00:00:01 <esowiki> [[Poolshark]] https://esolangs.org/w/index.php?diff=60168&oldid=60162 * Orby * (+640) /* Poolshark on the unit square */
00:03:31 <esowiki> [[Poolshark]] M https://esolangs.org/w/index.php?diff=60169&oldid=60168 * Orby * (+6) /* Poolshark on the unit square */
00:11:31 <esowiki> [[Poolshark]] M https://esolangs.org/w/index.php?diff=60170&oldid=60169 * Orby * (+112)
00:12:50 <esowiki> [[User:Orby]] https://esolangs.org/w/index.php?diff=60171&oldid=60006 * Orby * (+74)
00:17:02 -!- Lord_of_Life_ has joined.
00:18:19 <esowiki> [[Poolshark]] M https://esolangs.org/w/index.php?diff=60172&oldid=60170 * Orby * (+289) /* Irrational slope */
00:19:15 <esowiki> [[Poolshark]] M https://esolangs.org/w/index.php?diff=60173&oldid=60172 * Orby * (+10) /* Irrational slope */
00:20:00 -!- Lord_of_Life has quit (Ping timeout: 250 seconds).
00:20:00 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
00:24:16 -!- oerjan has joined.
00:27:44 <kmc> helloerjan
00:32:10 <kmc> oerjan: kimchi?
00:32:43 * oerjan may never have actually tasted it
00:33:54 <kmc> kimchi is tg
00:34:15 <orbitaldecay> oerjan: you probably don't remember me, we played with picofuck a few years ago
00:36:09 -!- Phantom_Hoover has quit (Remote host closed the connection).
00:36:40 -!- b_jonas has quit (Quit: leaving).
00:42:19 <oerjan> hum i admit i don't remember. are you the same as orby on the wiki?
00:45:21 <orbitaldecay> it was a long time ago and I have been completely abscent since then
00:45:42 <orbitaldecay> I have a tendency to cycle through communities every couple years :)
00:47:19 <oerjan> . o O ( so you're orbiting? )
00:48:07 <oerjan> but if the orbit is decaying, it should be speeding up.
00:49:02 -!- tromp has quit (Remote host closed the connection).
00:49:14 <oerjan> ...or is the opposite also decay, like i understand the moon did before settling into always facing earth
00:50:04 <orbitaldecay> ha! yes, my rate of cycling is increasing with age
00:50:15 <orbitaldecay> at my death, I will be in all communities simultaneously
00:51:20 <orbitaldecay> Do you think there's any way to get tc out of https://esolangs.org/wiki/Poolshark ?
00:55:49 * oerjan sidles over to the wiki
01:02:38 <oerjan> orbitaldecay: i suspect poolshark is "trivially" TC if you allow the boundary to be a suitable uncomputable curve (basically, just encode the answer you want for each program).
01:02:54 <oerjan> but that's so trivial it's not very interesting.
01:03:15 <orbitaldecay> yeah, I thought of that too, but no that is not interesting
01:04:24 <orbitaldecay> For any output it's trivial to construct a poolshark language that generates it, but the interesting question is whether or not a poolshark language exists that can generate any computable output
01:06:16 <orbitaldecay> You get neat sequences from the unit square once you start feeding it irrational numbers
01:06:57 <orbitaldecay> but I don't understand it well enough yet to say if there's some construction that can generate arbitrary computable sequences
01:07:46 <oerjan> have you seen trajedy?
01:08:19 <oerjan> that's a tiny bit similar
01:14:22 <int-e> So, conjecture... this is TC if the boundary of D is composed of straight line and parabola segments.
01:15:45 <orbitaldecay> int-e: hmm, can you elaborate? how would the parabola and line segment map to output?
01:16:16 <oerjan> i also had a hunch conic sections would help
01:16:17 <int-e> (I want to work with state encoded in the offset of a bundlet of parallel rays; and I think two parabola segments can be used to compress or expand such a bundle, by making one of them a scaled version of the other with the scaling centered at the focal point)
01:17:41 <oerjan> perhaps you could even simulate the rational transformations trajedy uses
01:18:05 <orbitaldecay> the idea behind poolshark comes from the field of dynamical billiards where there is a thing called a bunimovich stadium which might be useful https://en.wikipedia.org/wiki/Dynamical_billiards
01:18:08 <int-e> orbitaldecay: I have not thought about output; it's really a secondary concern to my mind (if everything else fails one can just look at whether the thing halts on a particular output or not.
01:18:52 -!- tromp has joined.
01:21:06 <int-e> But there are details to be worked out; what I have is only obviously enough for a stack machine.
01:21:53 <orbitaldecay> yeah, this is a good start. Ixnay on the bunimovich stadium is too smooth, no halting
01:22:16 <int-e> Trajedy has this delicate feature that a ray passes through gaps between adjacent mirrors... is something similar possible in Poolshark? (Let D consist of two closed squares adjoint at a corner, and aim a ray directly at that corner... does the program halt there and then or does it pass through to the other side?)
01:22:49 <orbitaldecay> that is a good question. let me think about it for a minute.
01:22:53 <oerjan> int-e: you don't need the _second_ derivative to be smooth, so...
01:23:10 -!- tromp has quit (Ping timeout: 246 seconds).
01:23:17 -!- S_Gautam has quit (Quit: Connection closed for inactivity).
01:23:34 <int-e> oerjan: yeah but it's too far into the night for me to wreck my brain like this... so I'm sticking to something smooth at simple :)
01:24:42 <int-e> I'm happy that two parabolas make a linear map. I thought about using ellipses (jumping from focal point to focal point) and decided that I wouldn't want to compose the resulting nonlinear maps
01:25:11 <oerjan> i was imagining two curves that join with the meeting point having a derivative without it being continuous... but i think that may require the limit not to exist
01:26:46 <int-e> is the wikipedia link worth adding to the wiki page?
01:27:37 <int-e> Anyway. I need to sleep, then work. Perhaps I'll revisit this in the evening :)
01:28:49 <esowiki> [[Poolshark]] M https://esolangs.org/w/index.php?diff=60174&oldid=60173 * Orby * (+139)
01:30:08 <HackEso> /srv/hackeso-code/multibot_cmds/lib/limits: line 5: exec: ello: not found
01:30:40 <HackEso> 4699:2014-07-27 <ellioẗt> ` rm bin/{node,ello} \ 4071:2013-11-22 <mrhmous̈e> chmod +x bin/ello \ 4070:2013-11-22 <mrhmous̈e> mv raw.php* bin/ello \ 4068:2013-11-22 <mrhmous̈e> chmod +x bin/ello \ 4067:2013-11-22 <mrhmous̈e> mv raw.php* bin/ello \ 4058:2013-11-22 <mrhmous̈e> echo "(function(){var e,l,o,t;t=process.argv[2],(null!=t?t.length:void 0)||(console.log(\'Usage: ello <name>\'),process.exit()),o=/(.*)(e)$/i,e=/(.*)([bcdfghjklmnpqrstvwxz])([o0]
01:30:42 <HackEso> bin/h! \ bin/?h \ bin/thanks \ bin/hyphenate.fi \ bin/hi \ bin/blessyou \ bin/unh \ bin/?hh \ bin/shmify \ bin/h
01:30:59 <HackEso> bin/ello//#!/usr/bin/env node \ // Generated by CoffeeScript 1.6.2 \ (function() { \ var consonant_then_o, ell_manglable, ends_with_consonant, ends_with_consonant_then_vowel, name, starts_with_o; \ \ name = process.argv[2]; \ \ if (!(name != null ? name.length : void 0)) { \ console.log('Usage: ello <name>'); \ process.exit(); \ } \ \ consonant_then_o = /(.*)([bcdfghjklmnpqrstvwxz])([o0].*)/i; \ \ ends_with_consonant = /[bcdfghjklmn
01:31:14 <esowiki> [[Poolshark]] https://esolangs.org/w/index.php?diff=60175&oldid=60174 * Orby * (+105) /* See also */
01:37:19 <oerjan> right, the mean value theorem implies that if a function is differentiable in an open interval, then the derivative there must be an accumulation point of the surrounding derivatives
01:37:46 <shachaf> Did you know integration of computable functions on computable reals is computable?
01:38:00 <oerjan> so if there's a limit of derivatives toward a point, it must be the derivative there if it exists
01:38:07 <shachaf> I heard that it was possible but didn't know how it worked.
01:42:05 <oerjan> orbitaldecay: which means you cannot really have a single exceptional but non-halting point at the meeting of two parabolas
01:42:19 <shachaf> It's a bit trickier than it seems at first.
01:42:40 <shachaf> In particular you can bound a computable function : [0,1] -> R
01:43:50 <oerjan> shachaf: the trickiness i see is, what if it blows up when you get close to a _non_-computable point?
01:45:24 <oerjan> well if you want to confirm it's bounded i'd imagine you'd want to cover [0,1] by intervals where it can be approximated in each
01:45:39 <oerjan> but how do you know those cover everything, even non-computable points?
01:45:42 <shachaf> I mean, all functions are bounded.
01:46:55 <oerjan> i know computable functions are supposedly automatically continuous
01:47:08 <oerjan> but it occurs to me that that's only obvious at computable argument points
01:48:58 <shachaf> But f is a function on computable reals
01:51:03 <oerjan> yes, but not having it on non-computable points means you cannot apply the classical boundedness theorems directly
01:51:28 <esowiki> [[Poolshark]] M https://esolangs.org/w/index.php?diff=60176&oldid=60175 * Orby * (+13) /* Slope of 1 */ making clear what is not
01:52:06 <oerjan> so i guess what i'm saying then is that i don't see why they're automatically bounded, in that case
01:52:47 <shachaf> Yes, it wasn't obvious to me that you can compute the bounds.
01:53:27 <shachaf> In particular it seemed like if you can just sample f at different points to different precisions, that doesn't tell you whether it blows up somewhere else.
01:53:37 <oerjan> yeah but it's not obvious to me that they even exist non-constructively
01:57:28 <oerjan> hm i suppose you'd compute in parallel
01:57:56 <oerjan> and then if things blow up, you'd show you somehow get a constructive counterexample out of it
01:59:08 <shachaf> see https://pdfs.semanticscholar.org/d739/04971c7e9cd54f5fc12a054feda0ef74b0ba.pdf hth
01:59:22 <shachaf> it doesn't even need to be parallel?
02:03:23 <oerjan> that'll be the second math paper i give up reading this week hth
02:03:53 <oerjan> my brain doesn't have the stamina for wading through long technical stuff any longer
02:04:14 <shachaf> Anyway one of the things is, when you ask f for the value of f(x) to some precision, it asks you about the value of x to some precision(s).
02:04:40 <shachaf> And you know how much precision it asked you for, which you can use for this.
02:04:53 <oerjan> that part i actually understood
02:06:30 <oerjan> if f is actually defined on non-computable computable real numbers :P then that is enough.
02:06:46 <esowiki> [[Poolshark]] M https://esolangs.org/w/index.php?diff=60177&oldid=60176 * Orby * (+0) /* See also */
02:07:21 -!- xkapastel has quit (Quit: Connection closed for inactivity).
02:08:28 <oerjan> (by which i mean, functions from precision to approximation that are not actually computable)
02:09:15 <shachaf> I think those are just called reals, though, not computable reals.
02:47:04 -!- tromp has joined.
02:51:40 -!- tromp has quit (Ping timeout: 250 seconds).
02:51:42 <shachaf> Did you know you can define typedefs to function types in C?
02:51:53 <shachaf> typedef int foo(int, int);
02:56:36 <esowiki> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=60178&oldid=60158 * Orangeyy * (+246)
02:59:45 <esowiki> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=60179&oldid=60178 * Orangeyy * (+83)
03:04:19 <esowiki> [[User:Timtomtoaster]] https://esolangs.org/w/index.php?diff=60180&oldid=60165 * Orangeyy * (+27)
03:04:28 <esowiki> [[User:Timtomtoaster]] https://esolangs.org/w/index.php?diff=60181&oldid=60180 * Orangeyy * (+1)
03:07:49 <fizzie> Sure, I've never understood why APIs tend to use typedefs of pointers to functions rather than the function types.
03:08:50 <fizzie> `typedef void sighandler_t(int); sighandler_t *signal(int signum, sighandler_t *handler);` would be so much better than `typedef void (*sighandler_t)(int); sighandler_t signal(int signum, sighandler_t handler);`
03:08:50 <HackEso> /srv/hackeso-code/multibot_cmds/lib/limits: line 5: exec: typedef: not found
03:09:23 <fizzie> It would be even gooder if you could use the function typedef to define functions, but I guess that's complicated re parameter names.
03:09:47 <fizzie> I'm guessing you can use them for declarations that aren't a definition, though.
03:13:22 <esowiki> [[User:Orangeyy]] N https://esolangs.org/w/index.php?oldid=60182 * Orangeyy * (+103) Created page with "Some kid who can kinda program Formerly [[User:Timtomtoaster]]. I might make something new eventually."
03:20:28 -!- FreeFull has quit.
03:23:33 -!- orbitaldecay has quit (Ping timeout: 256 seconds).
03:57:54 <esowiki> [[Esolang:Introduce yourself]] M https://esolangs.org/w/index.php?diff=60183&oldid=60179 * Oerjan * (-2) wut
03:58:11 <oerjan> esowiki: you lost my pun tdnh
04:27:18 -!- aloril_ has quit (Remote host closed the connection).
04:32:07 -!- copumpkin has quit (Ping timeout: 245 seconds).
04:35:01 -!- tromp has joined.
04:35:56 -!- aloril has joined.
04:39:34 -!- tromp has quit (Ping timeout: 244 seconds).
05:29:55 -!- aloril has quit (Remote host closed the connection).
05:33:05 -!- aloril has joined.
05:42:59 -!- arseniiv has joined.
06:23:13 -!- tromp has joined.
06:27:40 -!- tromp has quit (Ping timeout: 246 seconds).
07:15:52 <oerjan> i'm sure if moloch von zinzer had been there, he'd have wondered how none of them expected this.
07:28:11 <oerjan> well maybe the jägers did
07:41:42 -!- sebbu has quit (Read error: Connection reset by peer).
07:42:05 -!- sebbu has joined.
07:57:04 -!- tromp has joined.
08:22:28 -!- arseniiv has quit (Ping timeout: 246 seconds).
08:26:41 -!- arseniiv has joined.
08:29:16 -!- oerjan has quit (Quit: Nite).
08:37:49 -!- S_Gautam has joined.
08:49:18 -!- orbitaldecay has joined.
09:02:59 -!- orbitaldecay has quit (Ping timeout: 256 seconds).
09:14:19 -!- AnotherTest has joined.
10:07:23 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60184&oldid=60167 * A * (+160) /* Continuation on the above by User:A */
10:08:32 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60185&oldid=60184 * A * (-38) /* Attempt by User:A */
10:45:09 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60186&oldid=60185 * A * (+293) /* Tail procedure */
10:47:29 -!- S_Gautam has quit (Quit: Connection closed for inactivity).
10:47:53 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60187&oldid=60186 * A * (+94) /* Tail procedure */
10:49:45 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60188&oldid=60187 * A * (+177) /* =Bit comparison */
11:02:50 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60189&oldid=60188 * A * (-191)
11:06:46 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60190&oldid=60189 * A * (+27) This is Turing complete if only bits are neccesary for Turing completeness.
11:06:59 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60191&oldid=60190 * A * (+2) /* References */
11:07:34 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60192&oldid=60191 * A * (+40) /* Continuation on the above by User:A */
11:09:30 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60193&oldid=60192 * A * (+7) /* References */
11:40:27 -!- AnnaSeo has joined.
12:18:14 -!- Lord_of_Life_ has joined.
12:20:26 -!- Lord_of_Life has quit (Ping timeout: 244 seconds).
12:20:26 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
12:43:18 -!- copumpkin has joined.
13:01:06 -!- S_Gautam has joined.
13:10:42 -!- AnotherTest has quit (Ping timeout: 264 seconds).
13:22:00 -!- AnnaSeo has quit (Quit: <chnamsguy> i just need to know if you are a man).
13:34:07 -!- AnotherTest has joined.
14:07:23 -!- orbitaldecay has joined.
14:10:22 <esowiki> [[Bitch]] https://esolangs.org/w/index.php?diff=60194&oldid=60193 * A * (+77) /* References */
14:23:33 -!- wob_jonas has joined.
15:31:01 -!- Essadon has joined.
15:51:27 -!- moei has joined.
16:57:58 -!- mniip has quit (Ping timeout: 600 seconds).
16:58:58 -!- xkapastel has joined.
17:13:56 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
17:16:22 -!- mniip has joined.
17:16:29 -!- vertrex has quit (Ping timeout: 244 seconds).
17:16:40 -!- vertrex has joined.
17:16:40 -!- vertrex has quit (Changing host).
17:16:40 -!- vertrex has joined.
17:23:32 -!- tromp has quit (Remote host closed the connection).
17:42:50 -!- AnotherTest has quit (Ping timeout: 244 seconds).
17:47:34 -!- LKoen has joined.
17:53:29 -!- tromp has joined.
17:55:08 -!- tromp has quit (Remote host closed the connection).
17:56:07 -!- tromp has joined.
18:01:52 -!- AnotherTest has joined.
18:11:31 -!- LKoen has quit (Remote host closed the connection).
18:30:48 -!- tromp has quit (Remote host closed the connection).
18:40:18 <esowiki> [[Caeos]] https://esolangs.org/w/index.php?diff=60195&oldid=60161 * Areallycoolusername * (+2)
18:51:50 -!- b_jonas has joined.
18:52:07 -!- tromp has joined.
18:52:08 <int-e> fizzie: aww esowiki does not allow svg?
18:53:54 <esowiki> [[Special:Log/upload]] upload * Int-e * uploaded "[[File:Sharkscale.png]]": scaling a shark
18:58:00 -!- FreeFull has joined.
18:59:14 <esowiki> [[Talk:Poolshark]] N https://esolangs.org/w/index.php?oldid=60197 * Int-e * (+206) scaling sharks
19:03:09 <orin> why does "git rebase so slow"
19:03:36 <orin> er, why does git "rebase" so slowly
19:06:53 -!- LKoen has joined.
19:23:11 <kmc> are you rebasing past a lot of commits?
19:23:20 <kmc> it essentially has to apply each diff one by one
19:28:25 <orin> kmc: it would be nice if it had a loading bar then
19:28:56 <orin> rewinding head... 222/123123
19:28:58 <kmc> there's probably an obscure flag for it
19:29:07 <kmc> yeah if you want to rebase past 100,000 commits you're gonna have a bad time
19:29:28 <kmc> orin: some people run their git repo on a ramdisk for faster operations
19:29:33 <kmc> I have not found an improvement that way
19:29:38 <kmc> I think Git is not terribly optimized
19:29:45 <kmc> a lot of it is shell scripts or used to be
19:30:16 <shachaf> it is a shame so many people have settled on git when there are plenty of improvements to be made in version control
19:32:30 <HackEso> olist 1157: shachaf oerjan Sgeo FireFly boily nortti b_jonas
19:34:32 -!- arseniiv has quit (Ping timeout: 255 seconds).
19:42:07 <esowiki> [[Special:Log/upload]] upload * Int-e * uploaded "[[File:Magicsharkmirror.png]]": magic mirror for directing sharks
19:45:24 <fizzie> int-e: Not by intention, just by accident.
19:45:39 <fizzie> int-e: From a quick glance, it's just that it's not an enabled-by-default MediaWiki feature.
19:46:45 <fizzie> It'd probably be possible to enable it.
19:47:38 <esowiki> [[Talk:Poolshark]] https://esolangs.org/w/index.php?diff=60199&oldid=60197 * Int-e * (+765) enter the magic mirror
19:48:49 <int-e> It's not a big deal... svg just feels smoother. but of course it's also a way to transport Javascript... so perhaps it's better to leave it turned off.
19:49:40 <int-e> Though that's a very sad reason.
19:49:52 <int-e> But valid, I'm afraid.
19:53:36 <fizzie> MediaWiki blocks any JS in SVG even when it's enabled.
19:53:44 <fizzie> Of course that's assuming they've managed to catch it all.
19:53:46 <esowiki> [[Talk:Poolshark]] https://esolangs.org/w/index.php?diff=60200&oldid=60199 * Int-e * (+225) fewer obstructions
19:55:15 -!- Phantom_Hoover has joined.
20:01:29 <fizzie> @tell ais523 Re canonical way to clear top 32 bits of eax, looks like in at least some contexts GCC will use "mov eax, eax" for that, which makes sense.
20:07:21 <shachaf> I added a thing to my status bar that shows me the size in bytes of the primary and clipboard selection, whether they contain any newlines, and whether they are equal.
20:07:35 <shachaf> And also a keybinding to clear them both.
20:07:56 <fizzie> That sounds pretty good.
20:08:09 <fizzie> Especially the newline indicator, maybe I could dare to paste from PDFs to IRC with one of those.
20:08:38 <shachaf> Daring to paste to IRC was the goal.
20:09:02 <shachaf> imo can you believe we're still using these archaic terminal uis that don't support paste properly
20:09:05 <esowiki> [[Talk:Poolshark]] https://esolangs.org/w/index.php?diff=60201&oldid=60200 * Int-e * (+237)
20:12:33 <orin> god damn it it has been 2 hours
20:12:49 <int-e> hmm hmm. "ETSI as a global standards body has been engaged for the past three years in developing Transport Layer Security implementation technical standards"
20:18:20 <orin> spacex rocket launch in 11hr30min
20:21:36 <kmc> i usually paste into url bar first
20:21:40 <kmc> to erase newlines
20:21:58 <kmc> orin: it's not giving you a progress indication?
20:22:03 <kmc> it usually does...
20:22:07 <kmc> are you sure you don't want to merge instead?
20:22:20 <kmc> anyway you could look at how many commits there are past the rebase point as a progress
20:25:12 <shachaf> As far as I can tell it's not really possible to watch for clipboard changes in X?
20:25:30 <fizzie> int-e: Is this about the eTLS scam?
20:27:00 <int-e> where "e" stands for "backdoor-ready", I believe ;-)
20:27:12 <int-e> (at least that has a nice buzzy ring to it)
20:27:27 <int-e> (and it's far more accurate than "enterprise")
20:28:20 <int-e> They also write that "security capabilities [will] go dark". https://csrc.nist.gov/CSRC/media/Publications/sp/800-52/rev-2/draft/documents/sp800-52r2-draft2-comments-received.pdf#page=14
20:29:29 <kmc> shachaf: do you poll?
20:29:38 <kmc> shachaf: I think X has various extensions that let you sniff all messages?
20:30:37 <shachaf> I just poll once a second, yep.
20:30:41 -!- S_Gautam has quit (Quit: Connection closed for inactivity).
20:30:57 <shachaf> That seems kind of excessive for a small feature. But maybe it's possible?
20:31:07 <b_jonas> kmc: and if two clients try that, then they'll be spinning each otwards their nose, like when two dogs mutually try to sniff each other's butt
20:31:54 <shachaf> Sniffing messages doesn't generate messages presumably.
20:32:15 <b_jonas> but kmc said "sniff all messages"
20:32:49 <kmc> snort all messages
20:32:59 <kmc> you would have to be careful to ignore your own set-clipboard messages, though
20:34:34 <shachaf> and in theory it's a protocol but you can't even implement it yourself if you want opengl or whatever
20:34:51 <shachaf> so you gotta use xlib which is a double mess
20:36:43 <int-e> isn't there an alternative mess (xcb)?
20:37:38 <shachaf> Should I use xcb instead of xlib?
20:37:57 <shachaf> You still need xlib for glx but maybe you can do the rest with xcb.
20:42:54 <int-e> I don't know. I've written one xcb program and that was ages ago... I think the main point was to allow more asynchronicity which is rather pointless for small programs.
20:43:18 <shachaf> What kind of asynchronicity?
20:43:42 <int-e> It does seem to do GLX though. https://xcb.freedesktop.org/manual/group__XCB__Glx__API.html
20:45:01 <int-e> Well, submitting several requests simultaneously and process replys as they come along the socket. A lot of the libX APIs are synchronous at first; a function will make a request to the X server and wait for the reply.
20:46:55 <shachaf> I thought I'd be good just polling on the file descriptor.
20:48:28 <int-e> for most requests, xcb hands out cookied that you can wait on later. http://paste.debian.net/1070921/
20:50:36 <int-e> It's very low level... I'm not sure I'd really want to go there :P
20:54:38 <int-e> hah, http://www.inkscapeforum.com/images/smilies/tool_pen.gif ... treating tool icons as smileys :)
20:55:07 <shachaf> What should I use instead?
20:55:26 <shachaf> By the way, did you know Gtk renders widgets using CSS nowadays?
20:56:08 <int-e> Hmm, vaguely. I remember tweaking an evince color with CSS at some point.
20:56:47 <b_jonas> shachaf: isn't it Qt that does that?
20:57:37 <int-e> it's not all that surprising
20:57:51 <int-e> (once you allow themes)
21:03:59 <esowiki> [[Special:Log/upload]] overwrite * Int-e * uploaded a new version of "[[File:Magicsharkmirror.png]]": add "bad" circle
21:05:50 <esowiki> [[Talk:Poolshark]] https://esolangs.org/w/index.php?diff=60203&oldid=60201 * Int-e * (+219)
21:09:26 <esowiki> [[Talk:Poolshark]] M https://esolangs.org/w/index.php?diff=60204&oldid=60203 * Int-e * (+1) typo
21:12:44 <orbitaldecay> going to need to think about this, brain is a little tired atm, but thank you for articulating your idea!
21:13:08 <esowiki> [[ALLSCII]] N https://esolangs.org/w/index.php?oldid=60205 * Cortex * (+451) Created page with "'''ALLSCII''' is a stack-based esolang created by [[User:Cortex|]] where every printable ASCII character is (will be) a valid command. == Commands == {| class="wikitable" |-..."
21:14:28 -!- neanias has left.
21:18:58 <int-e> orbitaldecay: I guess my main point in the "magic mirror" construction is that without smoothness, you may have too much power built into a single reflection already. OTOH, this enables a divisibility test so reducing from Fractran becomes a breeze, which seems fair enough... (the region D and all reflections remain computable; TCness is emerging from doing an unbounded number of reflections.)
21:20:07 <int-e> I don't understand Mediawiki's caching.
21:20:30 <b_jonas> if in doubt, blame caching
21:20:51 <int-e> well, (https://esolangs.org/wiki/Talk:Poolshark is showing an old image version for me, after showing the right one a couple of minutes ago.
21:20:56 <orbitaldecay> got it. excellent. I appreciate your thoughts :)
21:22:10 <int-e> Restarted the browser, now it's fine... (too many caches is what I'm blaming)
21:22:27 <shachaf> int-e: I see, X calls can just block waiting for a reply whenever they want?
21:23:05 <int-e> shachaf: I think so.
21:25:34 <int-e> orbitaldecay: and the construction is a variation on an analysis (or calculus) exercise... construct a function which is differentiable on the rationals, with an arbitrary function Q -> R as the partial derivative. (Q can be replaced by a countable subset of the reals)
21:26:02 <int-e> hmm. I should not use "partial derivative" like that :P
21:26:17 <int-e> (I mean it's a partial function...)
21:26:57 <orbitaldecay> yeah, I vaguely recall doing something like that in an introductory real analysis course once
21:29:46 <orbitaldecay> I've been out of school for too long. My brain is rotting haha
21:30:00 <orbitaldecay> going through this will be a good exercise for the evening
21:31:49 <Phantom_Hoover> me too obviously since my intuition says that Q is dense in R so you can't arbitrarily pick derivatives on all points in Q like that
21:32:42 <shachaf> int-e: Man, the xcb API looks big and complicated.
21:32:53 <shachaf> Is this the canonical way to do things?
21:32:59 <int-e> shachaf: blame the X protocol :)
21:33:10 -!- shikhin has changed nick to shi.
21:33:21 <int-e> no, the canonical way is to use a toolkit library that hides all the ugliness (and adds some of its own instead)
21:33:47 <b_jonas> shachaf: the canonical way is to use xcb for negotiating and bootstrapping saner protocols
21:34:14 <int-e> Dunno. GTK is the only one I've programmed with.
21:34:34 <b_jonas> probably xkb and whatever that other one is for input, and some protocol for mapping some video memory and drawing straight on them for output
21:34:42 <shachaf> I tried to write some GTK code and it was so demoralizing I gave up programming.
21:35:27 <b_jonas> I don't know how it all works
21:35:38 <b_jonas> all the toolkits suck these days, but you can't do much about that
21:37:00 <int-e> The postmodern approach is to use an HTML renderer. (Electron...)
21:37:12 <shachaf> That is too postmodern for me.
21:39:08 <int-e> SDL is nice for what it does.
21:46:00 -!- shi has changed nick to shikhin.
21:46:34 <shachaf> int-e: One of the things I don't get about these APIs is why they're full of mallocs.
21:47:41 <int-e> they tend to be OO, and "malloc" is how you do "new" in C?
21:49:41 <shachaf> but why would a low-level api be OO
21:50:45 <int-e> uh if you did not refer to toolkit libraries maybe you should be more specific than writing "these"
21:52:08 <int-e> (toolkit libraries tend to think of themselves as highlevel)
21:53:55 <orin> THREEEE HOURS!!!!!
21:54:20 <orin> First, rewinding head to replay your work on top of it...
21:54:28 <orin> AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
21:56:56 <int-e> shachaf: but if you meant xcb... the X protocol has some variable length replies, so hiding the allocation in the library is *good*. for uniformity reasons it's reasonable to treat all the other data in the same way.
21:57:29 <shachaf> Can't it just tell me the size and let me allocate?
22:00:51 <b_jonas> shachaf: you can just implement that yourself directly
22:02:14 <shachaf> And the glX calls take a Display * argument so I don't think they can be used with my own implementation of the protocol?
22:25:11 -!- moei has quit (Quit: Leaving...).
22:44:39 -!- J_Arcane_ has joined.
22:45:25 -!- ocharles_ has joined.
22:52:36 -!- Cale has quit (*.net *.split).
22:52:38 -!- ocharles has quit (*.net *.split).
22:52:38 -!- J_Arcane has quit (*.net *.split).
22:52:42 -!- J_Arcane_ has changed nick to J_Arcane.
22:52:55 -!- ocharles_ has changed nick to ocharles.
22:56:47 <orin> FOUR HOUUUURRRRRS
22:59:03 -!- Cale has joined.
23:05:03 -!- orbitaldecay has quit (Ping timeout: 256 seconds).
23:12:09 -!- orbitaldecay has joined.
23:12:37 <orbitaldecay> int-e: apparently a method for the construction of continuous functions that are only differentiable at rational points wasn't discovered until 2013 (!)
23:13:01 <orbitaldecay> so we're not in completely uncharted territory
23:15:24 <orbitaldecay> doing it the other way, constructing a continuous function that is only differentiable at irrational points, is apparently easy
23:16:25 <int-e> yeah I can do that by adding sawtooth functions. (period 1/n, amplitude 1/n^2, sum over all n)
23:17:20 <int-e> or step functions to get something that is discontinuous at all rational numbers, and has derivative 0 elsewhere
23:18:20 <int-e> but fortunately, for that magic mirror I just don't care what happens for irrational offsets
23:18:41 <orbitaldecay> yeah, I think the easiest might be to have it not even be differentiable at rational points so it just halts
23:19:13 -!- Melvar has quit (Quit: rebooting).
23:27:31 -!- Melvar has joined.
23:39:20 <HackEso> wc: invalid option -- ' ' \ Try 'wc --help' for more information.
23:39:23 <HackEso> /srv/hackeso-code/multibot_cmds/lib/limits: line 5: exec: `wc: not found
23:45:58 <kmc> everything is buggy shit. nobody gives a shit if anything works. people don't give a fuck about picking good tools. the claim that software as practice today is "engineering" is completely embarrassing
23:46:16 <kmc> standards are so low and everyone is just okay with it for some reason
23:47:07 <kmc> this is one of the reasons i'm giving up on software
23:47:38 <kmc> fuck it, and fuck everything
23:51:53 <Phantom_Hoover> most notably that the stakes in software are, i think, actually quite low
23:52:48 <Phantom_Hoover> certainly lower than stuff that has historically been called engineering
23:53:19 <kmc> if software came with guarantees and you could sue
23:53:25 <kmc> then companies would have to buy insurance
23:53:33 <kmc> and if their practices are shit then their insurance will be too expensive
23:53:39 <kmc> but the software would also be more expensive
23:54:42 <Phantom_Hoover> sure but i think the reason that doesn't happen is that it's actually not that important when software breaks
23:55:25 <Phantom_Hoover> ive particularly noticed that with security stuff where, like, i feel there's this gulf between what security researchers say which is that security Matters and is Immensely Important
23:56:03 <Phantom_Hoover> and the real world in which bash, openssl and every fucking cpu on the planet all had security holes you could drive a truck through for a decade and it didn't matter until security researchers kicked up a fuss over it
23:57:51 <kmc> it mattered to a bunch of people who got hacked and their lives ruined
23:58:16 <kmc> it doesn't matter to average joe but it's one of the main drivers behind the modern panopticon authoritarian state
23:59:18 <Phantom_Hoover> and the panopticon authoritarian state is a political problem that must be solved with political action
23:59:31 <kmc> yeah but we don't have to make it so easy for them