elliott> (Turning off everything but axes would be reasonable.)
20:01:02 <elliott> Crawl's skills settings are kind of complicated.
20:01:42 <elliott> Ngevd: You're still using the arrow keys, aren't you. :(
20:02:57 <Ngevd> I just forget about diagonals
20:03:11 <elliott> Ngevd: Did you leave that scroll there on the previous level?
20:03:15 <elliott> (Ctrl+F scroll <return> to check.)
20:03:28 <elliott> Try a. to go back there and pick it up.
20:03:40 <oerjan> hm it appears that DMM has filled up the LMOO queue by drawing other other people
20:03:54 <elliott> (The advantage of using o rather than walking around manually, apart from being less tedious, is that it makes sure you don't miss picking anything up>0
20:04:50 <elliott> You should chop corpses after killing them with "c", unless they show up as green text. (So you can eat them when you get hungry.)
20:05:52 <elliott> Eating is... a bad idea when you're in combat.
20:06:29 <elliott> 21:05 <Gretell> Taneb the Shield-Bearer (L3 MiFi), blasted by Sigmund (magic dart) in D:3, with 91 points after 1086 turns and 0:08:54.
20:13:07 <Ngevd> Well, neither Vty nor hscurses want to install
20:18:56 <elliott> vty should be easier to install.
20:19:26 <nortti> Ngevd: what are you trying to do?
20:21:10 <Ngevd> elliott, http://hpaste.org/68200
20:21:42 <elliott> I think there might be a curses binding on Hackage that works with PDCurses. Maybe.
20:22:09 <elliott> Oh, it's http://hackage.haskell.org/package/nanocurses. But that package has been broken for years.
20:25:49 <zzo38> If you are writiing a roguelike, a few things to consider are: * Separate front-end and back-end * PC/NPC unification * If using Haskell, the "extensible-data" package
20:29:05 <nortti> http://www.terminally-incoherent.com/blog/2010/05/18/lets-use-a-methaphor-charlie-and-his-car/
20:30:48 <elliott> Challenge: Write introductory paragraph that can start no blog post of value. Submission: "Few weeks ago I posted a rant about stupid people not knowing how to internet. Most of the readers could relate, but some folks criticized me for being elitist. I’m pretty sure these people were just trolling but I guess they were successful enough to make me want to revisit this post."
20:32:02 <elliott> Ngevd: Anyway, you could just...
20:32:05 <elliott> Ngevd: I dunno, use SDL or something.
20:32:14 <elliott> Or write it in Python, that has libtcod bindings. :p
20:54:29 <Ngevd> I was using ansi-terminal, but input is dodhy?
20:55:55 <elliott> ansi-terminal can't do the kind of things you'd want.
20:56:08 <elliott> I suppose you could just redraw the entire screen each time... :p
20:57:15 <olsner> elliott: thanks, you just made me not read nortti's link
20:57:35 <elliott> olsner: But now you've missed out. :(
20:57:58 <Ngevd> That blog is awful
20:58:22 <elliott> You can miss out on awful things, too!
20:58:25 <nortti> Ngevd: why do you think so
20:58:48 <Ngevd> Because Charlie is an awful name!
20:59:06 <elliott> Ngevd: Did you know that every teenager consumes porn?
20:59:07 <elliott> http://www.terminally-incoherent.com/blog/2012/05/02/is-your-teen-browsing-the-pr0n/
20:59:07 <olsner> now I just need to learn how to internet, to make sure that blog poster doesn't make fun of me
20:59:16 -!- blacksun1 has joined.
20:59:22 <elliott> Ngevd's cover is blast wide open.
20:59:23 <Ngevd> elliott, FSVO porn
20:59:29 <HackEgo> blacksun1: 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.)
21:00:07 <Ngevd> I occasionally have disturbing deviantArt binges
21:00:19 -!- blacksun1 has quit (Client Quit).
21:07:03 -!- Ngevd has quit (Quit: Goodbye).
21:26:26 -!- nortti has quit (Quit: nortti).
21:32:44 -!- itidus20 has joined.
21:35:53 -!- itidus21 has quit (Ping timeout: 245 seconds).
22:06:20 -!- Patashu has joined.
22:10:25 -!- augur has quit (Remote host closed the connection).
22:38:25 <Gregor> <Ngevd> I occasionally have disturbing deviantArt binges <elliott> TELL US NO MORE
22:38:46 <Gregor> Welp, time to look up "pony" on Deviant Art and see how many pages I have to go through 'til I get to pony porn.
22:39:50 <elliott> Gregor: That's some cover you have there.
22:39:59 <elliott> How long have you been waiting for a tangentially-related IRC discussion to reply to with that?
22:40:47 <Gregor> I was going to say that right now regardless, it was merely happy happenstance that it vaguely fit a previous conversation.
22:48:00 -!- oerjan has quit (Quit: Lost terminal).
22:49:18 <Gregor> <Lumpio-> ( ﾟ∀ﾟ) // I honestly can't make heads or tails of this smiley.
22:49:25 <Gregor> Is it smiling? Freaked out? A bird?
22:50:24 <elliott> It looks like the O RLY owl to me.
22:50:27 <Gregor> That's pretty understandable.
22:50:45 -!- variable has changed nick to const.
22:52:16 <Lumpio-> http://shoppies-img.jp/res_items3/746/7465945/PT_110601202610190.jpg
22:52:35 <Lumpio-> I guess an exaggerated version of ( °▽°)
22:52:36 <elliott> My system's rendering of it is better.
22:52:43 <elliott> The dors are wider than just the dot, and to the left of the character.
22:53:21 <Lumpio-> Yeah that's a pretty odd rendering
22:53:30 <Lumpio-> They're not dots, they're degree signs.
22:53:41 <elliott> Degree signs are just dots with ego.
22:53:43 <Lumpio-> They're supposed to be circles.
22:54:09 <elliott> They turn into hollow circles when I increase the font size.
22:54:19 -!- derdon has quit (Remote host closed the connection).
22:54:23 <Lumpio-> Just leave it on the higher setting from now on
22:54:30 <elliott> It's hard to display the hollowness without distorting the sizes at a font size like that.
22:54:38 <elliott> Lumpio-: I already interpreted it as a circle.
22:54:49 <elliott> Lumpio-: Also, that is a different character.
22:54:54 <Lumpio-> Maybe you should change fonts
22:54:57 <elliott> That one renders as larger and more hollow here.
22:55:44 <elliott> http://ompldr.org/vZG51cQ Compare your original smiley to your "So you can see my" line.
22:55:59 <Lumpio-> It's not a degree sign, it's a half-width handakuten mark
22:56:48 <Lumpio-> (Which should also preferably be a circle, mind you)
22:57:05 <elliott> It is a circle, just a small one.
22:57:05 <Lumpio-> ugh, mozc's standard dictionary uses a full-width handakuten mark for those
22:57:22 <Lumpio-> Yeah but the proportions are off.
22:57:23 <elliott> Japanese text isn't really well-suited to this font size on displays with standard pixels per inch.
22:57:43 <elliott> (Even though this display has an above-average PPI.)
22:57:45 <Lumpio-> My font size is the same or smaller tbh
22:58:00 <elliott> It would probably be better at this size with more hinting.
22:58:12 <Lumpio-> Yeah, full hinting is a must.
22:58:18 <elliott> But I don't use Japanese on IRC, so :p
22:58:22 <Lumpio-> Otherwise it's just black blobs.
22:59:01 <Lumpio-> Enjoy your blobs then I guess
22:59:09 <Lumpio-> Full hinting + stroke elimination is a must for small text
22:59:47 <Lumpio-> On my resolution even 書 has enough horizontal lines to require stroke elimination
23:00:00 <elliott> If there's any justice, increasing PPI will obsolete that (and antialiasing) soon enough.
23:00:02 <elliott> (There isn't any justice, though.)
23:00:14 <elliott> (Especially since most OSes are woefully badly-equipped to handle larger PPIs.)
23:00:28 <Lumpio-> (Let me guess, "except OSX")
23:00:36 <elliott> No, OS X is terrible at it.
23:00:51 <elliott> OS X just assumes everything is 100 PPI and uses bitmap images for most things.
23:01:00 <Lumpio-> Anyways PPI hasn't increased on your usual laptop displays for years
23:01:12 <Lumpio-> Consumers seem to be content with "full HD"
23:01:29 <Lumpio-> Hard to find displays with better resolution.
23:02:05 <elliott> This display is ~128 PPI, which is nice. (1440x900 at 13")
23:02:07 <Lumpio-> Actually I doubt most desktop displays go higher than that either.
23:02:20 <pikhq_> 96 DPI is still standard. Alas.
23:02:32 <elliott> My other laptop is the same size but 1366x768, which is smaller than I would like.
23:02:47 <Lumpio-> LVDS1 connected 1440x900+1920+0 (normal left inverted right x axis y axis) 367mm x 229mm
23:03:02 <Lumpio-> I wonder if xrandr can calculate DPI for me, too lazy to
23:03:33 <elliott> "578 characters were initially added, though some characters unsuitable for names such as 怨 (grudge, resent), 痔 (haemorrhoids) and 屍 (corpse) were removed as a result of public feedback."
23:03:50 <elliott> You mean I can't call my kid Grudge Haemorrhoidscorpse?
23:04:07 <elliott> So much for free speech. :(
23:04:53 <Lumpio-> I don't think that would go through in most countries...
23:04:54 <pikhq_> elliott: You still can.
23:05:13 <pikhq_> It just can't be 痔屍 怨.
23:05:20 <pikhq_> You'll have to spell it out.
23:06:20 <pikhq_> グラッジ ヘモロイヅコープす morelike
23:07:14 <Lumpio-> Or KURAJI HEMOROIDOKOUPUSU
23:07:36 <pikhq_> Lumpio-: Nonono. Kuraji HEMOROIDOKOUPUSU
23:07:41 <Lumpio-> Yeah, google translate doesn't do it justice
23:07:41 <elliott> "Piles of corpses of resentment"
23:07:47 <elliott> Well, that's my new metal album name.
23:07:53 <Lumpio-> pikhq_: ...I capitalize randomly because Japan doesn't have capitalization.
23:08:14 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:19 <elliott> 悪の宇宙へようこそハローお名前とご住所を記入してください。
23:08:30 <Lumpio-> pikhq_: ...I capitalize randomly because Japan doesn't have capitalization.
23:08:34 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:36 <Lumpio-> pikhq_: ...I capitalize randomly because Japan doesn't have capitalization.
23:08:37 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:37 <elliott> 悪の宇宙へようこそハローお名前とご住所を記入してください。
23:08:40 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:43 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:46 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:49 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:51 <Lumpio-> I reject your reality and substitute my own.
23:08:52 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:55 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:08:58 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:09:01 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:09:04 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:09:07 <pikhq_> Lumpio-: When you romanise names, if you do all-caps, the bit in all-caps is a last name.
23:09:11 <Lumpio-> elliott: "Welcome to the space of evil, hello. Please write down your name and address"?
23:09:19 <elliott> 音声学アルカトラズは、オステオパシーの解釈の赦免をバカ-うねる。
23:09:51 <elliott> "Phonetics Alcatraz osteopathic moron-undulate exegesis pardoning."
23:10:00 <elliott> Through the filter of GOOGLE.
23:11:05 <Lumpio-> Phonetics Alcatraz stupid-twist the pardon of the analysis of osteopathy
23:11:20 <zzo38> pikhq_:That is what I eould have say. You have to use kana, only that specific kanji disallow
23:11:37 <elliott> Piles of Corpses of Resentment, by The Pardon of the Analysis of Osteopathy
23:13:18 <elliott> You have to admit, it's catchy.
23:14:55 -!- cheater has quit (Ping timeout: 260 seconds).
23:17:01 -!- zzo38 has quit (Remote host closed the connection).
23:18:41 -!- augur has joined.
23:18:44 -!- azaq23 has quit (Quit: Leaving.).
23:26:57 -!- Patashu has quit (Quit: MSN: Patashu@hotmail.com , Gmail: Patashu0@gmail.com , AIM: Patashu0 , YIM: patashu2 , Skype: patashu0 .).
23:27:05 -!- Patashu has joined.